{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.6455026455026456,
  "eval_steps": 500,
  "global_step": 3500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "grad_norm": 1.7509137392044067,
      "learning_rate": 9.974804736709499e-06,
      "loss": 1.9586,
      "step": 10
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.5435916185379028,
      "learning_rate": 9.949609473418998e-06,
      "loss": 1.764,
      "step": 20
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.209004521369934,
      "learning_rate": 9.924414210128497e-06,
      "loss": 1.5738,
      "step": 30
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.036686897277832,
      "learning_rate": 9.899218946837995e-06,
      "loss": 1.456,
      "step": 40
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.8039250373840332,
      "learning_rate": 9.874023683547493e-06,
      "loss": 1.3517,
      "step": 50
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.668268620967865,
      "learning_rate": 9.848828420256992e-06,
      "loss": 1.2773,
      "step": 60
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.6465680003166199,
      "learning_rate": 9.823633156966492e-06,
      "loss": 1.2494,
      "step": 70
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.669222891330719,
      "learning_rate": 9.79843789367599e-06,
      "loss": 1.2465,
      "step": 80
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.7393785715103149,
      "learning_rate": 9.773242630385489e-06,
      "loss": 1.2073,
      "step": 90
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.6821632385253906,
      "learning_rate": 9.748047367094986e-06,
      "loss": 1.1725,
      "step": 100
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.698268711566925,
      "learning_rate": 9.722852103804486e-06,
      "loss": 1.1724,
      "step": 110
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.6640620231628418,
      "learning_rate": 9.697656840513983e-06,
      "loss": 1.1656,
      "step": 120
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.8017461895942688,
      "learning_rate": 9.672461577223483e-06,
      "loss": 1.1764,
      "step": 130
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.6652551293373108,
      "learning_rate": 9.64726631393298e-06,
      "loss": 1.1706,
      "step": 140
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.8237988352775574,
      "learning_rate": 9.62207105064248e-06,
      "loss": 1.1231,
      "step": 150
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.8982694149017334,
      "learning_rate": 9.59687578735198e-06,
      "loss": 1.1417,
      "step": 160
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.7859475016593933,
      "learning_rate": 9.571680524061477e-06,
      "loss": 1.1382,
      "step": 170
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.9249159693717957,
      "learning_rate": 9.546485260770976e-06,
      "loss": 1.1363,
      "step": 180
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.221252202987671,
      "learning_rate": 9.521289997480474e-06,
      "loss": 1.1382,
      "step": 190
    },
    {
      "epoch": 0.15,
      "grad_norm": 4.636524200439453,
      "learning_rate": 9.498614260519024e-06,
      "loss": 1.0979,
      "step": 200
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.7521413564682007,
      "learning_rate": 9.473418997228522e-06,
      "loss": 1.0645,
      "step": 210
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.8328731060028076,
      "learning_rate": 9.44822373393802e-06,
      "loss": 1.0585,
      "step": 220
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.8342303037643433,
      "learning_rate": 9.423028470647519e-06,
      "loss": 1.0655,
      "step": 230
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.8196636438369751,
      "learning_rate": 9.397833207357018e-06,
      "loss": 1.0221,
      "step": 240
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.7511851787567139,
      "learning_rate": 9.372637944066516e-06,
      "loss": 1.046,
      "step": 250
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.8341330885887146,
      "learning_rate": 9.347442680776014e-06,
      "loss": 1.0331,
      "step": 260
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.9003998637199402,
      "learning_rate": 9.322247417485513e-06,
      "loss": 1.0386,
      "step": 270
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.79107266664505,
      "learning_rate": 9.297052154195013e-06,
      "loss": 1.0152,
      "step": 280
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.7756459712982178,
      "learning_rate": 9.27185689090451e-06,
      "loss": 1.0224,
      "step": 290
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.762424647808075,
      "learning_rate": 9.24666162761401e-06,
      "loss": 1.0517,
      "step": 300
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.7143459916114807,
      "learning_rate": 9.221466364323507e-06,
      "loss": 1.0106,
      "step": 310
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.7883962988853455,
      "learning_rate": 9.196271101033007e-06,
      "loss": 1.0004,
      "step": 320
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.7985430359840393,
      "learning_rate": 9.171075837742504e-06,
      "loss": 1.0171,
      "step": 330
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.8012735247612,
      "learning_rate": 9.145880574452004e-06,
      "loss": 1.0106,
      "step": 340
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.7331275939941406,
      "learning_rate": 9.120685311161502e-06,
      "loss": 1.0044,
      "step": 350
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.8667876124382019,
      "learning_rate": 9.095490047871001e-06,
      "loss": 1.0195,
      "step": 360
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.8839796781539917,
      "learning_rate": 9.0702947845805e-06,
      "loss": 1.0145,
      "step": 370
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.8900341391563416,
      "learning_rate": 9.045099521289998e-06,
      "loss": 1.0171,
      "step": 380
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.8628876805305481,
      "learning_rate": 9.019904257999496e-06,
      "loss": 1.0039,
      "step": 390
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.8550145626068115,
      "learning_rate": 8.994708994708995e-06,
      "loss": 1.015,
      "step": 400
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.8471419215202332,
      "learning_rate": 8.969513731418495e-06,
      "loss": 1.0114,
      "step": 410
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.8051427006721497,
      "learning_rate": 8.944318468127992e-06,
      "loss": 0.9998,
      "step": 420
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.8776746392250061,
      "learning_rate": 8.919123204837492e-06,
      "loss": 1.0109,
      "step": 430
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.9933453798294067,
      "learning_rate": 8.893927941546991e-06,
