{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.954084098598357,
  "eval_steps": 500,
  "global_step": 20500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": NaN,
      "learning_rate": 2.0000000000000002e-07,
      "loss": 2.0414,
      "step": 10
    },
    {
      "epoch": 0.0,
      "grad_norm": 53.515953063964844,
      "learning_rate": 6.5e-07,
      "loss": 1.8355,
      "step": 20
    },
    {
      "epoch": 0.01,
      "grad_norm": 27.920907974243164,
      "learning_rate": 1.15e-06,
      "loss": 1.5172,
      "step": 30
    },
    {
      "epoch": 0.01,
      "grad_norm": 20.47172737121582,
      "learning_rate": 1.65e-06,
      "loss": 1.2654,
      "step": 40
    },
    {
      "epoch": 0.01,
      "grad_norm": 8.903529167175293,
      "learning_rate": 2.1499999999999997e-06,
      "loss": 1.1095,
      "step": 50
    },
    {
      "epoch": 0.01,
      "grad_norm": 5.902841567993164,
      "learning_rate": 2.65e-06,
      "loss": 1.0017,
      "step": 60
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.9340908527374268,
      "learning_rate": 3.1500000000000003e-06,
      "loss": 0.8846,
      "step": 70
    },
    {
      "epoch": 0.02,
      "grad_norm": 8.701367378234863,
      "learning_rate": 3.6499999999999998e-06,
      "loss": 0.8572,
      "step": 80
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.3205362558364868,
      "learning_rate": 4.15e-06,
      "loss": 0.839,
      "step": 90
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.0334558486938477,
      "learning_rate": 4.65e-06,
      "loss": 0.7838,
      "step": 100
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.8086520433425903,
      "learning_rate": 5.15e-06,
      "loss": 0.7782,
      "step": 110
    },
    {
      "epoch": 0.03,
      "grad_norm": 29.50135040283203,
      "learning_rate": 5.65e-06,
      "loss": 0.7883,
      "step": 120
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.2708683013916016,
      "learning_rate": 6.15e-06,
      "loss": 0.7961,
      "step": 130
    },
    {
      "epoch": 0.03,
      "grad_norm": 33.43790817260742,
      "learning_rate": 6.650000000000001e-06,
      "loss": 0.7626,
      "step": 140
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.2334959506988525,
      "learning_rate": 7.15e-06,
      "loss": 0.7732,
      "step": 150
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.7489686012268066,
      "learning_rate": 7.65e-06,
      "loss": 0.7715,
      "step": 160
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.358823776245117,
      "learning_rate": 8.15e-06,
      "loss": 0.7778,
      "step": 170
    },
    {
      "epoch": 0.04,
      "grad_norm": 4.348245620727539,
      "learning_rate": 8.65e-06,
      "loss": 0.7619,
      "step": 180
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.9023125767707825,
      "learning_rate": 9.15e-06,
      "loss": 0.7746,
      "step": 190
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.9158114790916443,
      "learning_rate": 9.65e-06,
      "loss": 0.7592,
      "step": 200
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.6256226301193237,
      "learning_rate": 1.0150000000000001e-05,
      "loss": 0.7787,
      "step": 210
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.3805778920650482,
      "learning_rate": 1.065e-05,
      "loss": 0.7811,
      "step": 220
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.6968041062355042,
      "learning_rate": 1.115e-05,
      "loss": 0.73,
      "step": 230
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.9035410284996033,
      "learning_rate": 1.1650000000000002e-05,
      "loss": 0.7518,
      "step": 240
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.6727488040924072,
      "learning_rate": 1.215e-05,
      "loss": 0.7805,
      "step": 250
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.4153461158275604,
      "learning_rate": 1.2650000000000001e-05,
      "loss": 0.7706,
      "step": 260
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.4846651554107666,
      "learning_rate": 1.3150000000000001e-05,
      "loss": 0.7492,
      "step": 270
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.6631388664245605,
      "learning_rate": 1.3650000000000001e-05,
      "loss": 0.7681,
      "step": 280
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.6325013041496277,
      "learning_rate": 1.415e-05,
      "loss": 0.7753,
      "step": 290
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.877907395362854,
      "learning_rate": 1.465e-05,
      "loss": 0.7188,
      "step": 300
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.37142279744148254,
      "learning_rate": 1.515e-05,
      "loss": 0.7204,
      "step": 310
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7337246537208557,
      "learning_rate": 1.565e-05,
      "loss": 0.778,
      "step": 320
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.5847220420837402,
      "learning_rate": 1.6150000000000003e-05,
      "loss": 0.7288,
      "step": 330
    },
    {
      "epoch": 0.08,
      "grad_norm": 3.745180606842041,
      "learning_rate": 1.665e-05,
      "loss": 0.7531,
      "step": 340
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.357301265001297,
      "learning_rate": 1.7150000000000004e-05,
      "loss": 0.7448,
      "step": 350
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.9032486081123352,
      "learning_rate": 1.765e-05,
      "loss": 0.7651,
      "step": 360
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.2864232361316681,
      "learning_rate": 1.815e-05,
      "loss": 0.7193,
      "step": 370
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.8560261726379395,
      "learning_rate": 1.865e-05,
      "loss": 0.7421,
      "step": 380
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.2971792221069336,
      "learning_rate": 1.915e-05,
      "loss": 0.7171,
      "step": 390
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.40850627422332764,
      "learning_rate": 1.9650000000000003e-05,
      "loss": 0.7459,
      "step": 400
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.3934139311313629,
      "learning_rate": 2.0150000000000002e-05,
      "loss": 0.7205,
      "step": 410
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.8674131631851196,
      "learning_rate": 2.065e-05,
      "loss": 0.752,
      "step": 420
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.36707818508148193,
      "learning_rate": 2.115e-05,
      "loss": 0.7494,
      "step": 430
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.5044310092926025,
      "learning_rate": 2.165e-05,
      "loss": 0.7595,
      "step": 440
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.9025150537490845,
      "learning_rate": 2.215e-05,
      "loss": 0.7379,
      "step": 450
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.5004140138626099,
      "learning_rate": 2.265e-05,
      "loss": 0.7886,
      "step": 460
    },
    {
      "epoch": 0.11,
      "grad_norm": 3.544482707977295,
      "learning_rate": 2.3150000000000004e-05,
      "loss": 0.7259,
      "step": 470
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.33023321628570557,
      "learning_rate": 2.365e-05,
      "loss": 0.7333,
      "step": 480
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.3548080325126648,
      "learning_rate": 2.415e-05,
      "loss": 0.7527,
      "step": 490
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.42450150847435,
      "learning_rate": 2.465e-05,
      "loss": 0.7443,
      "step": 500
    },
    {
      "epoch": 0.12,
      "eval_loss": 0.7428915500640869,
      "eval_runtime": 68.3143,
      "eval_samples_per_second": 29.276,
      "eval_steps_per_second": 0.922,
      "step": 500
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.3299656808376312,
      "learning_rate": 2.515e-05,
      "loss": 0.7447,
      "step": 510
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.27258285880088806,
      "learning_rate": 2.5650000000000003e-05,
      "loss": 0.7278,
      "step": 520
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.23753009736537933,
      "learning_rate": 2.6150000000000002e-05,
      "loss": 0.7721,
      "step": 530
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.1074901819229126,
      "learning_rate": 2.6650000000000004e-05,
      "loss": 0.7383,
      "step": 540
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.428067147731781,
      "learning_rate": 2.7150000000000003e-05,
      "loss": 0.7508,
      "step": 550
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.35956478118896484,
      "learning_rate": 2.7650000000000005e-05,
      "loss": 0.7461,
      "step": 560
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.23017314076423645,
      "learning_rate": 2.815e-05,
      "loss": 0.7463,
      "step": 570
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.20370900630950928,
      "learning_rate": 2.865e-05,
      "loss": 0.7405,
      "step": 580
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.24586541950702667,
      "learning_rate": 2.915e-05,
      "loss": 0.6964,
      "step": 590
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.27857884764671326,
      "learning_rate": 2.965e-05,
      "loss": 0.7376,
      "step": 600
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.2655833065509796,
      "learning_rate": 3.015e-05,
      "loss": 0.7496,
      "step": 610
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.17174670100212097,
      "learning_rate": 3.065e-05,
      "loss": 0.7219,
      "step": 620
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.22892743349075317,
      "learning_rate": 3.115e-05,
      "loss": 0.7111,
      "step": 630
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.5913789868354797,
      "learning_rate": 3.1650000000000004e-05,
      "loss": 0.7097,
      "step": 640
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.25620004534721375,
      "learning_rate": 3.215e-05,
      "loss": 0.7306,
      "step": 650
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.3311476409435272,
      "learning_rate": 3.265e-05,
      "loss": 0.7406,
      "step": 660
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.4708893597126007,
      "learning_rate": 3.3150000000000006e-05,
      "loss": 0.7071,
      "step": 670
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.22469288110733032,
      "learning_rate": 3.3650000000000005e-05,
      "loss": 0.7116,
      "step": 680
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.2909330725669861,
      "learning_rate": 3.415e-05,
      "loss": 0.7592,
      "step": 690
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.21274766325950623,
      "learning_rate": 3.465e-05,
      "loss": 0.7144,
      "step": 700
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.23929230868816376,
      "learning_rate": 3.515e-05,
      "loss": 0.7382,
      "step": 710
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.3021218180656433,
      "learning_rate": 3.565e-05,
      "loss": 0.7287,
      "step": 720
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.23557321727275848,
      "learning_rate": 3.615e-05,
      "loss": 0.7475,
      "step": 730
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.22648084163665771,
      "learning_rate": 3.665e-05,
      "loss": 0.7457,
      "step": 740
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.2707761824131012,
      "learning_rate": 3.715e-05,
      "loss": 0.7322,
      "step": 750
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.2799661457538605,
      "learning_rate": 3.765e-05,
      "loss": 0.7682,
      "step": 760
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.326861172914505,
      "learning_rate": 3.8150000000000006e-05,
      "loss": 0.73,
      "step": 770
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.31985723972320557,
      "learning_rate": 3.8650000000000004e-05,
      "loss": 0.744,
      "step": 780
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.21637533605098724,
      "learning_rate": 3.915e-05,
      "loss": 0.7161,
      "step": 790
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.26102888584136963,
      "learning_rate": 3.965e-05,
      "loss": 0.7553,
      "step": 800
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.2469174712896347,
      "learning_rate": 4.015000000000001e-05,
      "loss": 0.7444,
      "step": 810
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.2022310495376587,
      "learning_rate": 4.065e-05,
      "loss": 0.7089,
      "step": 820
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.1729898750782013,
      "learning_rate": 4.115e-05,
      "loss": 0.7122,
      "step": 830
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.22450707852840424,
      "learning_rate": 4.165e-05,
      "loss": 0.7099,
      "step": 840
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.22389698028564453,
      "learning_rate": 4.215e-05,
      "loss": 0.7445,
      "step": 850
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.24962696433067322,
      "learning_rate": 4.265e-05,
      "loss": 0.7052,
      "step": 860
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.23780404031276703,
      "learning_rate": 4.315e-05,
      "loss": 0.7423,
      "step": 870
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.2434270679950714,
      "learning_rate": 4.3650000000000004e-05,
      "loss": 0.7291,
      "step": 880
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.3332350552082062,
      "learning_rate": 4.415e-05,
      "loss": 0.7306,
      "step": 890
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.5084495544433594,
      "learning_rate": 4.465e-05,
      "loss": 0.735,
      "step": 900
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.27552103996276855,
      "learning_rate": 4.5150000000000006e-05,
      "loss": 0.7082,
      "step": 910
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.2184416800737381,
      "learning_rate": 4.5650000000000005e-05,
      "loss": 0.7066,
      "step": 920
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.21738290786743164,
      "learning_rate": 4.6150000000000004e-05,
      "loss": 0.6977,
      "step": 930
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.21888123452663422,
      "learning_rate": 4.665e-05,
      "loss": 0.7083,
      "step": 940
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.2516074776649475,
      "learning_rate": 4.715e-05,
      "loss": 0.7451,
      "step": 950
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.39361557364463806,
      "learning_rate": 4.765e-05,
      "loss": 0.7013,
      "step": 960
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.2461744099855423,
      "learning_rate": 4.815e-05,
      "loss": 0.7325,
      "step": 970
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.2554089426994324,
      "learning_rate": 4.8650000000000003e-05,
      "loss": 0.7384,
      "step": 980
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.2717374861240387,
      "learning_rate": 4.915e-05,
      "loss": 0.7172,
      "step": 990
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.35919347405433655,
      "learning_rate": 4.965e-05,
      "loss": 0.6851,
      "step": 1000
    },
    {
      "epoch": 0.24,
      "eval_loss": 0.7169972658157349,
      "eval_runtime": 67.4497,
      "eval_samples_per_second": 29.652,
      "eval_steps_per_second": 0.934,
      "step": 1000
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.19895273447036743,
      "learning_rate": 4.999999713608037e-05,
      "loss": 0.7104,
      "step": 1010
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.21741873025894165,
      "learning_rate": 4.999994622197174e-05,
      "loss": 0.7217,
      "step": 1020
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.3415576219558716,
      "learning_rate": 4.999983166535371e-05,
      "loss": 0.716,
      "step": 1030
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.22503221035003662,
      "learning_rate": 4.99996534665179e-05,
      "loss": 0.7166,
      "step": 1040
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.22469323873519897,
      "learning_rate": 4.999941162591795e-05,
      "loss": 0.7173,
      "step": 1050