      "loss": 1.0099,
      "step": 440
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.8829633593559265,
      "learning_rate": 8.868732678256489e-06,
      "loss": 1.0201,
      "step": 450
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.8752363324165344,
      "learning_rate": 8.843537414965987e-06,
      "loss": 1.0022,
      "step": 460
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.9361273050308228,
      "learning_rate": 8.818342151675486e-06,
      "loss": 1.0075,
      "step": 470
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.8231716156005859,
      "learning_rate": 8.793146888384985e-06,
      "loss": 0.984,
      "step": 480
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.8183861374855042,
      "learning_rate": 8.767951625094483e-06,
      "loss": 0.9871,
      "step": 490
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.9514871835708618,
      "learning_rate": 8.74275636180398e-06,
      "loss": 1.024,
      "step": 500
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.9753857254981995,
      "learning_rate": 8.71756109851348e-06,
      "loss": 0.9926,
      "step": 510
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.9237353205680847,
      "learning_rate": 8.69236583522298e-06,
      "loss": 0.9709,
      "step": 520
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.9065663814544678,
      "learning_rate": 8.667170571932477e-06,
      "loss": 0.9848,
      "step": 530
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.8715184926986694,
      "learning_rate": 8.641975308641975e-06,
      "loss": 1.0091,
      "step": 540
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.8322390913963318,
      "learning_rate": 8.616780045351474e-06,
      "loss": 0.9876,
      "step": 550
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.9560284614562988,
      "learning_rate": 8.591584782060974e-06,
      "loss": 1.0234,
      "step": 560
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.8381783366203308,
      "learning_rate": 8.566389518770471e-06,
      "loss": 0.9788,
      "step": 570
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.8607673645019531,
      "learning_rate": 8.541194255479971e-06,
      "loss": 0.9816,
      "step": 580
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.8169811367988586,
      "learning_rate": 8.515998992189469e-06,
      "loss": 1.0092,
      "step": 590
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.9223540425300598,
      "learning_rate": 8.490803728898968e-06,
      "loss": 0.9885,
      "step": 600
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.7612660527229309,
      "learning_rate": 8.465608465608466e-06,
      "loss": 0.97,
      "step": 610
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.860916256904602,
      "learning_rate": 8.440413202317965e-06,
      "loss": 0.9921,
      "step": 620
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.807750403881073,
      "learning_rate": 8.415217939027463e-06,
      "loss": 0.987,
      "step": 630
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.7962011694908142,
      "learning_rate": 8.390022675736962e-06,
      "loss": 1.0134,
      "step": 640
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.8323624730110168,
      "learning_rate": 8.364827412446462e-06,
      "loss": 0.9769,
      "step": 650
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.9545472860336304,
      "learning_rate": 8.33963214915596e-06,
      "loss": 0.9862,
      "step": 660
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.0507547855377197,
      "learning_rate": 8.314436885865457e-06,
      "loss": 0.9977,
      "step": 670
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.9005000591278076,
      "learning_rate": 8.289241622574956e-06,
      "loss": 0.9897,
      "step": 680
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.8091966509819031,
      "learning_rate": 8.264046359284456e-06,
      "loss": 1.0188,
      "step": 690
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.8602038025856018,
      "learning_rate": 8.238851095993954e-06,
      "loss": 0.9997,
      "step": 700
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.9347642660140991,
      "learning_rate": 8.213655832703453e-06,
      "loss": 0.9904,
      "step": 710
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.8802648782730103,
      "learning_rate": 8.188460569412952e-06,
      "loss": 0.9765,
      "step": 720
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.8487595915794373,
      "learning_rate": 8.16326530612245e-06,
      "loss": 1.0184,
      "step": 730
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.9282223582267761,
      "learning_rate": 8.138070042831948e-06,
      "loss": 1.0047,
      "step": 740
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.0162054300308228,
      "learning_rate": 8.112874779541447e-06,
      "loss": 0.9909,
      "step": 750
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.8241488337516785,
      "learning_rate": 8.087679516250947e-06,
      "loss": 0.9976,
      "step": 760
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.069097638130188,
      "learning_rate": 8.062484252960444e-06,
      "loss": 0.9874,
      "step": 770
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.9084659218788147,
      "learning_rate": 8.037288989669942e-06,
      "loss": 0.9751,
      "step": 780
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.9863806962966919,
      "learning_rate": 8.012093726379441e-06,
      "loss": 0.9815,
      "step": 790
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.9838799834251404,
      "learning_rate": 7.98689846308894e-06,
      "loss": 1.0019,
      "step": 800
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.0254228115081787,
      "learning_rate": 7.961703199798438e-06,
      "loss": 0.9918,
      "step": 810
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.9409896731376648,
      "learning_rate": 7.936507936507936e-06,
      "loss": 0.9899,
      "step": 820
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.9909108281135559,
      "learning_rate": 7.911312673217436e-06,
      "loss": 0.9666,
      "step": 830
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.9930652976036072,
      "learning_rate": 7.886117409926935e-06,
      "loss": 0.9643,
      "step": 840
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.983561098575592,
      "learning_rate": 7.860922146636433e-06,
      "loss": 0.9895,
      "step": 850
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.9257413148880005,
      "learning_rate": 7.835726883345932e-06,
      "loss": 0.9933,
      "step": 860
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.9350329041481018,
      "learning_rate": 7.81053162005543e-06,
      "loss": 0.9725,
      "step": 870
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.0069069862365723,
      "learning_rate": 7.78533635676493e-06,
      "loss": 0.9739,
      "step": 880
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.9367712736129761,
      "learning_rate": 7.760141093474427e-06,
      "loss": 0.9835,
      "step": 890
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.867667555809021,
      "learning_rate": 7.734945830183926e-06,
      "loss": 0.9937,
      "step": 900
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.9225870370864868,
      "learning_rate": 7.709750566893424e-06,
      "loss": 0.9771,
      "step": 910
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.9846392869949341,
      "learning_rate": 7.684555303602923e-06,
      "loss": 0.9988,
      "step": 920
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.8313792943954468,
      "learning_rate": 7.659360040312423e-06,
      "loss": 0.9594,
      "step": 930
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.9007426500320435,
      "learning_rate": 7.63416477702192e-06,
      "loss": 0.9641,
      "step": 940
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.9949538111686707,
      "learning_rate": 7.608969513731419e-06,
      "loss": 0.9981,
      "step": 950
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.0600723028182983,
      "learning_rate": 7.583774250440918e-06,
      "loss": 0.9799,
      "step": 960
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.124242901802063,
      "learning_rate": 7.558578987150417e-06,
      "loss": 0.9923,
      "step": 970
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.1153829097747803,
      "learning_rate": 7.533383723859915e-06,
      "loss": 0.976,
      "step": 980
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.996911346912384,
      "learning_rate": 7.508188460569413e-06,
      "loss": 0.9952,
      "step": 990
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.866919219493866,
      "learning_rate": 7.482993197278913e-06,
      "loss": 0.9708,
      "step": 1000
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.0361676216125488,
      "learning_rate": 7.457797933988411e-06,
      "loss": 0.9757,
      "step": 1010
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.0245063304901123,
      "learning_rate": 7.432602670697909e-06,
      "loss": 0.9879,
      "step": 1020
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.9136086106300354,
      "learning_rate": 7.4074074074074075e-06,
      "loss": 0.962,
      "step": 1030
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.1000611782073975,
      "learning_rate": 7.382212144116907e-06,
      "loss": 0.9743,
      "step": 1040
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.0377795696258545,
      "learning_rate": 7.3570168808264054e-06,
      "loss": 0.9392,
      "step": 1050
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.9714633226394653,
      "learning_rate": 7.331821617535904e-06,
      "loss": 0.964,
      "step": 1060
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.027772068977356,
      "learning_rate": 7.306626354245403e-06,
      "loss": 0.9784,
      "step": 1070
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.9620116949081421,
      "learning_rate": 7.281431090954901e-06,
      "loss": 0.9803,
      "step": 1080
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.0313563346862793,
      "learning_rate": 7.2562358276644e-06,
      "loss": 0.9816,
      "step": 1090
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.058566927909851,
      "learning_rate": 7.231040564373898e-06,
      "loss": 0.9692,
      "step": 1100
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.9281766414642334,
      "learning_rate": 7.205845301083398e-06,
      "loss": 0.9508,
      "step": 1110
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.960678219795227,
      "learning_rate": 7.180650037792895e-06,
      "loss": 0.9437,
      "step": 1120
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.9975042343139648,
      "learning_rate": 7.155454774502394e-06,
      "loss": 0.97,
      "step": 1130
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.9827629327774048,
      "learning_rate": 7.130259511211892e-06,
      "loss": 0.9709,
      "step": 1140
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.084519386291504,
      "learning_rate": 7.105064247921392e-06,
      "loss": 0.9736,
      "step": 1150
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.1209474802017212,
      "learning_rate": 7.0798689846308895e-06,
      "loss": 0.9722,
      "step": 1160
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.0013864040374756,
      "learning_rate": 7.054673721340388e-06,
      "loss": 0.9671,
      "step": 1170
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.9566646218299866,
      "learning_rate": 7.0294784580498875e-06,
      "loss": 0.9767,
      "step": 1180
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.9742752909660339,
      "learning_rate": 7.004283194759386e-06,
      "loss": 0.9703,
      "step": 1190
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.1095930337905884,
      "learning_rate": 6.979087931468885e-06,
      "loss": 0.9817,
      "step": 1200
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.0295912027359009,
      "learning_rate": 6.953892668178382e-06,
      "loss": 0.9756,
      "step": 1210
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.0642352104187012,
      "learning_rate": 6.928697404887882e-06,
      "loss": 0.9503,
      "step": 1220
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.157819390296936,
      "learning_rate": 6.90350214159738e-06,
      "loss": 0.9765,
      "step": 1230
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.0977922677993774,
      "learning_rate": 6.878306878306879e-06,
      "loss": 0.9706,
      "step": 1240
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.9640308022499084,
      "learning_rate": 6.853111615016378e-06,
      "loss": 0.9578,
      "step": 1250
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.0732492208480835,
      "learning_rate": 6.827916351725876e-06,
      "loss": 0.9455,
      "step": 1260
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.9846688508987427,
      "learning_rate": 6.8027210884353745e-06,
      "loss": 0.9764,
      "step": 1270
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.8813304901123047,
      "learning_rate": 6.777525825144873e-06,
      "loss": 0.9842,
      "step": 1280
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.9812163710594177,
      "learning_rate": 6.752330561854372e-06,
      "loss": 0.9593,
      "step": 1290
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.0121666193008423,
      "learning_rate": 6.72713529856387e-06,
      "loss": 0.9623,
      "step": 1300
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.0399516820907593,
      "learning_rate": 6.701940035273369e-06,
      "loss": 0.9665,
      "step": 1310
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.9701886773109436,
      "learning_rate": 6.676744771982868e-06,
      "loss": 0.978,
      "step": 1320
    },
    {
      "epoch": 1.01,