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.24848666787147522,
      "learning_rate": 4.999910614416952e-05,
      "loss": 0.7442,
      "step": 1060
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.23361551761627197,
      "learning_rate": 4.999873702205027e-05,
      "loss": 0.7142,
      "step": 1070
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.2513304054737091,
      "learning_rate": 4.999830426049987e-05,
      "loss": 0.6998,
      "step": 1080
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.2846020758152008,
      "learning_rate": 4.999780786062003e-05,
      "loss": 0.6967,
      "step": 1090
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.2192545235157013,
      "learning_rate": 4.999724782367441e-05,
      "loss": 0.7259,
      "step": 1100
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.37123605608940125,
      "learning_rate": 4.999662415108872e-05,
      "loss": 0.7228,
      "step": 1110
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.3963713049888611,
      "learning_rate": 4.999593684445063e-05,
      "loss": 0.6848,
      "step": 1120
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.24653227627277374,
      "learning_rate": 4.9995185905509836e-05,
      "loss": 0.7181,
      "step": 1130
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.25218069553375244,
      "learning_rate": 4.999437133617799e-05,
      "loss": 0.7263,
      "step": 1140
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.20412948727607727,
      "learning_rate": 4.9993493138528765e-05,
      "loss": 0.6965,
      "step": 1150
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.21635311841964722,
      "learning_rate": 4.9992551314797775e-05,
      "loss": 0.7042,
      "step": 1160
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.21901027858257294,
      "learning_rate": 4.999154586738264e-05,
      "loss": 0.6967,
      "step": 1170
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.24104245007038116,
      "learning_rate": 4.9990476798842935e-05,
      "loss": 0.7009,
      "step": 1180
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.2433982491493225,
      "learning_rate": 4.998934411190018e-05,
      "loss": 0.7052,
      "step": 1190
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.5261262655258179,
      "learning_rate": 4.9988147809437876e-05,
      "loss": 0.7103,
      "step": 1200
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.36266064643859863,
      "learning_rate": 4.998688789450146e-05,
      "loss": 0.7028,
      "step": 1210
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.2509474754333496,
      "learning_rate": 4.9985564370298274e-05,
      "loss": 0.7047,
      "step": 1220
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.1925598382949829,
      "learning_rate": 4.9984177240197665e-05,
      "loss": 0.6943,
      "step": 1230
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.22494031488895416,
      "learning_rate": 4.998272650773083e-05,
      "loss": 0.6953,
      "step": 1240
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.24152888357639313,
      "learning_rate": 4.998121217659092e-05,
      "loss": 0.7181,
      "step": 1250
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.24296504259109497,
      "learning_rate": 4.997963425063297e-05,
      "loss": 0.6934,
      "step": 1260
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.2438274323940277,
      "learning_rate": 4.9977992733873906e-05,
      "loss": 0.7088,
      "step": 1270
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.21081580221652985,
      "learning_rate": 4.997628763049257e-05,
      "loss": 0.6833,
      "step": 1280
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.2111068218946457,
      "learning_rate": 4.9974518944829626e-05,
      "loss": 0.6922,
      "step": 1290
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.309772253036499,
      "learning_rate": 4.997268668138766e-05,
      "loss": 0.7195,
      "step": 1300
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.2199670672416687,
      "learning_rate": 4.997079084483105e-05,
      "loss": 0.6924,
      "step": 1310
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.21781674027442932,
      "learning_rate": 4.996883143998605e-05,
      "loss": 0.6873,
      "step": 1320
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.2969052791595459,
      "learning_rate": 4.996680847184072e-05,
      "loss": 0.7135,
      "step": 1330
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.23092682659626007,
      "learning_rate": 4.996472194554495e-05,
      "loss": 0.6827,
      "step": 1340
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.29615628719329834,
      "learning_rate": 4.996257186641042e-05,
      "loss": 0.6738,
      "step": 1350
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.23491479456424713,
      "learning_rate": 4.99603582399106e-05,
      "loss": 0.7396,
      "step": 1360
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.1986425668001175,
      "learning_rate": 4.9958081071680726e-05,
      "loss": 0.6787,
      "step": 1370
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.19658410549163818,
      "learning_rate": 4.99557403675178e-05,
      "loss": 0.66,
      "step": 1380
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.29404616355895996,
      "learning_rate": 4.995333613338057e-05,
      "loss": 0.6824,
      "step": 1390
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.23689547181129456,
      "learning_rate": 4.9950868375389514e-05,
      "loss": 0.7008,
      "step": 1400
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.22758638858795166,
      "learning_rate": 4.99483370998268e-05,
      "loss": 0.6946,
      "step": 1410
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.20259805023670197,
      "learning_rate": 4.994574231313634e-05,
      "loss": 0.6734,
      "step": 1420
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.21935436129570007,
      "learning_rate": 4.994308402192366e-05,
      "loss": 0.7094,
      "step": 1430
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.23414015769958496,
      "learning_rate": 4.9940362232956026e-05,
      "loss": 0.7019,
      "step": 1440
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.2623121440410614,
      "learning_rate": 4.993757695316228e-05,
      "loss": 0.6793,
      "step": 1450
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.31137487292289734,
      "learning_rate": 4.993472818963295e-05,
      "loss": 0.6902,
      "step": 1460
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.22092680633068085,
      "learning_rate": 4.993181594962013e-05,
      "loss": 0.6912,
      "step": 1470
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.23723623156547546,
      "learning_rate": 4.992884024053754e-05,
      "loss": 0.652,
      "step": 1480
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.26028233766555786,
      "learning_rate": 4.9925801069960454e-05,
      "loss": 0.6969,
      "step": 1490
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.23856200277805328,
      "learning_rate": 4.992269844562572e-05,
      "loss": 0.6723,
      "step": 1500
    },
    {
      "epoch": 0.36,
      "eval_loss": 0.6912096738815308,
      "eval_runtime": 67.244,
      "eval_samples_per_second": 29.742,
      "eval_steps_per_second": 0.937,
      "step": 1500
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.167326420545578,
      "learning_rate": 4.9919532375431677e-05,
      "loss": 0.6897,
      "step": 1510
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.26183101534843445,
      "learning_rate": 4.991630286743823e-05,
      "loss": 0.6844,
      "step": 1520
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.25918036699295044,
      "learning_rate": 4.991300992986676e-05,
      "loss": 0.6645,
      "step": 1530
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.2354690134525299,
      "learning_rate": 4.99096535711001e-05,
      "loss": 0.6672,
      "step": 1540
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.24444998800754547,
      "learning_rate": 4.990623379968257e-05,
      "loss": 0.6749,
      "step": 1550
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.2562585771083832,
      "learning_rate": 4.990275062431989e-05,
      "loss": 0.6652,
      "step": 1560
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.22618699073791504,
      "learning_rate": 4.98992040538792e-05,
      "loss": 0.7163,
      "step": 1570
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.20957091450691223,
      "learning_rate": 4.9895594097389044e-05,
      "loss": 0.7053,
      "step": 1580
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.26973310112953186,
      "learning_rate": 4.989192076403928e-05,
      "loss": 0.6716,
      "step": 1590
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.24841250479221344,
      "learning_rate": 4.9888184063181154e-05,
      "loss": 0.6884,
      "step": 1600
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.25283706188201904,
      "learning_rate": 4.98843840043272e-05,
      "loss": 0.6632,
      "step": 1610
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.196313738822937,
      "learning_rate": 4.988052059715126e-05,
      "loss": 0.712,
      "step": 1620
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.21621187031269073,
      "learning_rate": 4.987659385148842e-05,
      "loss": 0.6732,
      "step": 1630
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.2180820256471634,
      "learning_rate": 4.987260377733502e-05,
      "loss": 0.7045,
      "step": 1640
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.20110243558883667,
      "learning_rate": 4.986855038484862e-05,
      "loss": 0.6604,
      "step": 1650
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.20813095569610596,
      "learning_rate": 4.9864433684347964e-05,
      "loss": 0.7051,
      "step": 1660
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.22026021778583527,
      "learning_rate": 4.9860253686312964e-05,
      "loss": 0.6752,
      "step": 1670
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.23409809172153473,
      "learning_rate": 4.9856010401384654e-05,
      "loss": 0.6805,
      "step": 1680
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.1982721984386444,
      "learning_rate": 4.985170384036521e-05,
      "loss": 0.6888,
      "step": 1690
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.24984286725521088,
      "learning_rate": 4.984733401421785e-05,
      "loss": 0.6741,
      "step": 1700
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.24465042352676392,
      "learning_rate": 4.9842900934066874e-05,
      "loss": 0.6781,
      "step": 1710
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.26656293869018555,
      "learning_rate": 4.98384046111976e-05,
      "loss": 0.7086,
      "step": 1720
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.23073184490203857,
      "learning_rate": 4.9833845057056336e-05,
      "loss": 0.6966,
      "step": 1730
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.23296333849430084,
      "learning_rate": 4.982922228325037e-05,
      "loss": 0.6769,
      "step": 1740
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.22282017767429352,
      "learning_rate": 4.982453630154794e-05,
      "loss": 0.6581,
      "step": 1750
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.22943562269210815,
      "learning_rate": 4.981978712387815e-05,
      "loss": 0.6803,
      "step": 1760
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.29295608401298523,
      "learning_rate": 4.9814974762331034e-05,
      "loss": 0.6514,
      "step": 1770
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.21845972537994385,
      "learning_rate": 4.981009922915743e-05,
      "loss": 0.691,
      "step": 1780
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.24616730213165283,
      "learning_rate": 4.980516053676903e-05,
      "loss": 0.6948,
      "step": 1790
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.2354935109615326,
      "learning_rate": 4.9800158697738264e-05,
      "loss": 0.6778,
      "step": 1800
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.3475668430328369,
      "learning_rate": 4.979509372479837e-05,
      "loss": 0.6626,
      "step": 1810
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.2817700207233429,
      "learning_rate": 4.9789965630843265e-05,
      "loss": 0.6724,
      "step": 1820
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.28306934237480164,
      "learning_rate": 4.978477442892758e-05,
      "loss": 0.6673,
      "step": 1830
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.2593139111995697,
      "learning_rate": 4.9779520132266575e-05,
      "loss": 0.6866,
      "step": 1840
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.2618165612220764,
      "learning_rate": 4.9774202754236145e-05,
      "loss": 0.6671,
      "step": 1850
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.22875259816646576,
      "learning_rate": 4.9768822308372784e-05,
      "loss": 0.6743,
      "step": 1860
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.20143887400627136,
      "learning_rate": 4.976337880837351e-05,
      "loss": 0.6866,
      "step": 1870
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.23163306713104248,
      "learning_rate": 4.975787226809587e-05,
      "loss": 0.6772,
      "step": 1880
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.21628862619400024,
      "learning_rate": 4.975230270155791e-05,
      "loss": 0.6706,
      "step": 1890
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.21055322885513306,
      "learning_rate": 4.9746670122938105e-05,
      "loss": 0.6536,
      "step": 1900
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.20045976340770721,
      "learning_rate": 4.974097454657534e-05,
      "loss": 0.6477,
      "step": 1910
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.20996223390102386,
      "learning_rate": 4.9735215986968874e-05,
      "loss": 0.6526,
      "step": 1920
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.1778746396303177,
      "learning_rate": 4.972939445877831e-05,
      "loss": 0.6951,
      "step": 1930
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.2416403591632843,
      "learning_rate": 4.972350997682354e-05,
      "loss": 0.6703,
      "step": 1940
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.23652783036231995,
      "learning_rate": 4.9717562556084735e-05,
      "loss": 0.6401,
      "step": 1950
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.23598232865333557,
      "learning_rate": 4.9711552211702274e-05,
      "loss": 0.6668,
      "step": 1960
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.22741574048995972,
      "learning_rate": 4.970547895897672e-05,
      "loss": 0.6765,
      "step": 1970
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.20418991148471832,
      "learning_rate": 4.96993428133688e-05,
      "loss": 0.6771,
      "step": 1980
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.24983160197734833,
      "learning_rate": 4.969314379049932e-05,
      "loss": 0.6656,
      "step": 1990
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.2511340379714966,
      "learning_rate": 4.968688190614919e-05,
      "loss": 0.6605,
      "step": 2000
    },
    {
      "epoch": 0.48,
      "eval_loss": 0.6730201244354248,
      "eval_runtime": 67.2807,
      "eval_samples_per_second": 29.726,
      "eval_steps_per_second": 0.936,
      "step": 2000
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.22646427154541016,
      "learning_rate": 4.96805571762593e-05,
      "loss": 0.6729,
      "step": 2010
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.2343759387731552,
      "learning_rate": 4.9674169616930574e-05,
      "loss": 0.6635,
      "step": 2020
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.23573775589466095,
      "learning_rate": 4.966771924442385e-05,
      "loss": 0.654,
      "step": 2030
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.2696162760257721,
      "learning_rate": 4.966120607515987e-05,
      "loss": 0.6728,
      "step": 2040
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.25032511353492737,