      "grad_norm": 1.0398330688476562,
      "learning_rate": 6.651549508692367e-06,
      "loss": 0.9515,
      "step": 1330
    },
    {
      "epoch": 1.01,
      "grad_norm": 1.038935661315918,
      "learning_rate": 6.626354245401865e-06,
      "loss": 0.958,
      "step": 1340
    },
    {
      "epoch": 1.02,
      "grad_norm": 1.0119668245315552,
      "learning_rate": 6.601158982111363e-06,
      "loss": 0.9347,
      "step": 1350
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.9660181403160095,
      "learning_rate": 6.575963718820862e-06,
      "loss": 0.9577,
      "step": 1360
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.9453220963478088,
      "learning_rate": 6.550768455530361e-06,
      "loss": 0.9375,
      "step": 1370
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.9495384693145752,
      "learning_rate": 6.525573192239859e-06,
      "loss": 0.9311,
      "step": 1380
    },
    {
      "epoch": 1.05,
      "grad_norm": 1.1077016592025757,
      "learning_rate": 6.500377928949359e-06,
      "loss": 0.9489,
      "step": 1390
    },
    {
      "epoch": 1.06,
      "grad_norm": 1.0165854692459106,
      "learning_rate": 6.4751826656588565e-06,
      "loss": 0.958,
      "step": 1400
    },
    {
      "epoch": 1.07,
      "grad_norm": 1.3001985549926758,
      "learning_rate": 6.449987402368355e-06,
      "loss": 0.9325,
      "step": 1410
    },
    {
      "epoch": 1.07,
      "grad_norm": 1.0839056968688965,
      "learning_rate": 6.424792139077854e-06,
      "loss": 0.9769,
      "step": 1420
    },
    {
      "epoch": 1.08,
      "grad_norm": 1.0319207906723022,
      "learning_rate": 6.399596875787353e-06,
      "loss": 0.9574,
      "step": 1430
    },
    {
      "epoch": 1.09,
      "grad_norm": 1.0123823881149292,
      "learning_rate": 6.374401612496851e-06,
      "loss": 0.9315,
      "step": 1440
    },
    {
      "epoch": 1.1,
      "grad_norm": 1.1985503435134888,
      "learning_rate": 6.349206349206349e-06,
      "loss": 0.9354,
      "step": 1450
    },
    {
      "epoch": 1.1,
      "grad_norm": 1.1550285816192627,
      "learning_rate": 6.324011085915849e-06,
      "loss": 0.9344,
      "step": 1460
    },
    {
      "epoch": 1.11,
      "grad_norm": 1.0712833404541016,
      "learning_rate": 6.298815822625347e-06,
      "loss": 0.9444,
      "step": 1470
    },
    {
      "epoch": 1.12,
      "grad_norm": 1.0778782367706299,
      "learning_rate": 6.273620559334846e-06,
      "loss": 0.935,
      "step": 1480
    },
    {
      "epoch": 1.13,
      "grad_norm": 1.0286056995391846,
      "learning_rate": 6.2484252960443435e-06,
      "loss": 0.9606,
      "step": 1490
    },
    {
      "epoch": 1.13,
      "grad_norm": 1.1475175619125366,
      "learning_rate": 6.223230032753843e-06,
      "loss": 0.9354,
      "step": 1500
    },
    {
      "epoch": 1.14,
      "grad_norm": 1.1092008352279663,
      "learning_rate": 6.1980347694633414e-06,
      "loss": 0.9556,
      "step": 1510
    },
    {
      "epoch": 1.15,
      "grad_norm": 1.0508161783218384,
      "learning_rate": 6.17283950617284e-06,
      "loss": 0.9388,
      "step": 1520
    },
    {
      "epoch": 1.16,
      "grad_norm": 1.0714831352233887,
      "learning_rate": 6.147644242882339e-06,
      "loss": 0.9334,
      "step": 1530
    },
    {
      "epoch": 1.16,
      "grad_norm": 1.0550581216812134,
      "learning_rate": 6.122448979591837e-06,
      "loss": 0.9398,
      "step": 1540
    },
    {
      "epoch": 1.17,
      "grad_norm": 1.1077096462249756,
      "learning_rate": 6.097253716301336e-06,
      "loss": 0.9486,
      "step": 1550
    },
    {
      "epoch": 1.18,
      "grad_norm": 1.0329755544662476,
      "learning_rate": 6.072058453010834e-06,
      "loss": 0.9635,
      "step": 1560
    },
    {
      "epoch": 1.19,
      "grad_norm": 1.085105061531067,
      "learning_rate": 6.046863189720334e-06,
      "loss": 0.9394,
      "step": 1570
    },
    {
      "epoch": 1.19,
      "grad_norm": 1.0389838218688965,
      "learning_rate": 6.021667926429831e-06,
      "loss": 0.9474,
      "step": 1580
    },
    {
      "epoch": 1.2,
      "grad_norm": 1.1028958559036255,
      "learning_rate": 5.99647266313933e-06,
      "loss": 0.9364,
      "step": 1590
    },
    {
      "epoch": 1.21,
      "grad_norm": 1.0558817386627197,
      "learning_rate": 5.971277399848829e-06,
      "loss": 0.9276,
      "step": 1600
    },
    {
      "epoch": 1.22,
      "grad_norm": 1.0861153602600098,
      "learning_rate": 5.946082136558328e-06,
      "loss": 0.9375,
      "step": 1610
    },
    {
      "epoch": 1.22,
      "grad_norm": 1.1104577779769897,
      "learning_rate": 5.920886873267826e-06,
      "loss": 0.948,
      "step": 1620
    },
    {
      "epoch": 1.23,
      "grad_norm": 1.0775477886199951,
      "learning_rate": 5.895691609977324e-06,
      "loss": 0.9334,
      "step": 1630
    },
    {
      "epoch": 1.24,
      "grad_norm": 1.0878546237945557,
      "learning_rate": 5.8704963466868235e-06,
      "loss": 0.952,
      "step": 1640
    },
    {
      "epoch": 1.25,
      "grad_norm": 1.0840810537338257,
      "learning_rate": 5.845301083396322e-06,
      "loss": 0.9312,
      "step": 1650
    },
    {
      "epoch": 1.25,
      "grad_norm": 1.2151501178741455,
      "learning_rate": 5.820105820105821e-06,
      "loss": 0.9549,
      "step": 1660
    },
    {
      "epoch": 1.26,
      "grad_norm": 1.0833278894424438,
      "learning_rate": 5.79491055681532e-06,
      "loss": 0.9557,
      "step": 1670
    },
    {
      "epoch": 1.27,
      "grad_norm": 1.0161901712417603,
      "learning_rate": 5.769715293524818e-06,
      "loss": 0.9312,
      "step": 1680
    },
    {
      "epoch": 1.28,
      "grad_norm": 1.075750708580017,
      "learning_rate": 5.744520030234316e-06,
      "loss": 0.9501,
      "step": 1690
    },
    {
      "epoch": 1.28,
      "grad_norm": 1.042380928993225,
      "learning_rate": 5.719324766943815e-06,
      "loss": 0.9381,
      "step": 1700
    },
    {
      "epoch": 1.29,
      "grad_norm": 1.009562373161316,
      "learning_rate": 5.694129503653314e-06,
      "loss": 0.9305,
      "step": 1710
    },
    {
      "epoch": 1.3,
      "grad_norm": 1.1201280355453491,
      "learning_rate": 5.668934240362812e-06,
      "loss": 0.9318,
      "step": 1720
    },
    {
      "epoch": 1.31,
      "grad_norm": 1.1156421899795532,
      "learning_rate": 5.6437389770723105e-06,
      "loss": 0.9445,
      "step": 1730
    },
    {
      "epoch": 1.32,
      "grad_norm": 1.0913002490997314,
      "learning_rate": 5.61854371378181e-06,
      "loss": 0.9537,
      "step": 1740
    },
    {
      "epoch": 1.32,
      "grad_norm": 1.041400671005249,
      "learning_rate": 5.593348450491308e-06,
      "loss": 0.9464,
      "step": 1750
    },
    {
      "epoch": 1.33,
      "grad_norm": 1.0327285528182983,
      "learning_rate": 5.568153187200807e-06,
      "loss": 0.9361,
      "step": 1760
    },
    {