      "learning_rate": 4.965463012571927e-05,
      "loss": 0.6743,
      "step": 2050
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.24621817469596863,
      "learning_rate": 4.964799141284247e-05,
      "loss": 0.6596,
      "step": 2060
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.23296469449996948,
      "learning_rate": 4.964128995342966e-05,
      "loss": 0.6471,
      "step": 2070
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.22769325971603394,
      "learning_rate": 4.963452576454082e-05,
      "loss": 0.6499,
      "step": 2080
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.2572191655635834,
      "learning_rate": 4.9627698863395564e-05,
      "loss": 0.6499,
      "step": 2090
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.21782545745372772,
      "learning_rate": 4.962080926737319e-05,
      "loss": 0.6641,
      "step": 2100
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.22925855219364166,
      "learning_rate": 4.9613856994012567e-05,
      "loss": 0.6618,
      "step": 2110
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.2127605825662613,
      "learning_rate": 4.960684206101214e-05,
      "loss": 0.6821,
      "step": 2120
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.2520267069339752,
      "learning_rate": 4.9599764486229865e-05,
      "loss": 0.6748,
      "step": 2130
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.20282019674777985,
      "learning_rate": 4.9592624287683176e-05,
      "loss": 0.6674,
      "step": 2140
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.46226009726524353,
      "learning_rate": 4.958542148354891e-05,
      "loss": 0.6422,
      "step": 2150
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.2241724133491516,
      "learning_rate": 4.957815609216329e-05,
      "loss": 0.6612,
      "step": 2160
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.23772084712982178,
      "learning_rate": 4.957082813202186e-05,
      "loss": 0.6614,
      "step": 2170
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.2049807757139206,
      "learning_rate": 4.9563437621779465e-05,
      "loss": 0.6827,
      "step": 2180
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.22666208446025848,
      "learning_rate": 4.955598458025015e-05,
      "loss": 0.6522,
      "step": 2190
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.26957106590270996,
      "learning_rate": 4.954846902640718e-05,
      "loss": 0.6817,
      "step": 2200
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.22548554837703705,
      "learning_rate": 4.954089097938294e-05,
      "loss": 0.6444,
      "step": 2210
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.22197945415973663,
      "learning_rate": 4.9533250458468914e-05,
      "loss": 0.6779,
      "step": 2220
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.22711919248104095,
      "learning_rate": 4.9525547483115617e-05,
      "loss": 0.6416,
      "step": 2230
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.22576060891151428,
      "learning_rate": 4.951778207293255e-05,
      "loss": 0.6598,
      "step": 2240
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.22785042226314545,
      "learning_rate": 4.950995424768818e-05,
      "loss": 0.6499,
      "step": 2250
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.20107567310333252,
      "learning_rate": 4.9502064027309836e-05,
      "loss": 0.6509,
      "step": 2260
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.23379752039909363,
      "learning_rate": 4.94941114318837e-05,
      "loss": 0.6551,
      "step": 2270
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.22154928743839264,
      "learning_rate": 4.948609648165475e-05,
      "loss": 0.6428,
      "step": 2280
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.23805582523345947,
      "learning_rate": 4.947801919702667e-05,
      "loss": 0.6633,
      "step": 2290
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.24756357073783875,
      "learning_rate": 4.946987959856188e-05,
      "loss": 0.6561,
      "step": 2300
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.23767632246017456,
      "learning_rate": 4.9461677706981374e-05,
      "loss": 0.6552,
      "step": 2310
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.32177987694740295,
      "learning_rate": 4.9453413543164775e-05,
      "loss": 0.6705,
      "step": 2320
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.29961881041526794,
      "learning_rate": 4.94450871281502e-05,
      "loss": 0.6551,
      "step": 2330
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.2768830955028534,
      "learning_rate": 4.943669848313427e-05,
      "loss": 0.6468,
      "step": 2340
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.24728989601135254,
      "learning_rate": 4.9428247629472e-05,
      "loss": 0.6506,
      "step": 2350
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.2641989290714264,
      "learning_rate": 4.941973458867677e-05,
      "loss": 0.6517,
      "step": 2360
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.18787774443626404,
      "learning_rate": 4.941115938242028e-05,
      "loss": 0.6355,
      "step": 2370
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.26642099022865295,
      "learning_rate": 4.940252203253248e-05,
      "loss": 0.6811,
      "step": 2380
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.24736198782920837,
      "learning_rate": 4.939382256100154e-05,
      "loss": 0.6493,
      "step": 2390
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.3060832619667053,
      "learning_rate": 4.938506098997374e-05,
      "loss": 0.6707,
      "step": 2400
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.24164025485515594,
      "learning_rate": 4.937623734175346e-05,
      "loss": 0.6508,
      "step": 2410
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.26685982942581177,
      "learning_rate": 4.936735163880313e-05,
      "loss": 0.6441,
      "step": 2420
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.17739726603031158,
      "learning_rate": 4.9358403903743124e-05,
      "loss": 0.6499,
      "step": 2430
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.23560577630996704,
      "learning_rate": 4.9349394159351735e-05,
      "loss": 0.645,
      "step": 2440
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.23137669265270233,
      "learning_rate": 4.9340322428565135e-05,
      "loss": 0.6615,
      "step": 2450
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.3485463857650757,
      "learning_rate": 4.933118873447728e-05,
      "loss": 0.6283,
      "step": 2460
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.20602945983409882,
      "learning_rate": 4.932199310033987e-05,
      "loss": 0.6831,
      "step": 2470
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.21541930735111237,
      "learning_rate": 4.931273554956227e-05,
      "loss": 0.6722,
      "step": 2480
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.2092689573764801,
      "learning_rate": 4.930341610571151e-05,
      "loss": 0.6609,
      "step": 2490
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.23792122304439545,
      "learning_rate": 4.9294034792512126e-05,
      "loss": 0.6475,
      "step": 2500
    },
    {
      "epoch": 0.6,
      "eval_loss": 0.6643247008323669,
      "eval_runtime": 67.0911,
      "eval_samples_per_second": 29.81,
      "eval_steps_per_second": 0.939,
      "step": 2500
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.2101823091506958,
      "learning_rate": 4.928459163384619e-05,
      "loss": 0.6604,
      "step": 2510
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.1859540194272995,
      "learning_rate": 4.927508665375321e-05,
      "loss": 0.6252,
      "step": 2520
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.2371070235967636,
      "learning_rate": 4.926551987643007e-05,
      "loss": 0.6413,
      "step": 2530
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.2149161994457245,
      "learning_rate": 4.9255891326230964e-05,
      "loss": 0.6558,
      "step": 2540
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.21805475652217865,
      "learning_rate": 4.9246201027667354e-05,
      "loss": 0.631,
      "step": 2550
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.22705689072608948,
      "learning_rate": 4.9236449005407895e-05,
      "loss": 0.6785,
      "step": 2560
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.21902668476104736,
      "learning_rate": 4.9226635284278355e-05,
      "loss": 0.657,
      "step": 2570
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.21482336521148682,
      "learning_rate": 4.9216759889261586e-05,
      "loss": 0.6514,
      "step": 2580
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.20008622109889984,
      "learning_rate": 4.9206822845497444e-05,
      "loss": 0.6516,
      "step": 2590
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.205527663230896,
      "learning_rate": 4.919682417828271e-05,
      "loss": 0.6336,
      "step": 2600
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.263959139585495,
      "learning_rate": 4.9186763913071065e-05,
      "loss": 0.6449,
      "step": 2610
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.21950668096542358,
      "learning_rate": 4.917664207547297e-05,
      "loss": 0.6916,
      "step": 2620
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.215659499168396,
      "learning_rate": 4.916645869125564e-05,
      "loss": 0.6441,
      "step": 2630
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.22789780795574188,
      "learning_rate": 4.915621378634301e-05,
      "loss": 0.6576,
      "step": 2640
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.27431604266166687,
      "learning_rate": 4.914590738681555e-05,
      "loss": 0.6487,
      "step": 2650
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.21077977120876312,
      "learning_rate": 4.913553951891036e-05,
      "loss": 0.6688,
      "step": 2660
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.2798519730567932,
      "learning_rate": 4.9125110209020954e-05,
      "loss": 0.6618,
      "step": 2670
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.24790987372398376,
      "learning_rate": 4.911461948369731e-05,
      "loss": 0.6598,
      "step": 2680
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.2008027732372284,
      "learning_rate": 4.91040673696457e-05,
      "loss": 0.6388,
      "step": 2690
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.23889446258544922,
      "learning_rate": 4.9093453893728733e-05,
      "loss": 0.6526,
      "step": 2700
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.1997964233160019,
      "learning_rate": 4.908277908296518e-05,
      "loss": 0.6521,
      "step": 2710
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.24335375428199768,
      "learning_rate": 4.907204296452997e-05,
      "loss": 0.6636,
      "step": 2720
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.23431982100009918,
      "learning_rate": 4.906124556575411e-05,
      "loss": 0.6557,
      "step": 2730
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.21532365679740906,
      "learning_rate": 4.90503869141246e-05,
      "loss": 0.628,
      "step": 2740
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.22962962090969086,
      "learning_rate": 4.903946703728436e-05,
      "loss": 0.6422,
      "step": 2750
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.22807002067565918,
      "learning_rate": 4.90284859630322e-05,
      "loss": 0.6631,
      "step": 2760
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.27668192982673645,
      "learning_rate": 4.90174437193227e-05,
      "loss": 0.6778,
      "step": 2770
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.2503020763397217,
      "learning_rate": 4.900634033426616e-05,
      "loss": 0.6483,
      "step": 2780
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.3033277094364166,
      "learning_rate": 4.8995175836128536e-05,
      "loss": 0.6572,
      "step": 2790
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.2776722311973572,
      "learning_rate": 4.898395025333136e-05,
      "loss": 0.662,
      "step": 2800
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.31251537799835205,
      "learning_rate": 4.897266361445165e-05,
      "loss": 0.6617,
      "step": 2810
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.38570669293403625,
      "learning_rate": 4.8961315948221884e-05,
      "loss": 0.6419,
      "step": 2820
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.23277431726455688,
      "learning_rate": 4.894990728352988e-05,
      "loss": 0.6373,
      "step": 2830
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.2519891858100891,
      "learning_rate": 4.893843764941874e-05,
      "loss": 0.6759,
      "step": 2840
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.2358052134513855,
      "learning_rate": 4.892690707508677e-05,
      "loss": 0.6826,
      "step": 2850
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.23103272914886475,
      "learning_rate": 4.8915315589887436e-05,
      "loss": 0.624,
      "step": 2860
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.2128864973783493,
      "learning_rate": 4.8903663223329245e-05,
      "loss": 0.654,
      "step": 2870
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.22060807049274445,
      "learning_rate": 4.889195000507568e-05,
      "loss": 0.6376,
      "step": 2880
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.24493062496185303,
      "learning_rate": 4.888017596494517e-05,
      "loss": 0.6648,
      "step": 2890
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.2580796480178833,
      "learning_rate": 4.886834113291094e-05,
      "loss": 0.6607,
      "step": 2900
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.26906147599220276,
      "learning_rate": 4.8856445539101016e-05,
      "loss": 0.6441,
      "step": 2910
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.2145451009273529,
      "learning_rate": 4.884448921379805e-05,
      "loss": 0.6543,
      "step": 2920
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.2156788855791092,
      "learning_rate": 4.8832472187439345e-05,
      "loss": 0.6185,
      "step": 2930
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.25632360577583313,
      "learning_rate": 4.882039449061673e-05,
      "loss": 0.639,
      "step": 2940
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.2496800720691681,
      "learning_rate": 4.8808256154076436e-05,
      "loss": 0.6377,
      "step": 2950
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.39942485094070435,
      "learning_rate": 4.8796057208719124e-05,
      "loss": 0.6083,
      "step": 2960
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.27986374497413635,
      "learning_rate": 4.8783797685599706e-05,
      "loss": 0.6528,
      "step": 2970
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.22164517641067505,
      "learning_rate": 4.877147761592733e-05,
      "loss": 0.6456,
      "step": 2980
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.29858505725860596,
      "learning_rate": 4.875909703106527e-05,
      "loss": 0.643,
      "step": 2990
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.22595979273319244,
      "learning_rate": 4.874665596253084e-05,
      "loss": 0.6419,
      "step": 3000
    },
    {
      "epoch": 0.72,
      "eval_loss": 0.6583871841430664,
      "eval_runtime": 67.1961,
      "eval_samples_per_second": 29.764,
      "eval_steps_per_second": 0.938,
      "step": 3000
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.19871610403060913,
      "learning_rate": 4.8734154441995364e-05,
      "loss": 0.6556,
      "step": 3010
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.2072724997997284,
      "learning_rate": 4.872159250128401e-05,
      "loss": 0.6764,
      "step": 3020
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.22515198588371277,
      "learning_rate": 4.87089701723758e-05,
      "loss": 0.6703,
      "step": 3030
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.2292303442955017,
      "learning_rate": 4.869628748740347e-05,
      "loss": 0.6288,
      "step": 3040
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.25133493542671204,
      "learning_rate": 4.8683544478653395e-05,
      "loss": 0.6652,
      "step": 3050
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.29863354563713074,
      "learning_rate": 4.867074117856555e-05,
      "loss": 0.6352,
      "step": 3060
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.2168973684310913,
      "learning_rate": 4.865787761973334e-05,
      "loss": 0.6562,
      "step": 3070
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.2312694936990738,
      "learning_rate": 4.864495383490363e-05,
      "loss": 0.6463,
      "step": 3080
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.2333671748638153,
      "learning_rate": 4.863196985697655e-05,
      "loss": 0.6267,
      "step": 3090
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.23559433221817017,
      "learning_rate": 4.86189257190055e-05,
      "loss": 0.623,
      "step": 3100
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.22314316034317017,
      "learning_rate": 4.860582145419703e-05,
      "loss": 0.6298,
      "step": 3110
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.22701682150363922,
      "learning_rate": 4.859265709591073e-05,
      "loss": 0.6359,
      "step": 3120
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.21683742105960846,
      "learning_rate": 4.857943267765919e-05,
      "loss": 0.6666,
      "step": 3130
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.2519407272338867,
      "learning_rate": 4.856614823310788e-05,
      "loss": 0.6731,
      "step": 3140
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.23536355793476105,
      "learning_rate": 4.855280379607509e-05,
      "loss": 0.629,
      "step": 3150
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.19676193594932556,
      "learning_rate": 4.853939940053184e-05,
      "loss": 0.6329,
      "step": 3160
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.23497731983661652,
      "learning_rate": 4.852593508060177e-05,
      "loss": 0.6584,
      "step": 3170
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.2515724003314972,
      "learning_rate": 4.8512410870561084e-05,
      "loss": 0.6177,
      "step": 3180
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.20130635797977448,
      "learning_rate": 4.8498826804838436e-05,
      "loss": 0.6645,
      "step": 3190
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.23828361928462982,
      "learning_rate": 4.8485182918014876e-05,
      "loss": 0.6401,
      "step": 3200
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.2598424255847931,
      "learning_rate": 4.847147924482371e-05,
      "loss": 0.6193,
      "step": 3210
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.20697768032550812,
      "learning_rate": 4.845771582015046e-05,
      "loss": 0.626,
      "step": 3220
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.19661648571491241,
      "learning_rate": 4.8443892679032775e-05,
      "loss": 0.6283,
      "step": 3230
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.22154510021209717,
      "learning_rate": 4.843000985666028e-05,
      "loss": 0.6571,
      "step": 3240
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.3334987759590149,
      "learning_rate": 4.841606738837458e-05,
      "loss": 0.6553,
      "step": 3250
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.6604078412055969,
      "learning_rate": 4.8402065309669085e-05,
      "loss": 0.6239,
      "step": 3260
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.22238047420978546,
      "learning_rate": 4.838800365618898e-05,
      "loss": 0.6725,
      "step": 3270
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.21566419303417206,
      "learning_rate": 4.837388246373108e-05,
      "loss": 0.6597,
      "step": 3280
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.9337890148162842,
      "learning_rate": 4.83597017682438e-05,
      "loss": 0.6454,
      "step": 3290
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.2716673016548157,
      "learning_rate": 4.8345461605827014e-05,
      "loss": 0.6526,
      "step": 3300
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.23388756811618805,
      "learning_rate": 4.8331162012732e-05,
      "loss": 0.6642,
      "step": 3310
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.21243353188037872,
      "learning_rate": 4.8316803025361304e-05,
      "loss": 0.6233,
      "step": 3320
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.28600814938545227,
      "learning_rate": 4.8302384680268684e-05,
      "loss": 0.6559,
      "step": 3330
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.19718654453754425,
      "learning_rate": 4.8287907014159004e-05,
      "loss": 0.6357,
      "step": 3340
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.2330627292394638,
      "learning_rate": 4.827337006388816e-05,
      "loss": 0.6692,
      "step": 3350
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.18776822090148926,
      "learning_rate": 4.8258773866462927e-05,
      "loss": 0.6804,
      "step": 3360
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.24929189682006836,
      "learning_rate": 4.8244118459040944e-05,
      "loss": 0.6455,
      "step": 3370
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.23884710669517517,
      "learning_rate": 4.8229403878930566e-05,
      "loss": 0.6345,
      "step": 3380
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.2579725682735443,
      "learning_rate": 4.821463016359078e-05,
      "loss": 0.6584,
      "step": 3390
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.2741701900959015,
      "learning_rate": 4.8199797350631136e-05,
      "loss": 0.6463,
      "step": 3400
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.2254386991262436,
      "learning_rate": 4.818490547781159e-05,
      "loss": 0.6399,
      "step": 3410
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.2172977328300476,
      "learning_rate": 4.816995458304249e-05,
      "loss": 0.661,
      "step": 3420
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.23464016616344452,
      "learning_rate": 4.815494470438441e-05,
      "loss": 0.6516,
      "step": 3430
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.27163979411125183,
      "learning_rate": 4.813987588004807e-05,
      "loss": 0.6451,
      "step": 3440
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.20852161943912506,
      "learning_rate": 4.8124748148394285e-05,
      "loss": 0.6394,
      "step": 3450
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.2224467694759369,
      "learning_rate": 4.81095615479338e-05,
      "loss": 0.6427,
      "step": 3460
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.2205616980791092,
      "learning_rate": 4.8094316117327245e-05,
      "loss": 0.6491,
      "step": 3470
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.21984492242336273,
      "learning_rate": 4.8079011895384985e-05,
      "loss": 0.6271,
      "step": 3480
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.29770565032958984,
      "learning_rate": 4.806364892106707e-05,
      "loss": 0.6048,
      "step": 3490
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.2440354973077774,
      "learning_rate": 4.8048227233483127e-05,
      "loss": 0.6307,
      "step": 3500
    },
    {
      "epoch": 0.85,
      "eval_loss": 0.6532164216041565,
      "eval_runtime": 67.3224,
      "eval_samples_per_second": 29.708,
      "eval_steps_per_second": 0.936,
      "step": 3500
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.25720447301864624,
      "learning_rate": 4.803274687189222e-05,
      "loss": 0.6337,
      "step": 3510
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.21051880717277527,
      "learning_rate": 4.8017207875702814e-05,
      "loss": 0.6398,
      "step": 3520
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.2977493107318878,
      "learning_rate": 4.800161028447261e-05,
      "loss": 0.6346,
      "step": 3530
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.19749422371387482,
      "learning_rate": 4.798595413790848e-05,
      "loss": 0.6325,
      "step": 3540
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.21522672474384308,
      "learning_rate": 4.7970239475866386e-05,
      "loss": 0.6244,
      "step": 3550
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.24835069477558136,
      "learning_rate": 4.7954466338351224e-05,
      "loss": 0.612,
      "step": 3560
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.25459495186805725,
      "learning_rate": 4.793863476551677e-05,
      "loss": 0.6317,
      "step": 3570
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.28858670592308044,
      "learning_rate": 4.7922744797665544e-05,
      "loss": 0.6485,
      "step": 3580
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.30466488003730774,
      "learning_rate": 4.790679647524873e-05,
      "loss": 0.6307,
      "step": 3590
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.36087316274642944,
      "learning_rate": 4.789078983886607e-05,
      "loss": 0.6749,
      "step": 3600
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.2436164766550064,
      "learning_rate": 4.787472492926575e-05,
      "loss": 0.6239,
      "step": 3610
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.2527044415473938,
      "learning_rate": 4.78586017873443e-05,
      "loss": 0.6584,
      "step": 3620
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.25492456555366516,
      "learning_rate": 4.784242045414651e-05,
      "loss": 0.6487,
      "step": 3630
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.2293190211057663,
      "learning_rate": 4.782618097086528e-05,
      "loss": 0.646,
      "step": 3640
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.7955949902534485,
      "learning_rate": 4.780988337884157e-05,
      "loss": 0.6391,
      "step": 3650
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.2331811785697937,
      "learning_rate": 4.779352771956425e-05,
      "loss": 0.6267,
      "step": 3660
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.23316140472888947,
      "learning_rate": 4.7777114034670006e-05,
      "loss": 0.6405,
      "step": 3670
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.2253628969192505,
      "learning_rate": 4.776064236594327e-05,
      "loss": 0.6415,
      "step": 3680
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.2378271222114563,
      "learning_rate": 4.774411275531606e-05,
      "loss": 0.6362,
      "step": 3690
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.22514688968658447,
      "learning_rate": 4.7727525244867896e-05,
      "loss": 0.6375,
      "step": 3700
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.20843249559402466,
      "learning_rate": 4.771087987682571e-05,
      "loss": 0.6166,
      "step": 3710
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.2776215076446533,
      "learning_rate": 4.7694176693563705e-05,
      "loss": 0.6441,
      "step": 3720
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.21339628100395203,
      "learning_rate": 4.767741573760327e-05,
      "loss": 0.6388,
      "step": 3730
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.24429263174533844,
      "learning_rate": 4.766059705161288e-05,
      "loss": 0.6263,
      "step": 3740
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.2099929004907608,
      "learning_rate": 4.764372067840795e-05,
      "loss": 0.6383,
      "step": 3750
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.22066868841648102,
      "learning_rate": 4.7626786660950784e-05,
      "loss": 0.6313,
      "step": 3760
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.22953028976917267,
      "learning_rate": 4.760979504235038e-05,
      "loss": 0.6264,
      "step": 3770
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.3549204170703888,
      "learning_rate": 4.759274586586242e-05,
      "loss": 0.6368,
      "step": 3780
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.24681401252746582,
      "learning_rate": 4.757563917488909e-05,
      "loss": 0.6422,
      "step": 3790
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.2417946755886078,
      "learning_rate": 4.755847501297898e-05,
      "loss": 0.6438,
      "step": 3800
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.2778252959251404,
      "learning_rate": 4.7541253423827006e-05,
      "loss": 0.6312,
      "step": 3810
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.20280030369758606,
      "learning_rate": 4.7523974451274275e-05,
      "loss": 0.6349,
      "step": 3820
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.24433667957782745,
      "learning_rate": 4.7506638139307966e-05,
      "loss": 0.6648,
      "step": 3830
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.3273336887359619,
      "learning_rate": 4.7489244532061225e-05,
      "loss": 0.6498,
      "step": 3840
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.23943009972572327,
      "learning_rate": 4.747179367381307e-05,
      "loss": 0.6358,
      "step": 3850
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.26670822501182556,
      "learning_rate": 4.745428560898824e-05,
      "loss": 0.6563,
      "step": 3860
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.18882083892822266,
      "learning_rate": 4.7436720382157116e-05,
      "loss": 0.6358,
      "step": 3870
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.23588411509990692,
      "learning_rate": 4.741909803803562e-05,
      "loss": 0.6396,
      "step": 3880
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.24063749611377716,
      "learning_rate": 4.740141862148503e-05,
      "loss": 0.6412,
      "step": 3890
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.19155845046043396,
      "learning_rate": 4.738368217751196e-05,
      "loss": 0.62,
      "step": 3900
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.20259159803390503,
      "learning_rate": 4.736588875126816e-05,
      "loss": 0.6172,
      "step": 3910
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.22824090719223022,
      "learning_rate": 4.734803838805048e-05,
      "loss": 0.6286,
      "step": 3920
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.2356242537498474,
      "learning_rate": 4.7330131133300686e-05,
      "loss": 0.6121,
      "step": 3930
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.22052697837352753,
      "learning_rate": 4.731216703260538e-05,
      "loss": 0.6402,
      "step": 3940
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.2817922532558441,
      "learning_rate": 4.7294146131695874e-05,
      "loss": 0.6256,
      "step": 3950
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.23476284742355347,
      "learning_rate": 4.7276068476448097e-05,
      "loss": 0.644,
      "step": 3960
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.25303345918655396,
      "learning_rate": 4.725793411288242e-05,
      "loss": 0.649,
      "step": 3970
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.2615244686603546,
      "learning_rate": 4.723974308716361e-05,
      "loss": 0.6158,
      "step": 3980
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.2798680365085602,
      "learning_rate": 4.722149544560067e-05,
      "loss": 0.6149,
      "step": 3990
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.19780561327934265,
      "learning_rate": 4.720319123464672e-05,
      "loss": 0.6167,
      "step": 4000
    },
    {
      "epoch": 0.97,
      "eval_loss": 0.6494756937026978,
      "eval_runtime": 67.2363,
      "eval_samples_per_second": 29.746,
      "eval_steps_per_second": 0.937,
      "step": 4000
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.22284342348575592,
      "learning_rate": 4.718483050089891e-05,
      "loss": 0.6282,
      "step": 4010
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.242294043302536,
      "learning_rate": 4.7166413291098246e-05,
      "loss": 0.6418,
      "step": 4020
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.19954660534858704,
      "learning_rate": 4.714793965212955e-05,
      "loss": 0.6293,
      "step": 4030
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.2239541858434677,