      "epoch": 1.34,
      "grad_norm": 1.1421902179718018,
      "learning_rate": 5.542957923910305e-06,
      "loss": 0.9764,
      "step": 1770
    },
    {
      "epoch": 1.35,
      "grad_norm": 1.202650785446167,
      "learning_rate": 5.517762660619804e-06,
      "loss": 0.9553,
      "step": 1780
    },
    {
      "epoch": 1.35,
      "grad_norm": 1.0642273426055908,
      "learning_rate": 5.492567397329303e-06,
      "loss": 0.9485,
      "step": 1790
    },
    {
      "epoch": 1.36,
      "grad_norm": 1.0749919414520264,
      "learning_rate": 5.467372134038801e-06,
      "loss": 0.9418,
      "step": 1800
    },
    {
      "epoch": 1.37,
      "grad_norm": 1.0974488258361816,
      "learning_rate": 5.442176870748301e-06,
      "loss": 0.9539,
      "step": 1810
    },
    {
      "epoch": 1.38,
      "grad_norm": 1.1204264163970947,
      "learning_rate": 5.416981607457798e-06,
      "loss": 0.9357,
      "step": 1820
    },
    {
      "epoch": 1.38,
      "grad_norm": 1.268764853477478,
      "learning_rate": 5.391786344167297e-06,
      "loss": 0.9512,
      "step": 1830
    },
    {
      "epoch": 1.39,
      "grad_norm": 1.159050703048706,
      "learning_rate": 5.366591080876795e-06,
      "loss": 0.9402,
      "step": 1840
    },
    {
      "epoch": 1.4,
      "grad_norm": 1.2613215446472168,
      "learning_rate": 5.341395817586295e-06,
      "loss": 0.9411,
      "step": 1850
    },
    {
      "epoch": 1.41,
      "grad_norm": 1.1373072862625122,
      "learning_rate": 5.3162005542957925e-06,
      "loss": 0.9752,
      "step": 1860
    },
    {
      "epoch": 1.41,
      "grad_norm": 1.144242763519287,
      "learning_rate": 5.291005291005291e-06,
      "loss": 0.9471,
      "step": 1870
    },
    {
      "epoch": 1.42,
      "grad_norm": 1.0989221334457397,
      "learning_rate": 5.2658100277147905e-06,
      "loss": 0.9348,
      "step": 1880
    },
    {
      "epoch": 1.43,
      "grad_norm": 1.0864170789718628,
      "learning_rate": 5.240614764424289e-06,
      "loss": 0.9435,
      "step": 1890
    },
    {
      "epoch": 1.44,
      "grad_norm": 1.1140406131744385,
      "learning_rate": 5.2154195011337876e-06,
      "loss": 0.9541,
      "step": 1900
    },
    {
      "epoch": 1.44,
      "grad_norm": 1.1856478452682495,
      "learning_rate": 5.190224237843285e-06,
      "loss": 0.9151,
      "step": 1910
    },
    {
      "epoch": 1.45,
      "grad_norm": 1.2324665784835815,
      "learning_rate": 5.165028974552785e-06,
      "loss": 0.9411,
      "step": 1920
    },
    {
      "epoch": 1.46,
      "grad_norm": 1.038913369178772,
      "learning_rate": 5.139833711262283e-06,
      "loss": 0.9487,
      "step": 1930
    },
    {
      "epoch": 1.47,
      "grad_norm": 1.040578842163086,
      "learning_rate": 5.114638447971782e-06,
      "loss": 0.943,
      "step": 1940
    },
    {
      "epoch": 1.47,
      "grad_norm": 1.289865493774414,
      "learning_rate": 5.089443184681281e-06,
      "loss": 0.9382,
      "step": 1950
    },
    {
      "epoch": 1.48,
      "grad_norm": 1.1952651739120483,
      "learning_rate": 5.064247921390779e-06,
      "loss": 0.9453,
      "step": 1960
    },
    {
      "epoch": 1.49,
      "grad_norm": 1.1922310590744019,
      "learning_rate": 5.0390526581002774e-06,
      "loss": 0.9242,
      "step": 1970
    },
    {
      "epoch": 1.5,
      "grad_norm": 1.1691875457763672,
      "learning_rate": 5.013857394809776e-06,
      "loss": 0.932,
      "step": 1980
    },
    {
      "epoch": 1.5,
      "grad_norm": 1.254301905632019,
      "learning_rate": 4.9886621315192745e-06,
      "loss": 0.9341,
      "step": 1990
    },
    {
      "epoch": 1.51,
      "grad_norm": 1.0548466444015503,
      "learning_rate": 4.963466868228773e-06,
      "loss": 0.9496,
      "step": 2000
    },
    {
      "epoch": 1.52,
      "grad_norm": 1.2470088005065918,
      "learning_rate": 4.938271604938272e-06,
      "loss": 0.9488,
      "step": 2010
    },
    {
      "epoch": 1.53,
      "grad_norm": 1.2311805486679077,
      "learning_rate": 4.91307634164777e-06,
      "loss": 0.942,
      "step": 2020
    },
    {
      "epoch": 1.53,
      "grad_norm": 1.1870156526565552,
      "learning_rate": 4.88788107835727e-06,
      "loss": 0.94,
      "step": 2030
    },
    {
      "epoch": 1.54,
      "grad_norm": 1.060556173324585,
      "learning_rate": 4.862685815066767e-06,
      "loss": 0.9223,
      "step": 2040
    },
    {
      "epoch": 1.55,
      "grad_norm": 1.1396002769470215,
      "learning_rate": 4.837490551776267e-06,
      "loss": 0.9588,
      "step": 2050
    },
    {
      "epoch": 1.56,
      "grad_norm": 1.2042090892791748,
      "learning_rate": 4.812295288485765e-06,
      "loss": 0.9534,
      "step": 2060
    },
    {
      "epoch": 1.56,
      "grad_norm": 1.1195515394210815,
      "learning_rate": 4.787100025195264e-06,
      "loss": 0.9528,
      "step": 2070
    },
    {
      "epoch": 1.57,
      "grad_norm": 1.290823221206665,
      "learning_rate": 4.761904761904762e-06,
      "loss": 0.9514,
      "step": 2080
    },
    {
      "epoch": 1.58,
      "grad_norm": 1.0271083116531372,
      "learning_rate": 4.736709498614261e-06,
      "loss": 0.9699,
      "step": 2090
    },
    {
      "epoch": 1.59,
      "grad_norm": 1.2110365629196167,
      "learning_rate": 4.7115142353237595e-06,
      "loss": 0.9358,
      "step": 2100
    },
    {
      "epoch": 1.59,
      "grad_norm": 1.1250057220458984,
      "learning_rate": 4.686318972033258e-06,
      "loss": 0.9491,
      "step": 2110
    },
    {
      "epoch": 1.6,
      "grad_norm": 1.1912841796875,
      "learning_rate": 4.661123708742757e-06,
      "loss": 0.9332,
      "step": 2120
    },
    {
      "epoch": 1.61,
      "grad_norm": 1.0299352407455444,
      "learning_rate": 4.635928445452255e-06,
      "loss": 0.9577,
      "step": 2130
    },
    {
      "epoch": 1.62,
      "grad_norm": 1.0603456497192383,
      "learning_rate": 4.610733182161754e-06,
      "loss": 0.9332,
      "step": 2140
    },
    {
      "epoch": 1.63,
      "grad_norm": 1.1462222337722778,
      "learning_rate": 4.585537918871252e-06,
      "loss": 0.9343,
      "step": 2150
    },
    {
      "epoch": 1.63,
      "grad_norm": 1.0181183815002441,
      "learning_rate": 4.560342655580751e-06,
      "loss": 0.9365,
      "step": 2160
    },
    {
      "epoch": 1.64,
      "grad_norm": 1.2271602153778076,
      "learning_rate": 4.53514739229025e-06,
      "loss": 0.9171,
      "step": 2170
    },
    {
      "epoch": 1.65,
      "grad_norm": 1.1205991506576538,
      "learning_rate": 4.509952128999748e-06,
      "loss": 0.9299,
      "step": 2180
    },
    {
      "epoch": 1.66,
      "grad_norm": 1.1556931734085083,
      "learning_rate": 4.484756865709247e-06,
      "loss": 0.9652,
      "step": 2190
    },
    {
      "epoch": 1.66,
      "grad_norm": 1.0930700302124023,
      "learning_rate": 4.459561602418746e-06,
      "loss": 0.9259,
      "step": 2200