      "learning_rate": 4.712940963102126e-05,
      "loss": 0.6441,
      "step": 4040
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.2324070930480957,
      "learning_rate": 4.711082327494536e-05,
      "loss": 0.6355,
      "step": 4050
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.27540045976638794,
      "learning_rate": 4.709218063121725e-05,
      "loss": 0.6476,
      "step": 4060
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.37015092372894287,
      "learning_rate": 4.7073481747295614e-05,
      "loss": 0.6161,
      "step": 4070
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.2678948938846588,
      "learning_rate": 4.7054726670782304e-05,
      "loss": 0.6404,
      "step": 4080
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.2190527468919754,
      "learning_rate": 4.703591544942224e-05,
      "loss": 0.6619,
      "step": 4090
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.23498305678367615,
      "learning_rate": 4.701704813110325e-05,
      "loss": 0.6297,
      "step": 4100
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.2132258415222168,
      "learning_rate": 4.6998124763855984e-05,
      "loss": 0.6462,
      "step": 4110
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.2967916429042816,
      "learning_rate": 4.697914539585376e-05,
      "loss": 0.6347,
      "step": 4120
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.2503927946090698,
      "learning_rate": 4.6960110075412473e-05,
      "loss": 0.6264,
      "step": 4130
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.2261209934949875,
      "learning_rate": 4.694101885099045e-05,
      "loss": 0.6351,
      "step": 4140
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.2164410799741745,
      "learning_rate": 4.692187177118832e-05,
      "loss": 0.626,
      "step": 4150
    },
    {
      "epoch": 1.01,
      "grad_norm": 0.2934332489967346,
      "learning_rate": 4.690266888474893e-05,
      "loss": 0.6243,
      "step": 4160
    },
    {
      "epoch": 1.01,
      "grad_norm": 0.23883183300495148,
      "learning_rate": 4.688341024055718e-05,
      "loss": 0.624,
      "step": 4170
    },
    {
      "epoch": 1.01,
      "grad_norm": 0.1730484664440155,
      "learning_rate": 4.686409588763991e-05,
      "loss": 0.6473,
      "step": 4180
    },
    {
      "epoch": 1.01,
      "grad_norm": 0.27762770652770996,
      "learning_rate": 4.6844725875165775e-05,
      "loss": 0.6508,
      "step": 4190
    },
    {
      "epoch": 1.01,
      "grad_norm": 0.243989497423172,
      "learning_rate": 4.682530025244514e-05,
      "loss": 0.6305,
      "step": 4200
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.21361881494522095,
      "learning_rate": 4.6805819068929925e-05,
      "loss": 0.5969,
      "step": 4210
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.22126366198062897,
      "learning_rate": 4.678628237421348e-05,
      "loss": 0.6192,
      "step": 4220
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.2569611370563507,
      "learning_rate": 4.6766690218030495e-05,
      "loss": 0.6235,
      "step": 4230
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.26861709356307983,
      "learning_rate": 4.674704265025683e-05,
      "loss": 0.6413,
      "step": 4240
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.19330435991287231,
      "learning_rate": 4.672733972090943e-05,
      "loss": 0.6413,
      "step": 4250
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.25647568702697754,
      "learning_rate": 4.6707581480146136e-05,
      "loss": 0.6411,
      "step": 4260
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.23879855871200562,
      "learning_rate": 4.6687767978265625e-05,
      "loss": 0.6188,
      "step": 4270
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.25180870294570923,
      "learning_rate": 4.666789926570725e-05,
      "loss": 0.6388,
      "step": 4280
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.26070743799209595,
      "learning_rate": 4.6647975393050904e-05,
      "loss": 0.6254,
      "step": 4290
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.2934954762458801,
      "learning_rate": 4.662799641101691e-05,
      "loss": 0.6089,
      "step": 4300
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.24827881157398224,
      "learning_rate": 4.6607962370465866e-05,
      "loss": 0.6233,
      "step": 4310
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.7880712151527405,
      "learning_rate": 4.658787332239856e-05,
      "loss": 0.6305,
      "step": 4320
    },
    {
      "epoch": 1.05,
      "grad_norm": 0.23941326141357422,
      "learning_rate": 4.6567729317955796e-05,
      "loss": 0.6449,
      "step": 4330
    },
    {
      "epoch": 1.05,
      "grad_norm": 0.21050909161567688,
      "learning_rate": 4.654753040841829e-05,
      "loss": 0.6324,
      "step": 4340
    },
    {
      "epoch": 1.05,
      "grad_norm": 0.2372525930404663,
      "learning_rate": 4.6527276645206516e-05,
      "loss": 0.6098,
      "step": 4350
    },
    {
      "epoch": 1.05,
      "grad_norm": 0.23655132949352264,
      "learning_rate": 4.65069680798806e-05,
      "loss": 0.6457,
      "step": 4360
    },
    {
      "epoch": 1.06,
      "grad_norm": 0.26810187101364136,
      "learning_rate": 4.648660476414017e-05,
      "loss": 0.6327,
      "step": 4370
    },
    {
      "epoch": 1.06,
      "grad_norm": 0.2635047137737274,
      "learning_rate": 4.6466186749824235e-05,
      "loss": 0.6359,
      "step": 4380
    },
    {
      "epoch": 1.06,
      "grad_norm": 0.2955858111381531,
      "learning_rate": 4.6445714088911076e-05,
      "loss": 0.6412,
      "step": 4390
    },
    {
      "epoch": 1.06,
      "grad_norm": 0.2971380949020386,
      "learning_rate": 4.6425186833518054e-05,
      "loss": 0.61,
      "step": 4400
    },
    {
      "epoch": 1.07,
      "grad_norm": 0.22041000425815582,
      "learning_rate": 4.6404605035901505e-05,
      "loss": 0.6494,
      "step": 4410
    },
    {
      "epoch": 1.07,
      "grad_norm": 0.2463519424200058,
      "learning_rate": 4.638396874845666e-05,
      "loss": 0.6223,
      "step": 4420
    },
    {
      "epoch": 1.07,
      "grad_norm": 0.2519693374633789,
      "learning_rate": 4.636327802371742e-05,
      "loss": 0.6399,
      "step": 4430
    },
    {
      "epoch": 1.07,
      "grad_norm": 0.23022256791591644,
      "learning_rate": 4.6342532914356284e-05,
      "loss": 0.6106,
      "step": 4440
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.21000300347805023,
      "learning_rate": 4.632173347318421e-05,
      "loss": 0.6463,
      "step": 4450
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.2383776307106018,
      "learning_rate": 4.630087975315045e-05,
      "loss": 0.6202,
      "step": 4460
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.2528972327709198,
      "learning_rate": 4.627997180734244e-05,
      "loss": 0.6401,
      "step": 4470
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.20749187469482422,
      "learning_rate": 4.625900968898565e-05,
      "loss": 0.6541,
      "step": 4480
    },
    {
      "epoch": 1.09,
      "grad_norm": 0.2283320277929306,
      "learning_rate": 4.623799345144348e-05,
      "loss": 0.6311,
      "step": 4490
    },
    {
      "epoch": 1.09,
      "grad_norm": 0.21833132207393646,
      "learning_rate": 4.6216923148217096e-05,
      "loss": 0.6272,
      "step": 4500
    },
    {
      "epoch": 1.09,
      "eval_loss": 0.6476519107818604,
      "eval_runtime": 67.279,
      "eval_samples_per_second": 29.727,
      "eval_steps_per_second": 0.936,
      "step": 4500
    },
    {
      "epoch": 1.09,
      "grad_norm": 0.2190130203962326,
      "learning_rate": 4.619579883294528e-05,
      "loss": 0.6405,
      "step": 4510
    },
    {
      "epoch": 1.09,
      "grad_norm": 0.32068029046058655,
      "learning_rate": 4.617462055940433e-05,
      "loss": 0.6101,
      "step": 4520
    },
    {
      "epoch": 1.09,
      "grad_norm": 0.2526598572731018,
      "learning_rate": 4.6153388381507886e-05,
      "loss": 0.6252,
      "step": 4530
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.19768927991390228,
      "learning_rate": 4.613210235330686e-05,
      "loss": 0.6138,
      "step": 4540
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.24341870844364166,
      "learning_rate": 4.611076252898919e-05,
      "loss": 0.6072,
      "step": 4550
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.2283903807401657,
      "learning_rate": 4.60893689628798e-05,
      "loss": 0.6396,
      "step": 4560
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.21138328313827515,
      "learning_rate": 4.606792170944041e-05,
      "loss": 0.6306,
      "step": 4570
    },
    {
      "epoch": 1.11,
      "grad_norm": 0.4939797818660736,
      "learning_rate": 4.604642082326944e-05,
      "loss": 0.6259,
      "step": 4580
    },
    {
      "epoch": 1.11,
      "grad_norm": 0.22861245274543762,
      "learning_rate": 4.60248663591018e-05,
      "loss": 0.6274,
      "step": 4590
    },
    {
      "epoch": 1.11,
      "grad_norm": 0.35914289951324463,
      "learning_rate": 4.6003258371808825e-05,
      "loss": 0.6357,
      "step": 4600
    },
    {
      "epoch": 1.11,
      "grad_norm": 0.2684130072593689,
      "learning_rate": 4.598159691639809e-05,
      "loss": 0.6299,
      "step": 4610
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.2696840763092041,
      "learning_rate": 4.5959882048013294e-05,
      "loss": 0.6224,
      "step": 4620
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.21357202529907227,
      "learning_rate": 4.59381138219341e-05,
      "loss": 0.6119,
      "step": 4630
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.20233683288097382,
      "learning_rate": 4.591629229357601e-05,
      "loss": 0.5956,
      "step": 4640
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.2905120849609375,
      "learning_rate": 4.5894417518490225e-05,
      "loss": 0.6106,
      "step": 4650
    },
    {
      "epoch": 1.13,
      "grad_norm": 0.2116033285856247,
      "learning_rate": 4.5872489552363475e-05,
      "loss": 0.6218,
      "step": 4660
    },
    {
      "epoch": 1.13,
      "grad_norm": 0.22969362139701843,
      "learning_rate": 4.585050845101791e-05,
      "loss": 0.6419,
      "step": 4670
    },
    {
      "epoch": 1.13,
      "grad_norm": 0.23150619864463806,
      "learning_rate": 4.582847427041097e-05,
      "loss": 0.6128,
      "step": 4680
    },
    {
      "epoch": 1.13,
      "grad_norm": 0.23322267830371857,
      "learning_rate": 4.580638706663517e-05,
      "loss": 0.6114,
      "step": 4690
    },
    {
      "epoch": 1.14,
      "grad_norm": 0.23113666474819183,
      "learning_rate": 4.578424689591805e-05,
      "loss": 0.6173,
      "step": 4700
    },
    {
      "epoch": 1.14,
      "grad_norm": 0.2632519602775574,
      "learning_rate": 4.5762053814621975e-05,
      "loss": 0.643,
      "step": 4710
    },
    {
      "epoch": 1.14,
      "grad_norm": 0.19099248945713043,
      "learning_rate": 4.573980787924399e-05,
      "loss": 0.6201,
      "step": 4720
    },
    {
      "epoch": 1.14,
      "grad_norm": 0.23567742109298706,
      "learning_rate": 4.5717509146415705e-05,
      "loss": 0.6163,
      "step": 4730
    },
    {
      "epoch": 1.15,
      "grad_norm": 0.21497173607349396,
      "learning_rate": 4.5695157672903144e-05,
      "loss": 0.6226,
      "step": 4740
    },
    {
      "epoch": 1.15,
      "grad_norm": 0.22423885762691498,
      "learning_rate": 4.567275351560658e-05,
      "loss": 0.6052,
      "step": 4750
    },
    {
      "epoch": 1.15,
      "grad_norm": 0.2486392855644226,
      "learning_rate": 4.5650296731560396e-05,
      "loss": 0.6102,
      "step": 4760
    },
    {
      "epoch": 1.15,
      "grad_norm": 0.19872063398361206,
      "learning_rate": 4.562778737793298e-05,
      "loss": 0.633,
      "step": 4770
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.19403629004955292,
      "learning_rate": 4.560522551202651e-05,
      "loss": 0.6074,
      "step": 4780
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.2101258486509323,
      "learning_rate": 4.558261119127686e-05,
      "loss": 0.6019,
      "step": 4790
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.2554053068161011,
      "learning_rate": 4.555994447325344e-05,
      "loss": 0.6095,
      "step": 4800
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.3209736943244934,
      "learning_rate": 4.5537225415659054e-05,
      "loss": 0.6095,
      "step": 4810
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.1882716864347458,
      "learning_rate": 4.551445407632973e-05,
      "loss": 0.624,
      "step": 4820
    },
    {
      "epoch": 1.17,
      "grad_norm": 0.219781294465065,
      "learning_rate": 4.549163051323461e-05,
      "loss": 0.6212,
      "step": 4830
    },
    {
      "epoch": 1.17,
      "grad_norm": 0.1968742161989212,
      "learning_rate": 4.5468754784475764e-05,
      "loss": 0.6088,
      "step": 4840
    },
    {
      "epoch": 1.17,
      "grad_norm": 0.23859383165836334,
      "learning_rate": 4.5445826948288074e-05,
      "loss": 0.6429,
      "step": 4850
    },
    {
      "epoch": 1.17,
      "grad_norm": 0.21909688413143158,
      "learning_rate": 4.542284706303906e-05,
      "loss": 0.6163,
      "step": 4860
    },
    {
      "epoch": 1.18,
      "grad_norm": 0.23673193156719208,
      "learning_rate": 4.539981518722876e-05,
      "loss": 0.6261,
      "step": 4870
    },
    {
      "epoch": 1.18,
      "grad_norm": 0.23555058240890503,
      "learning_rate": 4.537673137948954e-05,
      "loss": 0.598,
      "step": 4880
    },
    {
      "epoch": 1.18,
      "grad_norm": 0.3052713871002197,
      "learning_rate": 4.5353595698586e-05,
      "loss": 0.6485,
      "step": 4890
    },
    {
      "epoch": 1.18,
      "grad_norm": 0.2587834298610687,
      "learning_rate": 4.533040820341477e-05,
      "loss": 0.6104,
      "step": 4900
    },
    {
      "epoch": 1.19,
      "grad_norm": 0.2710438370704651,
      "learning_rate": 4.53071689530044e-05,
      "loss": 0.6286,
      "step": 4910
    },
    {
      "epoch": 1.19,
      "grad_norm": 0.2709255516529083,
      "learning_rate": 4.528387800651517e-05,
      "loss": 0.6199,
      "step": 4920
    },
    {
      "epoch": 1.19,
      "grad_norm": 0.3711196780204773,
      "learning_rate": 4.5260535423239e-05,
      "loss": 0.6303,
      "step": 4930
    },
    {
      "epoch": 1.19,
      "grad_norm": 0.2359853833913803,
      "learning_rate": 4.523714126259923e-05,
      "loss": 0.6232,
      "step": 4940
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.2610563039779663,
      "learning_rate": 4.5213695584150495e-05,
      "loss": 0.6266,
      "step": 4950
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.23091551661491394,
      "learning_rate": 4.519019844757863e-05,
      "loss": 0.6264,
      "step": 4960
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.20323573052883148,
      "learning_rate": 4.516664991270041e-05,
      "loss": 0.6137,
      "step": 4970
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.2040981501340866,
      "learning_rate": 4.5143050039463476e-05,
      "loss": 0.6433,
      "step": 4980
    },
    {
      "epoch": 1.21,
      "grad_norm": 0.24421969056129456,
      "learning_rate": 4.511939888794617e-05,
      "loss": 0.6221,
      "step": 4990
    },
    {
      "epoch": 1.21,
      "grad_norm": 0.22230570018291473,
      "learning_rate": 4.5095696518357375e-05,
      "loss": 0.6002,
      "step": 5000
    },
    {
      "epoch": 1.21,
      "eval_loss": 0.6445114016532898,
      "eval_runtime": 67.1925,
      "eval_samples_per_second": 29.765,
      "eval_steps_per_second": 0.938,
      "step": 5000
    },
    {
      "epoch": 1.21,
      "grad_norm": 0.2543829381465912,
      "learning_rate": 4.5071942991036354e-05,
      "loss": 0.6598,
      "step": 5010
    },
    {
      "epoch": 1.21,
      "grad_norm": 0.29779505729675293,
      "learning_rate": 4.504813836645258e-05,
      "loss": 0.6196,
      "step": 5020
    },
    {
      "epoch": 1.22,
      "grad_norm": 0.2295481115579605,
      "learning_rate": 4.5024282705205635e-05,
      "loss": 0.6227,
      "step": 5030
    },
    {
      "epoch": 1.22,
      "grad_norm": 0.234407439827919,
      "learning_rate": 4.500037606802502e-05,
      "loss": 0.6363,
      "step": 5040
    },
    {
      "epoch": 1.22,
      "grad_norm": 0.2559088468551636,
      "learning_rate": 4.497641851577e-05,
      "loss": 0.6471,
      "step": 5050
    },
    {
      "epoch": 1.22,
      "grad_norm": 0.21824946999549866,
      "learning_rate": 4.495241010942945e-05,
      "loss": 0.6362,
      "step": 5060
    },
    {
      "epoch": 1.23,
      "grad_norm": 0.2934998571872711,
      "learning_rate": 4.4928350910121714e-05,
      "loss": 0.6285,
      "step": 5070
    },
    {
      "epoch": 1.23,
      "grad_norm": 0.2297659069299698,
      "learning_rate": 4.4904240979094436e-05,
      "loss": 0.6102,
      "step": 5080
    },
    {
      "epoch": 1.23,
      "grad_norm": 0.2708946466445923,
      "learning_rate": 4.488008037772441e-05,
      "loss": 0.6186,
      "step": 5090
    },
    {
      "epoch": 1.23,
      "grad_norm": 0.23141387104988098,
      "learning_rate": 4.485586916751743e-05,
      "loss": 0.6314,
      "step": 5100
    },
    {
      "epoch": 1.23,
      "grad_norm": 0.2494589388370514,
      "learning_rate": 4.483160741010809e-05,
      "loss": 0.6413,
      "step": 5110
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.2377118617296219,
      "learning_rate": 4.480729516725971e-05,
      "loss": 0.617,
      "step": 5120
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.2568589746952057,
      "learning_rate": 4.47829325008641e-05,
      "loss": 0.6143,
      "step": 5130
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.23955343663692474,
      "learning_rate": 4.475851947294145e-05,
      "loss": 0.6257,
      "step": 5140
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.2029058337211609,
      "learning_rate": 4.4734056145640135e-05,
      "loss": 0.6074,
      "step": 5150
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.3426890969276428,
      "learning_rate": 4.470954258123661e-05,
      "loss": 0.6511,
      "step": 5160
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.28450432419776917,
      "learning_rate": 4.4684978842135175e-05,
      "loss": 0.6227,
      "step": 5170
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.2370108813047409,
      "learning_rate": 4.4660364990867895e-05,
      "loss": 0.6198,
      "step": 5180
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.33506277203559875,
      "learning_rate": 4.463570109009441e-05,
      "loss": 0.6112,
      "step": 5190
    },
    {
      "epoch": 1.26,
      "grad_norm": 0.21070849895477295,
      "learning_rate": 4.461098720260173e-05,
      "loss": 0.606,
      "step": 5200
    },
    {
      "epoch": 1.26,
      "grad_norm": 0.3238009214401245,
      "learning_rate": 4.458622339130416e-05,
      "loss": 0.6413,
      "step": 5210
    },
    {
      "epoch": 1.26,
      "grad_norm": 0.23879234492778778,
      "learning_rate": 4.456140971924309e-05,
      "loss": 0.6229,
      "step": 5220
    },
    {
      "epoch": 1.26,
      "grad_norm": 0.24163304269313812,
      "learning_rate": 4.45365462495868e-05,
      "loss": 0.6268,
      "step": 5230
    },
    {
      "epoch": 1.27,
      "grad_norm": 0.22378313541412354,
      "learning_rate": 4.451163304563038e-05,
      "loss": 0.6218,
      "step": 5240
    },
    {
      "epoch": 1.27,
      "grad_norm": 0.21948544681072235,
      "learning_rate": 4.448667017079554e-05,
      "loss": 0.5945,
      "step": 5250
    },
    {
      "epoch": 1.27,
      "grad_norm": 0.2834113836288452,
      "learning_rate": 4.44616576886304e-05,
      "loss": 0.6449,
      "step": 5260
    },
    {
      "epoch": 1.27,
      "grad_norm": 0.2503283619880676,
      "learning_rate": 4.443659566280938e-05,
      "loss": 0.6243,
      "step": 5270
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.35525229573249817,
      "learning_rate": 4.441148415713303e-05,
      "loss": 0.6283,
      "step": 5280
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.2495538890361786,
      "learning_rate": 4.438632323552786e-05,
      "loss": 0.5959,
      "step": 5290
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.24201400578022003,
      "learning_rate": 4.436111296204617e-05,
      "loss": 0.6242,
      "step": 5300
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.24746832251548767,
      "learning_rate": 4.4335853400865906e-05,
      "loss": 0.5989,
      "step": 5310
    },
    {
      "epoch": 1.29,
      "grad_norm": 0.2049688845872879,
      "learning_rate": 4.4310544616290475e-05,
      "loss": 0.6375,
      "step": 5320
    },
    {
      "epoch": 1.29,
      "grad_norm": 0.2205064594745636,
      "learning_rate": 4.42851866727486e-05,
      "loss": 0.6253,
      "step": 5330
    },
    {
      "epoch": 1.29,
      "grad_norm": 0.2159041166305542,
      "learning_rate": 4.425977963479415e-05,
      "loss": 0.6027,
      "step": 5340
    },
    {
      "epoch": 1.29,
      "grad_norm": 0.2615651488304138,
      "learning_rate": 4.423432356710597e-05,
      "loss": 0.6134,
      "step": 5350
    },
    {
      "epoch": 1.3,
      "grad_norm": 0.23875150084495544,
      "learning_rate": 4.420881853448771e-05,
      "loss": 0.5926,
      "step": 5360
    },
    {
      "epoch": 1.3,
      "grad_norm": 0.25981611013412476,
      "learning_rate": 4.418326460186769e-05,
      "loss": 0.623,
      "step": 5370
    },
    {
      "epoch": 1.3,
      "grad_norm": 0.5433134436607361,
      "learning_rate": 4.415766183429871e-05,
      "loss": 0.6143,
      "step": 5380
    },
    {
      "epoch": 1.3,
      "grad_norm": 0.23302482068538666,
      "learning_rate": 4.413201029695788e-05,
      "loss": 0.625,
      "step": 5390
    },
    {
      "epoch": 1.3,
      "grad_norm": 0.2442236691713333,
      "learning_rate": 4.410631005514646e-05,
      "loss": 0.623,
      "step": 5400
    },
    {
      "epoch": 1.31,
      "grad_norm": 0.25277549028396606,
      "learning_rate": 4.4080561174289713e-05,
      "loss": 0.6298,
      "step": 5410
    },
    {
      "epoch": 1.31,
      "grad_norm": 0.26706215739250183,
      "learning_rate": 4.405476371993673e-05,
      "loss": 0.5957,
      "step": 5420
    },
    {
      "epoch": 1.31,
      "grad_norm": 0.19037441909313202,
      "learning_rate": 4.4028917757760216e-05,
      "loss": 0.6228,
      "step": 5430
    },
    {
      "epoch": 1.31,
      "grad_norm": 0.21772147715091705,
      "learning_rate": 4.4003023353556406e-05,
      "loss": 0.6402,
      "step": 5440
    },
    {
      "epoch": 1.32,
      "grad_norm": 0.19529011845588684,
      "learning_rate": 4.3977080573244836e-05,
      "loss": 0.5973,
      "step": 5450
    },
    {
      "epoch": 1.32,
      "grad_norm": 0.22065111994743347,
      "learning_rate": 4.395108948286818e-05,
      "loss": 0.6181,
      "step": 5460
    },
    {
      "epoch": 1.32,
      "grad_norm": 0.3971559405326843,
      "learning_rate": 4.3925050148592124e-05,
      "loss": 0.6217,
      "step": 5470
    },
    {
      "epoch": 1.32,
      "grad_norm": 0.2587713897228241,
      "learning_rate": 4.389896263670514e-05,
      "loss": 0.5931,
      "step": 5480
    },
    {
      "epoch": 1.33,
      "grad_norm": 0.23226861655712128,
      "learning_rate": 4.387282701361837e-05,
      "loss": 0.6207,
      "step": 5490
    },
    {
      "epoch": 1.33,
      "grad_norm": 0.2062036246061325,
      "learning_rate": 4.3846643345865425e-05,
      "loss": 0.6303,
      "step": 5500
    },
    {
      "epoch": 1.33,
      "eval_loss": 0.6429007649421692,
      "eval_runtime": 67.2472,
      "eval_samples_per_second": 29.741,
      "eval_steps_per_second": 0.937,
      "step": 5500
    },
    {
      "epoch": 1.33,
      "grad_norm": 0.2225368171930313,
      "learning_rate": 4.3820411700102216e-05,
      "loss": 0.6123,
      "step": 5510
    },
    {
      "epoch": 1.33,
      "grad_norm": 0.2737579941749573,
      "learning_rate": 4.3794132143106795e-05,
      "loss": 0.6255,
      "step": 5520
    },
    {
      "epoch": 1.34,
      "grad_norm": 0.2531569004058838,
      "learning_rate": 4.376780474177918e-05,
      "loss": 0.5988,
      "step": 5530
    },
    {
      "epoch": 1.34,
      "grad_norm": 0.27578508853912354,
      "learning_rate": 4.37414295631412e-05,
      "loss": 0.6075,
      "step": 5540
    },
    {
      "epoch": 1.34,
      "grad_norm": 0.2573574185371399,
      "learning_rate": 4.37150066743363e-05,
      "loss": 0.6217,
      "step": 5550
    },
    {
      "epoch": 1.34,
      "grad_norm": 0.28215906023979187,
      "learning_rate": 4.3688536142629376e-05,
      "loss": 0.6202,
      "step": 5560
    },
    {
      "epoch": 1.35,
      "grad_norm": 0.23517820239067078,
      "learning_rate": 4.366201803540663e-05,
      "loss": 0.6056,
      "step": 5570
    },
    {
      "epoch": 1.35,
      "grad_norm": 0.26566624641418457,
      "learning_rate": 4.363545242017534e-05,
      "loss": 0.6046,
      "step": 5580
    },
    {
      "epoch": 1.35,
      "grad_norm": 0.20942912995815277,
      "learning_rate": 4.360883936456377e-05,
      "loss": 0.5939,
      "step": 5590
    },
    {
      "epoch": 1.35,
      "grad_norm": 0.24037909507751465,
      "learning_rate": 4.358217893632093e-05,
      "loss": 0.6218,
      "step": 5600
    },
    {
      "epoch": 1.36,
      "grad_norm": 0.22285695374011993,
      "learning_rate": 4.355547120331641e-05,
      "loss": 0.6332,
      "step": 5610
    },
    {
      "epoch": 1.36,
      "grad_norm": 0.2597566545009613,
      "learning_rate": 4.352871623354025e-05,
      "loss": 0.6626,
      "step": 5620
    },
    {
      "epoch": 1.36,
      "grad_norm": 0.37357744574546814,
      "learning_rate": 4.350191409510275e-05,
      "loss": 0.6143,
      "step": 5630
    },
    {
      "epoch": 1.36,
      "grad_norm": 0.24381165206432343,
      "learning_rate": 4.347506485623424e-05,
      "loss": 0.6369,
      "step": 5640
    },
    {
      "epoch": 1.37,
      "grad_norm": 0.27438944578170776,
      "learning_rate": 4.344816858528501e-05,
      "loss": 0.6535,
      "step": 5650
    },
    {
      "epoch": 1.37,
      "grad_norm": 0.20593568682670593,
      "learning_rate": 4.342122535072505e-05,
      "loss": 0.6251,
      "step": 5660
    },
    {
      "epoch": 1.37,
      "grad_norm": 0.25618818402290344,
      "learning_rate": 4.33942352211439e-05,
      "loss": 0.6182,
      "step": 5670
    },
    {
      "epoch": 1.37,
      "grad_norm": 0.20692601799964905,
      "learning_rate": 4.33671982652505e-05,
      "loss": 0.6261,
      "step": 5680
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.2003064602613449,
      "learning_rate": 4.334011455187298e-05,
      "loss": 0.6331,
      "step": 5690
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.22887033224105835,
      "learning_rate": 4.3312984149958516e-05,
      "loss": 0.6292,
      "step": 5700
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.21389944851398468,
      "learning_rate": 4.328580712857313e-05,
      "loss": 0.6094,
      "step": 5710
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.26887232065200806,
      "learning_rate": 4.32585835569015e-05,
      "loss": 0.5942,
      "step": 5720
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.30249106884002686,
      "learning_rate": 4.3231313504246875e-05,
      "loss": 0.6473,
      "step": 5730
    },
    {
      "epoch": 1.39,
      "grad_norm": 0.2598324418067932,
      "learning_rate": 4.320399704003075e-05,
      "loss": 0.6198,
      "step": 5740
    },
    {
      "epoch": 1.39,
      "grad_norm": 0.27700358629226685,
      "learning_rate": 4.317663423379283e-05,
      "loss": 0.6378,
      "step": 5750
    },
    {
      "epoch": 1.39,
      "grad_norm": 0.22245779633522034,
      "learning_rate": 4.3149225155190766e-05,
      "loss": 0.6458,
      "step": 5760
    },
    {
      "epoch": 1.39,
      "grad_norm": 0.27054449915885925,
      "learning_rate": 4.3121769874e-05,
      "loss": 0.6347,
      "step": 5770
    },
    {
      "epoch": 1.4,
      "grad_norm": 0.22811339795589447,
      "learning_rate": 4.309426846011362e-05,
      "loss": 0.6382,
      "step": 5780
    },
    {
      "epoch": 1.4,
      "grad_norm": 0.27327001094818115,
      "learning_rate": 4.306672098354211e-05,
      "loss": 0.6114,
      "step": 5790
    },
    {
      "epoch": 1.4,
      "grad_norm": 0.24687005579471588,
      "learning_rate": 4.3039127514413266e-05,
      "loss": 0.5949,
      "step": 5800
    },
    {
      "epoch": 1.4,
      "grad_norm": 0.39029061794281006,
      "learning_rate": 4.3011488122971945e-05,
      "loss": 0.6281,
      "step": 5810
    },
    {
      "epoch": 1.41,
      "grad_norm": 0.2590855360031128,
      "learning_rate": 4.298380287957989e-05,
      "loss": 0.6068,
      "step": 5820
    },
    {
      "epoch": 1.41,
      "grad_norm": 0.18194353580474854,
      "learning_rate": 4.29560718547156e-05,
      "loss": 0.6114,
      "step": 5830
    },
    {
      "epoch": 1.41,
      "grad_norm": 0.8544417023658752,
      "learning_rate": 4.292829511897409e-05,
      "loss": 0.6336,
      "step": 5840
    },
    {
      "epoch": 1.41,
      "grad_norm": 0.23495988547801971,
      "learning_rate": 4.290047274306678e-05,
      "loss": 0.6185,
      "step": 5850
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.20022150874137878,
      "learning_rate": 4.2872604797821246e-05,
      "loss": 0.6001,
      "step": 5860
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.40812554955482483,
      "learning_rate": 4.284469135418108e-05,
      "loss": 0.6365,
      "step": 5870
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.23241877555847168,
      "learning_rate": 4.28167324832057e-05,
      "loss": 0.6046,
      "step": 5880
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.26784956455230713,
      "learning_rate": 4.2788728256070165e-05,
      "loss": 0.604,
      "step": 5890
    },
    {
      "epoch": 1.43,
      "grad_norm": 0.25370949506759644,
      "learning_rate": 4.2760678744065005e-05,
      "loss": 0.6368,
      "step": 5900
    },
    {
      "epoch": 1.43,
      "grad_norm": 0.2346085160970688,
      "learning_rate": 4.273258401859602e-05,
      "loss": 0.5944,
      "step": 5910
    },
    {
      "epoch": 1.43,
      "grad_norm": 0.32089290022850037,
      "learning_rate": 4.2704444151184117e-05,
      "loss": 0.6124,
      "step": 5920
    },
    {
      "epoch": 1.43,
      "grad_norm": 0.19013577699661255,
      "learning_rate": 4.267625921346513e-05,
      "loss": 0.6167,
      "step": 5930
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.27597373723983765,
      "learning_rate": 4.2648029277189616e-05,
      "loss": 0.6002,
      "step": 5940
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.2857840657234192,
      "learning_rate": 4.261975441422268e-05,
      "loss": 0.5885,
      "step": 5950
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.22939664125442505,
      "learning_rate": 4.259143469654382e-05,
      "loss": 0.6086,
      "step": 5960
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.24333885312080383,
      "learning_rate": 4.25630701962467e-05,
      "loss": 0.6066,
      "step": 5970
    },
    {
      "epoch": 1.45,
      "grad_norm": 0.2737683057785034,
      "learning_rate": 4.253466098553899e-05,
      "loss": 0.6338,
      "step": 5980
    },
    {
      "epoch": 1.45,
      "grad_norm": 0.23578789830207825,
      "learning_rate": 4.2506207136742196e-05,
      "loss": 0.6088,
      "step": 5990
    },
    {
      "epoch": 1.45,
      "grad_norm": 0.3092876076698303,
      "learning_rate": 4.247770872229143e-05,
      "loss": 0.6405,
      "step": 6000
    },
    {
      "epoch": 1.45,
      "eval_loss": 0.6421223282814026,
      "eval_runtime": 67.2062,
      "eval_samples_per_second": 29.759,
      "eval_steps_per_second": 0.937,
      "step": 6000
    },
    {
      "epoch": 1.45,
      "grad_norm": 0.24720308184623718,
      "learning_rate": 4.2449165814735294e-05,
      "loss": 0.6489,
      "step": 6010
    },
    {
      "epoch": 1.45,
      "grad_norm": 0.2425301969051361,
      "learning_rate": 4.242057848673561e-05,
      "loss": 0.6237,
      "step": 6020