    },
    {
      "epoch": 1.67,
      "grad_norm": 1.207714319229126,
      "learning_rate": 4.434366339128244e-06,
      "loss": 0.9529,
      "step": 2210
    },
    {
      "epoch": 1.68,
      "grad_norm": 1.1223605871200562,
      "learning_rate": 4.409171075837743e-06,
      "loss": 0.9397,
      "step": 2220
    },
    {
      "epoch": 1.69,
      "grad_norm": 1.2369412183761597,
      "learning_rate": 4.3839758125472415e-06,
      "loss": 0.9246,
      "step": 2230
    },
    {
      "epoch": 1.69,
      "grad_norm": 1.232971429824829,
      "learning_rate": 4.35878054925674e-06,
      "loss": 0.9349,
      "step": 2240
    },
    {
      "epoch": 1.7,
      "grad_norm": 1.1932899951934814,
      "learning_rate": 4.333585285966239e-06,
      "loss": 0.9495,
      "step": 2250
    },
    {
      "epoch": 1.71,
      "grad_norm": 1.2230783700942993,
      "learning_rate": 4.308390022675737e-06,
      "loss": 0.9205,
      "step": 2260
    },
    {
      "epoch": 1.72,
      "grad_norm": 1.2365373373031616,
      "learning_rate": 4.283194759385236e-06,
      "loss": 0.9214,
      "step": 2270
    },
    {
      "epoch": 1.72,
      "grad_norm": 1.132912278175354,
      "learning_rate": 4.257999496094734e-06,
      "loss": 0.9154,
      "step": 2280
    },
    {
      "epoch": 1.73,
      "grad_norm": 1.2143046855926514,
      "learning_rate": 4.232804232804233e-06,
      "loss": 0.9485,
      "step": 2290
    },
    {
      "epoch": 1.74,
      "grad_norm": 1.2785279750823975,
      "learning_rate": 4.207608969513731e-06,
      "loss": 0.9462,
      "step": 2300
    },
    {
      "epoch": 1.75,
      "grad_norm": 1.3493692874908447,
      "learning_rate": 4.182413706223231e-06,
      "loss": 0.9382,
      "step": 2310
    },
    {
      "epoch": 1.75,
      "grad_norm": 1.1612365245819092,
      "learning_rate": 4.1572184429327285e-06,
      "loss": 0.9407,
      "step": 2320
    },
    {
      "epoch": 1.76,
      "grad_norm": 1.2210137844085693,
      "learning_rate": 4.132023179642228e-06,
      "loss": 0.9196,
      "step": 2330
    },
    {
      "epoch": 1.77,
      "grad_norm": 1.1422805786132812,
      "learning_rate": 4.1068279163517265e-06,
      "loss": 0.9426,
      "step": 2340
    },
    {
      "epoch": 1.78,
      "grad_norm": 1.1776541471481323,
      "learning_rate": 4.081632653061225e-06,
      "loss": 0.9436,
      "step": 2350
    },
    {
      "epoch": 1.78,
      "grad_norm": 1.2455426454544067,
      "learning_rate": 4.0564373897707236e-06,
      "loss": 0.9299,
      "step": 2360
    },
    {
      "epoch": 1.79,
      "grad_norm": 1.146103858947754,
      "learning_rate": 4.031242126480222e-06,
      "loss": 0.9353,
      "step": 2370
    },
    {
      "epoch": 1.8,
      "grad_norm": 1.1490944623947144,
      "learning_rate": 4.006046863189721e-06,
      "loss": 0.9373,
      "step": 2380
    },
    {
      "epoch": 1.81,
      "grad_norm": 1.1841084957122803,
      "learning_rate": 3.980851599899219e-06,
      "loss": 0.9214,
      "step": 2390
    },
    {
      "epoch": 1.81,
      "grad_norm": 1.360938310623169,
      "learning_rate": 3.955656336608718e-06,
      "loss": 0.9233,
      "step": 2400
    },
    {
      "epoch": 1.82,
      "grad_norm": 1.1501877307891846,
      "learning_rate": 3.930461073318216e-06,
      "loss": 0.9149,
      "step": 2410
    },
    {
      "epoch": 1.83,
      "grad_norm": 1.3008352518081665,
      "learning_rate": 3.905265810027715e-06,
      "loss": 0.9592,
      "step": 2420
    },
    {
      "epoch": 1.84,
      "grad_norm": 1.2323644161224365,
      "learning_rate": 3.8800705467372134e-06,
      "loss": 0.9295,
      "step": 2430
    },
    {
      "epoch": 1.84,
      "grad_norm": 1.1721792221069336,
      "learning_rate": 3.854875283446712e-06,
      "loss": 0.9431,
      "step": 2440
    },
    {
      "epoch": 1.85,
      "grad_norm": 1.2879486083984375,
      "learning_rate": 3.829680020156211e-06,
      "loss": 0.9463,
      "step": 2450
    },
    {
      "epoch": 1.86,
      "grad_norm": 1.1511075496673584,
      "learning_rate": 3.8044847568657095e-06,
      "loss": 0.9664,
      "step": 2460
    },
    {
      "epoch": 1.87,
      "grad_norm": 1.3236780166625977,
      "learning_rate": 3.7792894935752085e-06,
      "loss": 0.9368,
      "step": 2470
    },
    {
      "epoch": 1.87,
      "grad_norm": 1.1940069198608398,
      "learning_rate": 3.7540942302847066e-06,
      "loss": 0.9548,
      "step": 2480
    },
    {
      "epoch": 1.88,
      "grad_norm": 1.1677494049072266,
      "learning_rate": 3.7288989669942056e-06,
      "loss": 0.9592,
      "step": 2490
    },
    {
      "epoch": 1.89,
      "grad_norm": 1.1518640518188477,
      "learning_rate": 3.7037037037037037e-06,
      "loss": 0.9497,
      "step": 2500
    },
    {
      "epoch": 1.9,
      "grad_norm": 1.159209966659546,
      "learning_rate": 3.6785084404132027e-06,
      "loss": 0.9368,
      "step": 2510
    },
    {
      "epoch": 1.9,
      "grad_norm": 1.2133903503417969,
      "learning_rate": 3.6533131771227017e-06,
      "loss": 0.9257,
      "step": 2520
    },
    {
      "epoch": 1.91,
      "grad_norm": 1.329829454421997,
      "learning_rate": 3.6281179138322e-06,
      "loss": 0.9544,
      "step": 2530
    },
    {
      "epoch": 1.92,
      "grad_norm": 1.264733910560608,
      "learning_rate": 3.602922650541699e-06,
      "loss": 0.9525,
      "step": 2540
    },
    {
      "epoch": 1.93,
      "grad_norm": 1.2355499267578125,
      "learning_rate": 3.577727387251197e-06,
      "loss": 0.95,
      "step": 2550
    },
    {
      "epoch": 1.93,
      "grad_norm": 1.1711816787719727,
      "learning_rate": 3.552532123960696e-06,
      "loss": 0.9443,
      "step": 2560
    },
    {
      "epoch": 1.94,
      "grad_norm": 1.1221129894256592,
      "learning_rate": 3.527336860670194e-06,
      "loss": 0.9422,
      "step": 2570
    },
    {
      "epoch": 1.95,
      "grad_norm": 1.154335379600525,
      "learning_rate": 3.502141597379693e-06,
      "loss": 0.9462,
      "step": 2580
    },
    {
      "epoch": 1.96,
      "grad_norm": 1.2431056499481201,
      "learning_rate": 3.476946334089191e-06,
      "loss": 0.9305,
      "step": 2590
    },
    {
      "epoch": 1.97,
      "grad_norm": 1.1966361999511719,
      "learning_rate": 3.45175107079869e-06,
      "loss": 0.9537,
      "step": 2600
    },
    {
      "epoch": 1.97,
      "grad_norm": 1.2401050329208374,
      "learning_rate": 3.426555807508189e-06,
      "loss": 0.9262,
      "step": 2610
    },
    {
      "epoch": 1.98,
      "grad_norm": 1.1483770608901978,
      "learning_rate": 3.4013605442176872e-06,
      "loss": 0.9515,
      "step": 2620
    },
    {
      "epoch": 1.99,
      "grad_norm": 1.1387736797332764,
      "learning_rate": 3.376165280927186e-06,
      "loss": 0.9293,
      "step": 2630
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.2311975955963135,
      "learning_rate": 3.3509700176366843e-06,