    },
    {
      "epoch": 1.46,
      "grad_norm": 0.1967163234949112,
      "learning_rate": 4.239194681106733e-05,
      "loss": 0.6256,
      "step": 6030
    },
    {
      "epoch": 1.46,
      "grad_norm": 0.24658535420894623,
      "learning_rate": 4.2363270860618264e-05,
      "loss": 0.5891,
      "step": 6040
    },
    {
      "epoch": 1.46,
      "grad_norm": 0.22012916207313538,
      "learning_rate": 4.233455070838895e-05,
      "loss": 0.6072,
      "step": 6050
    },
    {
      "epoch": 1.46,
      "grad_norm": 0.32060471177101135,
      "learning_rate": 4.230578642749245e-05,
      "loss": 0.62,
      "step": 6060
    },
    {
      "epoch": 1.47,
      "grad_norm": 0.2524150013923645,
      "learning_rate": 4.2276978091154164e-05,
      "loss": 0.6002,
      "step": 6070
    },
    {
      "epoch": 1.47,
      "grad_norm": 0.22831355035305023,
      "learning_rate": 4.2248125772711635e-05,
      "loss": 0.6245,
      "step": 6080
    },
    {
      "epoch": 1.47,
      "grad_norm": 0.3438965678215027,
      "learning_rate": 4.221922954561439e-05,
      "loss": 0.6127,
      "step": 6090
    },
    {
      "epoch": 1.47,
      "grad_norm": 0.20607294142246246,
      "learning_rate": 4.2190289483423704e-05,
      "loss": 0.6074,
      "step": 6100
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.3653343915939331,
      "learning_rate": 4.216130565981249e-05,
      "loss": 0.6156,
      "step": 6110
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.19855797290802002,
      "learning_rate": 4.213227814856501e-05,
      "loss": 0.6107,
      "step": 6120
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.2605169415473938,
      "learning_rate": 4.210320702357678e-05,
      "loss": 0.6246,
      "step": 6130
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.2300351858139038,
      "learning_rate": 4.207409235885435e-05,
      "loss": 0.6347,
      "step": 6140
    },
    {
      "epoch": 1.49,
      "grad_norm": 0.9489219188690186,
      "learning_rate": 4.2044934228515063e-05,
      "loss": 0.6217,
      "step": 6150
    },
    {
      "epoch": 1.49,
      "grad_norm": 0.2525247037410736,
      "learning_rate": 4.201573270678696e-05,
      "loss": 0.6115,
      "step": 6160
    },
    {
      "epoch": 1.49,
      "grad_norm": 0.2521913945674896,
      "learning_rate": 4.1986487868008525e-05,
      "loss": 0.5965,
      "step": 6170
    },
    {
      "epoch": 1.49,
      "grad_norm": 0.24830523133277893,
      "learning_rate": 4.195719978662851e-05,
      "loss": 0.618,
      "step": 6180
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.23050466179847717,
      "learning_rate": 4.192786853720576e-05,
      "loss": 0.6135,
      "step": 6190
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.21613763272762299,
      "learning_rate": 4.1898494194409e-05,
      "loss": 0.6155,
      "step": 6200
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.22277826070785522,
      "learning_rate": 4.1869076833016676e-05,
      "loss": 0.611,
      "step": 6210
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.2443763017654419,
      "learning_rate": 4.1839616527916736e-05,
      "loss": 0.5888,
      "step": 6220
    },
    {
      "epoch": 1.51,
      "grad_norm": 0.513171374797821,
      "learning_rate": 4.181011335410645e-05,
      "loss": 0.6096,
      "step": 6230
    },
    {
      "epoch": 1.51,
      "grad_norm": 0.21167334914207458,
      "learning_rate": 4.178056738669221e-05,
      "loss": 0.6108,
      "step": 6240
    },
    {
      "epoch": 1.51,
      "grad_norm": 0.22910219430923462,
      "learning_rate": 4.175097870088937e-05,
      "loss": 0.6172,
      "step": 6250
    },
    {
      "epoch": 1.51,
      "grad_norm": 0.2219991236925125,
      "learning_rate": 4.172134737202202e-05,
      "loss": 0.6188,
      "step": 6260
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.266703337430954,
      "learning_rate": 4.16916734755228e-05,
      "loss": 0.6331,
      "step": 6270
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.23374725878238678,
      "learning_rate": 4.166195708693273e-05,
      "loss": 0.6173,
      "step": 6280
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.2778140902519226,
      "learning_rate": 4.163219828190099e-05,
      "loss": 0.6345,
      "step": 6290
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.23954761028289795,
      "learning_rate": 4.160239713618474e-05,
      "loss": 0.6205,
      "step": 6300
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.23126688599586487,
      "learning_rate": 4.1572553725648945e-05,
      "loss": 0.6136,
      "step": 6310
    },
    {
      "epoch": 1.53,
      "grad_norm": 0.209469735622406,
      "learning_rate": 4.1542668126266135e-05,
      "loss": 0.6181,
      "step": 6320
    },
    {
      "epoch": 1.53,
      "grad_norm": 0.24480338394641876,
      "learning_rate": 4.151274041411626e-05,
      "loss": 0.627,
      "step": 6330
    },
    {
      "epoch": 1.53,
      "grad_norm": 0.27108994126319885,
      "learning_rate": 4.148277066538648e-05,
      "loss": 0.6169,
      "step": 6340
    },
    {
      "epoch": 1.53,
      "grad_norm": 0.2842053472995758,
      "learning_rate": 4.145275895637095e-05,
      "loss": 0.6347,
      "step": 6350
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.23858144879341125,
      "learning_rate": 4.142270536347066e-05,
      "loss": 0.6172,
      "step": 6360
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.18353480100631714,
      "learning_rate": 4.139260996319321e-05,
      "loss": 0.5939,
      "step": 6370
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.24537691473960876,
      "learning_rate": 4.136247283215263e-05,
      "loss": 0.6219,
      "step": 6380
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.30072882771492004,
      "learning_rate": 4.1332294047069215e-05,
      "loss": 0.6368,
      "step": 6390
    },
    {
      "epoch": 1.55,
      "grad_norm": 0.2931046485900879,
      "learning_rate": 4.1302073684769256e-05,
      "loss": 0.6108,
      "step": 6400
    },
    {
      "epoch": 1.55,
      "grad_norm": 0.23465721309185028,
      "learning_rate": 4.1271811822184914e-05,
      "loss": 0.6262,
      "step": 6410
    },
    {
      "epoch": 1.55,
      "grad_norm": 0.21646186709403992,
      "learning_rate": 4.124150853635398e-05,
      "loss": 0.6221,
      "step": 6420
    },
    {
      "epoch": 1.55,
      "grad_norm": 0.19293636083602905,
      "learning_rate": 4.1211163904419716e-05,
      "loss": 0.6147,
      "step": 6430
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.2320639044046402,
      "learning_rate": 4.1180778003630624e-05,
      "loss": 0.5902,
      "step": 6440
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.24089358747005463,
      "learning_rate": 4.115035091134027e-05,
      "loss": 0.6307,
      "step": 6450
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.2302493005990982,
      "learning_rate": 4.1119882705007094e-05,
      "loss": 0.6283,
      "step": 6460
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.23400545120239258,
      "learning_rate": 4.108937346219417e-05,
      "loss": 0.612,
      "step": 6470
    },
    {
      "epoch": 1.57,
      "grad_norm": 0.21376444399356842,
      "learning_rate": 4.1058823260569056e-05,
      "loss": 0.6185,
      "step": 6480
    },
    {
      "epoch": 1.57,
      "grad_norm": 0.23962020874023438,
      "learning_rate": 4.1028232177903586e-05,
      "loss": 0.599,
      "step": 6490
    },
    {
      "epoch": 1.57,
      "grad_norm": 0.20366187393665314,
      "learning_rate": 4.099760029207366e-05,
      "loss": 0.6041,
      "step": 6500
    },
    {
      "epoch": 1.57,
      "eval_loss": 0.6386786103248596,
      "eval_runtime": 67.1728,
      "eval_samples_per_second": 29.774,
      "eval_steps_per_second": 0.938,
      "step": 6500
    },
    {
      "epoch": 1.57,
      "grad_norm": 0.19989852607250214,
      "learning_rate": 4.096692768105904e-05,
      "loss": 0.6166,
      "step": 6510
    },
    {
      "epoch": 1.58,
      "grad_norm": 0.2329440414905548,
      "learning_rate": 4.093621442294318e-05,
      "loss": 0.6111,
      "step": 6520
    },
    {
      "epoch": 1.58,
      "grad_norm": 0.3589790165424347,
      "learning_rate": 4.0905460595912995e-05,
      "loss": 0.5906,
      "step": 6530
    },
    {
      "epoch": 1.58,
      "grad_norm": 0.21661555767059326,
      "learning_rate": 4.0874666278258696e-05,
      "loss": 0.6191,
      "step": 6540
    },
    {
      "epoch": 1.58,
      "grad_norm": 0.21175633370876312,
      "learning_rate": 4.0843831548373536e-05,
      "loss": 0.6092,
      "step": 6550
    },
    {
      "epoch": 1.59,
      "grad_norm": 0.2863156199455261,
      "learning_rate": 4.0812956484753686e-05,
      "loss": 0.615,
      "step": 6560
    },
    {
      "epoch": 1.59,
      "grad_norm": 0.2971532344818115,
      "learning_rate": 4.078204116599796e-05,
      "loss": 0.6144,
      "step": 6570
    },
    {
      "epoch": 1.59,
      "grad_norm": 0.3574163019657135,
      "learning_rate": 4.0751085670807684e-05,
      "loss": 0.6297,
      "step": 6580
    },
    {
      "epoch": 1.59,
      "grad_norm": 0.23878632485866547,
      "learning_rate": 4.072009007798643e-05,
      "loss": 0.6164,
      "step": 6590
    },
    {
      "epoch": 1.59,
      "grad_norm": 0.19213823974132538,
      "learning_rate": 4.0689054466439865e-05,
      "loss": 0.6161,
      "step": 6600
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.23007014393806458,
      "learning_rate": 4.0657978915175524e-05,
      "loss": 0.5919,
      "step": 6610
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.6034734845161438,
      "learning_rate": 4.062686350330262e-05,
      "loss": 0.601,
      "step": 6620
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.234625443816185,
      "learning_rate": 4.059570831003184e-05,
      "loss": 0.6298,
      "step": 6630
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.2770217955112457,
      "learning_rate": 4.056451341467514e-05,
      "loss": 0.6066,
      "step": 6640
    },
    {
      "epoch": 1.61,
      "grad_norm": 0.3099740147590637,
      "learning_rate": 4.053327889664556e-05,
      "loss": 0.6296,
      "step": 6650
    },
    {
      "epoch": 1.61,
      "grad_norm": 0.2302497923374176,
      "learning_rate": 4.050200483545697e-05,
      "loss": 0.5892,
      "step": 6660
    },
    {
      "epoch": 1.61,
      "grad_norm": 0.24834463000297546,
      "learning_rate": 4.047069131072396e-05,
      "loss": 0.6186,
      "step": 6670
    },
    {
      "epoch": 1.61,
      "grad_norm": 0.4216982126235962,
      "learning_rate": 4.043933840216154e-05,
      "loss": 0.6059,
      "step": 6680
    },
    {
      "epoch": 1.62,
      "grad_norm": 0.19061365723609924,
      "learning_rate": 4.040794618958499e-05,
      "loss": 0.5777,
      "step": 6690
    },
    {
      "epoch": 1.62,
      "grad_norm": 0.27531304955482483,
      "learning_rate": 4.0376514752909644e-05,
      "loss": 0.6293,
      "step": 6700
    },
    {
      "epoch": 1.62,
      "grad_norm": 0.29000842571258545,
      "learning_rate": 4.034504417215071e-05,
      "loss": 0.6534,
      "step": 6710
    },
    {
      "epoch": 1.62,
      "grad_norm": 0.2087034285068512,
      "learning_rate": 4.0313534527423014e-05,
      "loss": 0.6039,
      "step": 6720
    },
    {
      "epoch": 1.63,
      "grad_norm": 0.23176586627960205,
      "learning_rate": 4.028198589894086e-05,
      "loss": 0.6382,
      "step": 6730
    },
    {
      "epoch": 1.63,
      "grad_norm": 0.23297011852264404,
      "learning_rate": 4.0250398367017754e-05,
      "loss": 0.6058,
      "step": 6740
    },
    {
      "epoch": 1.63,
      "grad_norm": 0.27167224884033203,
      "learning_rate": 4.021877201206628e-05,
      "loss": 0.6133,
      "step": 6750
    },
    {
      "epoch": 1.63,
      "grad_norm": 0.2260424941778183,
      "learning_rate": 4.018710691459783e-05,
      "loss": 0.6564,
      "step": 6760
    },
    {
      "epoch": 1.64,
      "grad_norm": 0.20277553796768188,
      "learning_rate": 4.015540315522242e-05,
      "loss": 0.6115,
      "step": 6770
    },
    {
      "epoch": 1.64,
      "grad_norm": 0.23194430768489838,
      "learning_rate": 4.012366081464851e-05,
      "loss": 0.5972,
      "step": 6780
    },
    {
      "epoch": 1.64,
      "grad_norm": 0.4496138095855713,
      "learning_rate": 4.009187997368275e-05,
      "loss": 0.601,
      "step": 6790
    },
    {
      "epoch": 1.64,
      "grad_norm": 0.2352319061756134,
      "learning_rate": 4.0060060713229806e-05,
      "loss": 0.6126,
      "step": 6800
    },
    {
      "epoch": 1.65,
      "grad_norm": 0.2365293949842453,
      "learning_rate": 4.002820311429218e-05,
      "loss": 0.5925,
      "step": 6810
    },
    {
      "epoch": 1.65,
      "grad_norm": 0.27040597796440125,
      "learning_rate": 3.9996307257969936e-05,
      "loss": 0.6373,
      "step": 6820
    },
    {
      "epoch": 1.65,
      "grad_norm": 0.29712024331092834,
      "learning_rate": 3.996437322546053e-05,
      "loss": 0.6174,
      "step": 6830
    },
    {
      "epoch": 1.65,
      "grad_norm": 0.2615072727203369,
      "learning_rate": 3.993240109805863e-05,
      "loss": 0.6315,
      "step": 6840
    },
    {
      "epoch": 1.66,
      "grad_norm": 0.19362391531467438,
      "learning_rate": 3.990039095715587e-05,
      "loss": 0.6553,
      "step": 6850
    },
    {
      "epoch": 1.66,
      "grad_norm": 0.27068230509757996,
      "learning_rate": 3.986834288424064e-05,
      "loss": 0.5997,
      "step": 6860
    },
    {
      "epoch": 1.66,
      "grad_norm": 0.24307383596897125,
      "learning_rate": 3.9836256960897916e-05,
      "loss": 0.6098,
      "step": 6870
    },
    {
      "epoch": 1.66,
      "grad_norm": 0.2581513226032257,
      "learning_rate": 3.980413326880902e-05,
      "loss": 0.618,
      "step": 6880
    },
    {
      "epoch": 1.67,
      "grad_norm": 0.2150333821773529,
      "learning_rate": 3.9771971889751426e-05,
      "loss": 0.6039,
      "step": 6890
    },
    {
      "epoch": 1.67,
      "grad_norm": 0.28339913487434387,
      "learning_rate": 3.9739772905598525e-05,
      "loss": 0.6122,
      "step": 6900
    },
    {
      "epoch": 1.67,
      "grad_norm": 0.5444418787956238,
      "learning_rate": 3.970753639831949e-05,
      "loss": 0.6242,
      "step": 6910
    },
    {
      "epoch": 1.67,
      "grad_norm": 0.2532576322555542,
      "learning_rate": 3.967526244997896e-05,
      "loss": 0.6069,
      "step": 6920
    },
    {
      "epoch": 1.67,
      "grad_norm": 0.2285206913948059,
      "learning_rate": 3.9642951142736936e-05,
      "loss": 0.6039,
      "step": 6930
    },
    {
      "epoch": 1.68,
      "grad_norm": 0.2928161919116974,
      "learning_rate": 3.961060255884849e-05,
      "loss": 0.6098,
      "step": 6940
    },
    {
      "epoch": 1.68,
      "grad_norm": 0.3057214617729187,
      "learning_rate": 3.957821678066359e-05,
      "loss": 0.6299,
      "step": 6950
    },
    {
      "epoch": 1.68,
      "grad_norm": 0.2918532192707062,
      "learning_rate": 3.9545793890626925e-05,
      "loss": 0.6069,
      "step": 6960
    },
    {
      "epoch": 1.68,
      "grad_norm": 0.20853939652442932,
      "learning_rate": 3.951333397127762e-05,
      "loss": 0.6228,
      "step": 6970
    },
    {
      "epoch": 1.69,
      "grad_norm": 0.2272331267595291,
      "learning_rate": 3.948083710524909e-05,
      "loss": 0.6051,
      "step": 6980
    },
    {
      "epoch": 1.69,
      "grad_norm": 0.22748872637748718,
      "learning_rate": 3.9448303375268794e-05,
      "loss": 0.6259,
      "step": 6990
    },
    {
      "epoch": 1.69,
      "grad_norm": 0.23295633494853973,
      "learning_rate": 3.9415732864158026e-05,
      "loss": 0.5912,
      "step": 7000
    },
    {
      "epoch": 1.69,
      "eval_loss": 0.6369640827178955,
      "eval_runtime": 67.3439,
      "eval_samples_per_second": 29.698,
      "eval_steps_per_second": 0.935,
      "step": 7000
    },
    {
      "epoch": 1.69,
      "grad_norm": 0.25066569447517395,
      "learning_rate": 3.9383125654831747e-05,
      "loss": 0.594,
      "step": 7010
    },
    {