      "loss": 0.9201,
      "step": 2640
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.1818970441818237,
      "learning_rate": 3.3257747543461833e-06,
      "loss": 0.9439,
      "step": 2650
    },
    {
      "epoch": 2.01,
      "grad_norm": 1.3574014902114868,
      "learning_rate": 3.3005794910556814e-06,
      "loss": 0.9235,
      "step": 2660
    },
    {
      "epoch": 2.02,
      "grad_norm": 1.1797312498092651,
      "learning_rate": 3.2753842277651804e-06,
      "loss": 0.9235,
      "step": 2670
    },
    {
      "epoch": 2.03,
      "grad_norm": 1.225716233253479,
      "learning_rate": 3.2501889644746794e-06,
      "loss": 0.9363,
      "step": 2680
    },
    {
      "epoch": 2.03,
      "grad_norm": 1.1875022649765015,
      "learning_rate": 3.2249937011841775e-06,
      "loss": 0.9155,
      "step": 2690
    },
    {
      "epoch": 2.04,
      "grad_norm": 1.2818158864974976,
      "learning_rate": 3.1997984378936765e-06,
      "loss": 0.9189,
      "step": 2700
    },
    {
      "epoch": 2.05,
      "grad_norm": 1.0735368728637695,
      "learning_rate": 3.1746031746031746e-06,
      "loss": 0.9183,
      "step": 2710
    },
    {
      "epoch": 2.06,
      "grad_norm": 1.2235422134399414,
      "learning_rate": 3.1494079113126736e-06,
      "loss": 0.9383,
      "step": 2720
    },
    {
      "epoch": 2.06,
      "grad_norm": 1.2761168479919434,
      "learning_rate": 3.1242126480221717e-06,
      "loss": 0.9161,
      "step": 2730
    },
    {
      "epoch": 2.07,
      "grad_norm": 1.1925853490829468,
      "learning_rate": 3.0990173847316707e-06,
      "loss": 0.9133,
      "step": 2740
    },
    {
      "epoch": 2.08,
      "grad_norm": 1.1860272884368896,
      "learning_rate": 3.0738221214411697e-06,
      "loss": 0.9317,
      "step": 2750
    },
    {
      "epoch": 2.09,
      "grad_norm": 1.285198450088501,
      "learning_rate": 3.048626858150668e-06,
      "loss": 0.9332,
      "step": 2760
    },
    {
      "epoch": 2.09,
      "grad_norm": 1.220641851425171,
      "learning_rate": 3.023431594860167e-06,
      "loss": 0.9464,
      "step": 2770
    },
    {
      "epoch": 2.1,
      "grad_norm": 1.31607186794281,
      "learning_rate": 2.998236331569665e-06,
      "loss": 0.9233,
      "step": 2780
    },
    {
      "epoch": 2.11,
      "grad_norm": 1.2238725423812866,
      "learning_rate": 2.973041068279164e-06,
      "loss": 0.9359,
      "step": 2790
    },
    {
      "epoch": 2.12,
      "grad_norm": 1.1632503271102905,
      "learning_rate": 2.947845804988662e-06,
      "loss": 0.9117,
      "step": 2800
    },
    {
      "epoch": 2.12,
      "grad_norm": 1.3732521533966064,
      "learning_rate": 2.922650541698161e-06,
      "loss": 0.9068,
      "step": 2810
    },
    {
      "epoch": 2.13,
      "grad_norm": 1.2329943180084229,
      "learning_rate": 2.89745527840766e-06,
      "loss": 0.9208,
      "step": 2820
    },
    {
      "epoch": 2.14,
      "grad_norm": 1.331434726715088,
      "learning_rate": 2.872260015117158e-06,
      "loss": 0.9353,
      "step": 2830
    },
    {
      "epoch": 2.15,
      "grad_norm": 1.3423517942428589,
      "learning_rate": 2.847064751826657e-06,
      "loss": 0.9216,
      "step": 2840
    },
    {
      "epoch": 2.15,
      "grad_norm": 1.2505260705947876,
      "learning_rate": 2.8218694885361552e-06,
      "loss": 0.9267,
      "step": 2850
    },
    {
      "epoch": 2.16,
      "grad_norm": 1.1538835763931274,
      "learning_rate": 2.796674225245654e-06,
      "loss": 0.9195,
      "step": 2860
    },
    {
      "epoch": 2.17,
      "grad_norm": 1.3361774682998657,
      "learning_rate": 2.7714789619551523e-06,
      "loss": 0.9252,
      "step": 2870
    },
    {
      "epoch": 2.18,
      "grad_norm": 1.2870272397994995,
      "learning_rate": 2.7462836986646513e-06,
      "loss": 0.9011,
      "step": 2880
    },
    {
      "epoch": 2.18,
      "grad_norm": 1.238709807395935,
      "learning_rate": 2.7210884353741503e-06,
      "loss": 0.9203,
      "step": 2890
    },
    {
      "epoch": 2.19,
      "grad_norm": 1.1434613466262817,
      "learning_rate": 2.6958931720836484e-06,
      "loss": 0.9263,
      "step": 2900
    },
    {
      "epoch": 2.2,
      "grad_norm": 1.2599645853042603,
      "learning_rate": 2.6706979087931474e-06,
      "loss": 0.9224,
      "step": 2910
    },
    {
      "epoch": 2.21,
      "grad_norm": 1.3294990062713623,
      "learning_rate": 2.6455026455026455e-06,
      "loss": 0.9116,
      "step": 2920
    },
    {
      "epoch": 2.21,
      "grad_norm": 1.2516047954559326,
      "learning_rate": 2.6203073822121445e-06,
      "loss": 0.9287,
      "step": 2930
    },
    {
      "epoch": 2.22,
      "grad_norm": 1.2115424871444702,
      "learning_rate": 2.5951121189216426e-06,
      "loss": 0.9384,
      "step": 2940
    },
    {
      "epoch": 2.23,
      "grad_norm": 1.1633473634719849,
      "learning_rate": 2.5699168556311416e-06,
      "loss": 0.9211,
      "step": 2950
    },
    {
      "epoch": 2.24,
      "grad_norm": 1.181342363357544,
      "learning_rate": 2.5447215923406406e-06,
      "loss": 0.9122,
      "step": 2960
    },
    {
      "epoch": 2.24,
      "grad_norm": 1.2719746828079224,
      "learning_rate": 2.5195263290501387e-06,
      "loss": 0.9302,
      "step": 2970
    },
    {
      "epoch": 2.25,
      "grad_norm": 1.292691707611084,
      "learning_rate": 2.4943310657596373e-06,
      "loss": 0.9226,
      "step": 2980
    },
    {
      "epoch": 2.26,
      "grad_norm": 1.1643900871276855,
      "learning_rate": 2.469135802469136e-06,
      "loss": 0.903,
      "step": 2990
    },
    {
      "epoch": 2.27,
      "grad_norm": 1.2362549304962158,
      "learning_rate": 2.443940539178635e-06,
      "loss": 0.9238,
      "step": 3000
    },
    {
      "epoch": 2.28,
      "grad_norm": 1.1967755556106567,
      "learning_rate": 2.4187452758881334e-06,
      "loss": 0.9372,
      "step": 3010
    },
    {
      "epoch": 2.28,
      "grad_norm": 1.1664007902145386,
      "learning_rate": 2.393550012597632e-06,
      "loss": 0.9249,
      "step": 3020
    },
    {
      "epoch": 2.29,
      "grad_norm": 1.39642333984375,
      "learning_rate": 2.3683547493071305e-06,
      "loss": 0.9381,
      "step": 3030
    },
    {
      "epoch": 2.3,
      "grad_norm": 1.2738609313964844,
      "learning_rate": 2.343159486016629e-06,
      "loss": 0.9345,
      "step": 3040
    },
    {
      "epoch": 2.31,
      "grad_norm": 1.2925204038619995,
      "learning_rate": 2.3179642227261276e-06,
      "loss": 0.9156,
      "step": 3050
    },
    {
      "epoch": 2.31,
      "grad_norm": 1.470628261566162,
      "learning_rate": 2.292768959435626e-06,
      "loss": 0.9128,
      "step": 3060
    },
    {
      "epoch": 2.32,
      "grad_norm": 1.280967354774475,
      "learning_rate": 2.267573696145125e-06,
      "loss": 0.9225,
      "step": 3070
    },
    {
      "epoch": 2.33,
      "grad_norm": 1.3174177408218384,
      "learning_rate": 2.2423784328546237e-06,
      "loss": 0.9138,
      "step": 3080
    },
    {
      "epoch": 2.34,
      "grad_norm": 1.3123756647109985,
      "learning_rate": 2.217183169564122e-06,
      "loss": 0.9284,
      "step": 3090
    },
    {
      "epoch": 2.34,
      "grad_norm": 1.232643723487854,
      "learning_rate": 2.1919879062736208e-06,
      "loss": 0.9201,
      "step": 3100
    },
    {
      "epoch": 2.35,
      "grad_norm": 1.207700490951538,
      "learning_rate": 2.1667926429831193e-06,
      "loss": 0.9061,
      "step": 3110
    },
    {
      "epoch": 2.36,
      "grad_norm": 1.2038023471832275,
      "learning_rate": 2.141597379692618e-06,
      "loss": 0.9437,
      "step": 3120
    },
    {
      "epoch": 2.37,
      "grad_norm": 1.4127525091171265,
      "learning_rate": 2.1164021164021164e-06,
      "loss": 0.9142,
      "step": 3130
    },
    {
      "epoch": 2.37,
      "grad_norm": 1.3315386772155762,
      "learning_rate": 2.0912068531116154e-06,
      "loss": 0.9217,
      "step": 3140
    },
    {
      "epoch": 2.38,
      "grad_norm": 1.3743528127670288,
      "learning_rate": 2.066011589821114e-06,
      "loss": 0.8931,
      "step": 3150
    },
    {
      "epoch": 2.39,
      "grad_norm": 1.2260786294937134,
      "learning_rate": 2.0408163265306125e-06,
      "loss": 0.9224,
      "step": 3160
    },
    {
      "epoch": 2.4,
      "grad_norm": 1.3002612590789795,
      "learning_rate": 2.015621063240111e-06,
      "loss": 0.9371,
      "step": 3170
    },
    {
      "epoch": 2.4,
      "grad_norm": 1.1567788124084473,
      "learning_rate": 1.9904257999496096e-06,
      "loss": 0.9244,
      "step": 3180
    },
    {
      "epoch": 2.41,
      "grad_norm": 1.2505953311920166,
      "learning_rate": 1.965230536659108e-06,
      "loss": 0.8985,
      "step": 3190
    },
    {
      "epoch": 2.42,
      "grad_norm": 1.3488541841506958,
      "learning_rate": 1.9400352733686067e-06,
      "loss": 0.9111,
      "step": 3200
    },
    {
      "epoch": 2.43,
      "grad_norm": 1.371126651763916,
      "learning_rate": 1.9148400100781057e-06,
      "loss": 0.9094,
      "step": 3210
    },
    {
      "epoch": 2.43,
      "grad_norm": 1.3531619310379028,
      "learning_rate": 1.8896447467876043e-06,
      "loss": 0.9353,
      "step": 3220
    },
    {
      "epoch": 2.44,
      "grad_norm": 1.261234998703003,
      "learning_rate": 1.8644494834971028e-06,
      "loss": 0.9635,
      "step": 3230
    },
    {
      "epoch": 2.45,
      "grad_norm": 1.2285544872283936,
      "learning_rate": 1.8392542202066014e-06,
      "loss": 0.9104,
      "step": 3240
    },
    {
      "epoch": 2.46,
      "grad_norm": 1.2340446710586548,
      "learning_rate": 1.8140589569161e-06,
      "loss": 0.9275,
      "step": 3250
    },
    {
      "epoch": 2.46,
      "grad_norm": 1.1888856887817383,
      "learning_rate": 1.7888636936255985e-06,
      "loss": 0.8929,
      "step": 3260
    },
    {
      "epoch": 2.47,
      "grad_norm": 1.3445130586624146,
      "learning_rate": 1.763668430335097e-06,
      "loss": 0.9233,
      "step": 3270
    },
    {
      "epoch": 2.48,
      "grad_norm": 1.2736765146255493,
      "learning_rate": 1.7384731670445956e-06,
      "loss": 0.9143,
      "step": 3280
    },
    {
      "epoch": 2.49,
      "grad_norm": 1.1950249671936035,
      "learning_rate": 1.7132779037540945e-06,
      "loss": 0.8981,
      "step": 3290
    },
    {
      "epoch": 2.49,
      "grad_norm": 1.3372317552566528,
      "learning_rate": 1.688082640463593e-06,
      "loss": 0.9149,
      "step": 3300
    },
    {
      "epoch": 2.5,
      "grad_norm": 1.2550235986709595,
      "learning_rate": 1.6628873771730917e-06,
      "loss": 0.9448,
      "step": 3310
    },
    {
      "epoch": 2.51,
      "grad_norm": 1.3083131313323975,
      "learning_rate": 1.6376921138825902e-06,
      "loss": 0.926,
      "step": 3320
    },
    {
      "epoch": 2.52,
      "grad_norm": 1.2111276388168335,
      "learning_rate": 1.6124968505920888e-06,
      "loss": 0.9237,
      "step": 3330
    },
    {
      "epoch": 2.52,
      "grad_norm": 1.1683611869812012,
      "learning_rate": 1.5873015873015873e-06,
      "loss": 0.9265,
      "step": 3340
    },
    {
      "epoch": 2.53,
      "grad_norm": 1.2934691905975342,
      "learning_rate": 1.5621063240110859e-06,
      "loss": 0.8935,
      "step": 3350
    },
    {
      "epoch": 2.54,
      "grad_norm": 1.2122243642807007,
      "learning_rate": 1.5369110607205848e-06,
      "loss": 0.9184,
      "step": 3360
    },
    {
      "epoch": 2.55,
      "grad_norm": 1.318337321281433,
      "learning_rate": 1.5117157974300834e-06,
      "loss": 0.919,
      "step": 3370
    },
    {
      "epoch": 2.55,
      "grad_norm": 1.2197314500808716,
      "learning_rate": 1.486520534139582e-06,
      "loss": 0.9118,
      "step": 3380
    },
    {
      "epoch": 2.56,
      "grad_norm": 1.1137281656265259,
      "learning_rate": 1.4613252708490805e-06,
      "loss": 0.9312,
      "step": 3390
    },
    {
      "epoch": 2.57,
      "grad_norm": 1.2075839042663574,
      "learning_rate": 1.436130007558579e-06,
      "loss": 0.9258,
      "step": 3400
    },
    {
      "epoch": 2.58,
      "grad_norm": 1.379468321800232,
      "learning_rate": 1.4109347442680776e-06,
      "loss": 0.9163,
      "step": 3410
    },
    {
      "epoch": 2.59,
      "grad_norm": 1.3524144887924194,
      "learning_rate": 1.3857394809775762e-06,
      "loss": 0.9211,
      "step": 3420
    },
    {
      "epoch": 2.59,
      "grad_norm": 1.342460036277771,
      "learning_rate": 1.3605442176870751e-06,
      "loss": 0.9107,
      "step": 3430
    },
    {
      "epoch": 2.6,
      "grad_norm": 1.3853424787521362,
      "learning_rate": 1.3353489543965737e-06,
      "loss": 0.9367,
      "step": 3440
    },
    {
      "epoch": 2.61,
      "grad_norm": 1.397850513458252,
      "learning_rate": 1.3101536911060723e-06,
      "loss": 0.9173,
      "step": 3450
    },
    {
      "epoch": 2.62,
      "grad_norm": 1.439322590827942,
      "learning_rate": 1.2849584278155708e-06,
      "loss": 0.9158,
      "step": 3460
    },
    {
      "epoch": 2.62,
      "grad_norm": 1.199365258216858,
      "learning_rate": 1.2597631645250694e-06,
      "loss": 0.9167,
      "step": 3470
    },
    {
      "epoch": 2.63,
      "grad_norm": 1.323708415031433,
      "learning_rate": 1.234567901234568e-06,
      "loss": 0.9087,
      "step": 3480
    },
    {
      "epoch": 2.64,
      "grad_norm": 1.323588252067566,
      "learning_rate": 1.2093726379440667e-06,
      "loss": 0.9217,
      "step": 3490
    },
    {
      "epoch": 2.65,
      "grad_norm": 1.3864338397979736,
      "learning_rate": 1.1841773746535652e-06,
      "loss": 0.9202,
      "step": 3500
    }
  ],
  "logging_steps": 10,
  "max_steps": 3969,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 3.1042768180419625e+18,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}