      "epoch": 1.7,
      "grad_norm": 0.1922658085823059,
      "learning_rate": 3.935048183029831e-05,
      "loss": 0.6048,
      "step": 7020
    },
    {
      "epoch": 1.7,
      "grad_norm": 0.23781736195087433,
      "learning_rate": 3.93178014736593e-05,
      "loss": 0.6002,
      "step": 7030
    },
    {
      "epoch": 1.7,
      "grad_norm": 0.25118449330329895,
      "learning_rate": 3.9285084668109276e-05,
      "loss": 0.6013,
      "step": 7040
    },
    {
      "epoch": 1.7,
      "grad_norm": 0.2312859296798706,
      "learning_rate": 3.925233149693561e-05,
      "loss": 0.6118,
      "step": 7050
    },
    {
      "epoch": 1.71,
      "grad_norm": 0.25878220796585083,
      "learning_rate": 3.921954204351826e-05,
      "loss": 0.6091,
      "step": 7060
    },
    {
      "epoch": 1.71,
      "grad_norm": 0.25731998682022095,
      "learning_rate": 3.9186716391329485e-05,
      "loss": 0.6221,
      "step": 7070
    },
    {
      "epoch": 1.71,
      "grad_norm": 0.5314686894416809,
      "learning_rate": 3.9153854623933785e-05,
      "loss": 0.6488,
      "step": 7080
    },
    {
      "epoch": 1.71,
      "grad_norm": 0.2735700309276581,
      "learning_rate": 3.912095682498752e-05,
      "loss": 0.5929,
      "step": 7090
    },
    {
      "epoch": 1.72,
      "grad_norm": 0.24641837179660797,
      "learning_rate": 3.908802307823883e-05,
      "loss": 0.6052,
      "step": 7100
    },
    {
      "epoch": 1.72,
      "grad_norm": 0.2285420298576355,
      "learning_rate": 3.905505346752733e-05,
      "loss": 0.6117,
      "step": 7110
    },
    {
      "epoch": 1.72,
      "grad_norm": 0.33809909224510193,
      "learning_rate": 3.902204807678398e-05,
      "loss": 0.6045,
      "step": 7120
    },
    {
      "epoch": 1.72,
      "grad_norm": 0.2631922662258148,
      "learning_rate": 3.8989006990030776e-05,
      "loss": 0.5957,
      "step": 7130
    },
    {
      "epoch": 1.73,
      "grad_norm": 0.2324107438325882,
      "learning_rate": 3.89559302913806e-05,
      "loss": 0.6154,
      "step": 7140
    },
    {
      "epoch": 1.73,
      "grad_norm": 0.28898167610168457,
      "learning_rate": 3.8922818065037e-05,
      "loss": 0.607,
      "step": 7150
    },
    {
      "epoch": 1.73,
      "grad_norm": 0.26003435254096985,
      "learning_rate": 3.888967039529398e-05,
      "loss": 0.6149,
      "step": 7160
    },
    {
      "epoch": 1.73,
      "grad_norm": 0.203714519739151,
      "learning_rate": 3.885648736653574e-05,
      "loss": 0.6087,
      "step": 7170
    },
    {
      "epoch": 1.74,
      "grad_norm": 0.24271652102470398,
      "learning_rate": 3.8823269063236525e-05,
      "loss": 0.6175,
      "step": 7180
    },
    {
      "epoch": 1.74,
      "grad_norm": 0.32109883427619934,
      "learning_rate": 3.8790015569960346e-05,
      "loss": 0.5913,
      "step": 7190
    },
    {
      "epoch": 1.74,
      "grad_norm": 0.23378537595272064,
      "learning_rate": 3.8756726971360835e-05,
      "loss": 0.6011,
      "step": 7200
    },
    {
      "epoch": 1.74,
      "grad_norm": 0.30031922459602356,
      "learning_rate": 3.872340335218096e-05,
      "loss": 0.6384,
      "step": 7210
    },
    {
      "epoch": 1.74,
      "grad_norm": 0.18759281933307648,
      "learning_rate": 3.869004479725285e-05,
      "loss": 0.6099,
      "step": 7220
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.18216826021671295,
      "learning_rate": 3.865665139149759e-05,
      "loss": 0.6137,
      "step": 7230
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.215232253074646,
      "learning_rate": 3.862322321992495e-05,
      "loss": 0.6235,
      "step": 7240
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.28549501299858093,
      "learning_rate": 3.858976036763323e-05,
      "loss": 0.6003,
      "step": 7250
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.2499440610408783,
      "learning_rate": 3.855626291980901e-05,
      "loss": 0.6031,
      "step": 7260
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.22503146529197693,
      "learning_rate": 3.852273096172694e-05,
      "loss": 0.6188,
      "step": 7270
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.2220824956893921,
      "learning_rate": 3.848916457874952e-05,
      "loss": 0.6354,
      "step": 7280
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.20054738223552704,
      "learning_rate": 3.845556385632688e-05,
      "loss": 0.6124,
      "step": 7290
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.28167736530303955,
      "learning_rate": 3.842192887999659e-05,
      "loss": 0.6066,
      "step": 7300
    },
    {
      "epoch": 1.77,
      "grad_norm": 0.23211318254470825,
      "learning_rate": 3.83882597353834e-05,
      "loss": 0.6366,
      "step": 7310
    },
    {
      "epoch": 1.77,
      "grad_norm": 0.24710510671138763,
      "learning_rate": 3.835455650819905e-05,
      "loss": 0.5974,
      "step": 7320
    },
    {
      "epoch": 1.77,
      "grad_norm": 0.20109450817108154,
      "learning_rate": 3.832081928424205e-05,
      "loss": 0.6194,
      "step": 7330
    },
    {
      "epoch": 1.77,
      "grad_norm": 0.228953555226326,
      "learning_rate": 3.828704814939742e-05,
      "loss": 0.6056,
      "step": 7340
    },
    {
      "epoch": 1.78,
      "grad_norm": 0.24477165937423706,
      "learning_rate": 3.8253243189636554e-05,
      "loss": 0.6425,
      "step": 7350
    },
    {
      "epoch": 1.78,
      "grad_norm": 0.2542911767959595,
      "learning_rate": 3.8219404491016956e-05,
      "loss": 0.6043,
      "step": 7360
    },
    {
      "epoch": 1.78,
      "grad_norm": 0.22980964183807373,
      "learning_rate": 3.8185532139681964e-05,
      "loss": 0.5909,
      "step": 7370
    },
    {
      "epoch": 1.78,
      "grad_norm": 0.2753266394138336,
      "learning_rate": 3.8151626221860636e-05,
      "loss": 0.6307,
      "step": 7380
    },
    {
      "epoch": 1.79,
      "grad_norm": 0.2397356778383255,
      "learning_rate": 3.811768682386747e-05,
      "loss": 0.6254,
      "step": 7390
    },
    {
      "epoch": 1.79,
      "grad_norm": 0.2399650514125824,
      "learning_rate": 3.808371403210219e-05,
      "loss": 0.6151,
      "step": 7400
    },
    {
      "epoch": 1.79,
      "grad_norm": 0.26702845096588135,
      "learning_rate": 3.8049707933049524e-05,
      "loss": 0.652,
      "step": 7410
    },
    {
      "epoch": 1.79,
      "grad_norm": 0.2288055717945099,
      "learning_rate": 3.8015668613279e-05,
      "loss": 0.5837,
      "step": 7420
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.25483304262161255,
      "learning_rate": 3.798159615944471e-05,
      "loss": 0.6001,
      "step": 7430
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.21478036046028137,
      "learning_rate": 3.79474906582851e-05,
      "loss": 0.6437,
      "step": 7440
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.24260424077510834,
      "learning_rate": 3.7913352196622765e-05,
      "loss": 0.5938,
      "step": 7450
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.2208358496427536,
      "learning_rate": 3.7879180861364155e-05,
      "loss": 0.6158,
      "step": 7460
    },
    {
      "epoch": 1.81,
      "grad_norm": 0.19732904434204102,
      "learning_rate": 3.7844976739499454e-05,
      "loss": 0.5967,
      "step": 7470
    },
    {
      "epoch": 1.81,
      "grad_norm": 0.2552759349346161,
      "learning_rate": 3.781073991810229e-05,
      "loss": 0.6201,
      "step": 7480
    },
    {
      "epoch": 1.81,
      "grad_norm": 0.22178280353546143,
      "learning_rate": 3.7776470484329554e-05,
      "loss": 0.5918,
      "step": 7490
    },
    {
      "epoch": 1.81,
      "grad_norm": 0.23287595808506012,
      "learning_rate": 3.7742168525421115e-05,
      "loss": 0.6121,
      "step": 7500
    },
    {
      "epoch": 1.81,
      "eval_loss": 0.6359540820121765,
      "eval_runtime": 67.1833,
      "eval_samples_per_second": 29.769,
      "eval_steps_per_second": 0.938,
      "step": 7500
    },
    {
      "epoch": 1.81,
      "grad_norm": 0.3245614469051361,
      "learning_rate": 3.7707834128699695e-05,
      "loss": 0.6239,
      "step": 7510
    },
    {
      "epoch": 1.82,
      "grad_norm": 0.20530427992343903,
      "learning_rate": 3.767346738157057e-05,
      "loss": 0.5835,
      "step": 7520
    },
    {
      "epoch": 1.82,
      "grad_norm": 0.23378130793571472,
      "learning_rate": 3.763906837152136e-05,
      "loss": 0.6247,
      "step": 7530
    },
    {
      "epoch": 1.82,
      "grad_norm": 0.2955145835876465,
      "learning_rate": 3.760463718612182e-05,
      "loss": 0.603,
      "step": 7540
    },
    {
      "epoch": 1.82,
      "grad_norm": 0.23302963376045227,
      "learning_rate": 3.757017391302363e-05,
      "loss": 0.607,
      "step": 7550
    },
    {
      "epoch": 1.83,
      "grad_norm": 0.2423810064792633,
      "learning_rate": 3.7535678639960145e-05,
      "loss": 0.6153,
      "step": 7560
    },
    {
      "epoch": 1.83,
      "grad_norm": 0.25609883666038513,
      "learning_rate": 3.7501151454746184e-05,
      "loss": 0.6355,
      "step": 7570
    },
    {
      "epoch": 1.83,
      "grad_norm": 0.22590801119804382,
      "learning_rate": 3.746659244527779e-05,
      "loss": 0.5878,
      "step": 7580
    },
    {
      "epoch": 1.83,
      "grad_norm": 0.3146507740020752,
      "learning_rate": 3.7432001699532066e-05,
      "loss": 0.6061,
      "step": 7590
    },
    {
      "epoch": 1.84,
      "grad_norm": 0.22446873784065247,
      "learning_rate": 3.739737930556685e-05,
      "loss": 0.593,
      "step": 7600
    },
    {
      "epoch": 1.84,
      "grad_norm": 0.19154445827007294,
      "learning_rate": 3.736272535152058e-05,
      "loss": 0.6231,
      "step": 7610
    },
    {
      "epoch": 1.84,
      "grad_norm": 0.22120869159698486,
      "learning_rate": 3.732803992561204e-05,
      "loss": 0.6028,
      "step": 7620
    },
    {
      "epoch": 1.84,
      "grad_norm": 0.22038020193576813,
      "learning_rate": 3.72933231161401e-05,
      "loss": 0.5896,
      "step": 7630
    },
    {
      "epoch": 1.85,
      "grad_norm": 0.22278408706188202,
      "learning_rate": 3.725857501148356e-05,
      "loss": 0.6177,
      "step": 7640
    },
    {
      "epoch": 1.85,
      "grad_norm": 0.2636203467845917,
      "learning_rate": 3.722379570010087e-05,
      "loss": 0.6137,
      "step": 7650
    },
    {
      "epoch": 1.85,
      "grad_norm": 0.2235359400510788,
      "learning_rate": 3.718898527052993e-05,
      "loss": 0.6156,
      "step": 7660
    },
    {
      "epoch": 1.85,
      "grad_norm": 0.24960432946681976,
      "learning_rate": 3.715414381138785e-05,
      "loss": 0.6181,
      "step": 7670
    },
    {
      "epoch": 1.86,
      "grad_norm": 0.2168387770652771,
      "learning_rate": 3.7119271411370736e-05,
      "loss": 0.5881,
      "step": 7680
    },
    {
      "epoch": 1.86,
      "grad_norm": 0.2279047667980194,
      "learning_rate": 3.708436815925348e-05,
      "loss": 0.626,
      "step": 7690
    },
    {
      "epoch": 1.86,
      "grad_norm": 0.23511163890361786,
      "learning_rate": 3.704943414388946e-05,
      "loss": 0.6249,
      "step": 7700
    },
    {
      "epoch": 1.86,
      "grad_norm": 0.2570114731788635,
      "learning_rate": 3.7014469454210425e-05,
      "loss": 0.605,
      "step": 7710
    },
    {
      "epoch": 1.87,
      "grad_norm": 0.2473715841770172,
      "learning_rate": 3.697947417922619e-05,
      "loss": 0.6271,
      "step": 7720
    },
    {
      "epoch": 1.87,
      "grad_norm": 0.2647952735424042,
      "learning_rate": 3.694444840802443e-05,
      "loss": 0.6002,
      "step": 7730
    },
    {
      "epoch": 1.87,
      "grad_norm": 0.21034955978393555,
      "learning_rate": 3.690939222977045e-05,
      "loss": 0.6136,
      "step": 7740
    },
    {
      "epoch": 1.87,
      "grad_norm": 0.2274123579263687,
      "learning_rate": 3.6874305733706967e-05,
      "loss": 0.5838,
      "step": 7750
    },
    {
      "epoch": 1.88,
      "grad_norm": 0.22881461679935455,
      "learning_rate": 3.6839189009153886e-05,
      "loss": 0.6043,
      "step": 7760
    },
    {
      "epoch": 1.88,
      "grad_norm": 0.2513927221298218,
      "learning_rate": 3.680404214550806e-05,
      "loss": 0.6119,
      "step": 7770
    },
    {
      "epoch": 1.88,
      "grad_norm": 0.23109400272369385,
      "learning_rate": 3.676886523224306e-05,
      "loss": 0.5832,
      "step": 7780
    },
    {
      "epoch": 1.88,
      "grad_norm": 0.2706112563610077,
      "learning_rate": 3.673365835890896e-05,
      "loss": 0.61,
      "step": 7790
    },
    {
      "epoch": 1.88,
      "grad_norm": 0.2524433732032776,
      "learning_rate": 3.669842161513211e-05,
      "loss": 0.5975,
      "step": 7800
    },
    {
      "epoch": 1.89,
      "grad_norm": 0.20997729897499084,
      "learning_rate": 3.6663155090614895e-05,
      "loss": 0.6122,
      "step": 7810
    },
    {
      "epoch": 1.89,
      "grad_norm": 0.21376433968544006,
      "learning_rate": 3.66278588751355e-05,
      "loss": 0.5988,
      "step": 7820
    },
    {
      "epoch": 1.89,
      "grad_norm": 0.23716038465499878,
      "learning_rate": 3.659253305854772e-05,
      "loss": 0.6063,
      "step": 7830
    },
    {
      "epoch": 1.89,
      "grad_norm": 0.2442622035741806,
      "learning_rate": 3.655717773078068e-05,
      "loss": 0.6223,
      "step": 7840
    },
    {
      "epoch": 1.9,
      "grad_norm": 0.2357431948184967,
      "learning_rate": 3.652179298183866e-05,
      "loss": 0.6066,
      "step": 7850
    },
    {
      "epoch": 1.9,
      "grad_norm": 0.2631849944591522,
      "learning_rate": 3.6486378901800814e-05,
      "loss": 0.5991,
      "step": 7860
    },
    {
      "epoch": 1.9,
      "grad_norm": 0.3156498968601227,
      "learning_rate": 3.645093558082098e-05,
      "loss": 0.6226,
      "step": 7870
    },
    {
      "epoch": 1.9,
      "grad_norm": 0.2296159863471985,
      "learning_rate": 3.641546310912741e-05,
      "loss": 0.603,
      "step": 7880
    },
    {
      "epoch": 1.91,
      "grad_norm": 0.21041952073574066,
      "learning_rate": 3.63799615770226e-05,
      "loss": 0.5871,
      "step": 7890
    },
    {
      "epoch": 1.91,
      "grad_norm": 0.2268192023038864,
      "learning_rate": 3.634443107488301e-05,
      "loss": 0.5855,
      "step": 7900
    },
    {
      "epoch": 1.91,
      "grad_norm": 0.24145358800888062,
      "learning_rate": 3.630887169315883e-05,
      "loss": 0.6153,
      "step": 7910
    },
    {
      "epoch": 1.91,
      "grad_norm": 0.2399817854166031,
      "learning_rate": 3.62732835223738e-05,
      "loss": 0.6173,
      "step": 7920
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.21857813000679016,
      "learning_rate": 3.6237666653124934e-05,
      "loss": 0.6205,
      "step": 7930
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.2975575029850006,
      "learning_rate": 3.62020211760823e-05,
      "loss": 0.6071,
      "step": 7940
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.23792515695095062,
      "learning_rate": 3.6166347181988805e-05,
      "loss": 0.6046,
      "step": 7950
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.26107633113861084,
      "learning_rate": 3.613064476165993e-05,
      "loss": 0.6321,
      "step": 7960
    },
    {
      "epoch": 1.93,
      "grad_norm": 0.25213637948036194,
      "learning_rate": 3.6094914005983546e-05,
      "loss": 0.5989,
      "step": 7970
    },
    {
      "epoch": 1.93,
      "grad_norm": 0.21624846756458282,
      "learning_rate": 3.605915500591964e-05,
      "loss": 0.6212,
      "step": 7980
    },
    {
      "epoch": 1.93,
      "grad_norm": 0.22757196426391602,
      "learning_rate": 3.6023367852500126e-05,
      "loss": 0.6179,
      "step": 7990
    },
    {
      "epoch": 1.93,
      "grad_norm": 0.24339766800403595,
      "learning_rate": 3.5987552636828544e-05,
      "loss": 0.613,
      "step": 8000
    },
    {
      "epoch": 1.93,
      "eval_loss": 0.6343603730201721,
      "eval_runtime": 67.246,
      "eval_samples_per_second": 29.742,
      "eval_steps_per_second": 0.937,
      "step": 8000
    },
    {
      "epoch": 1.94,
      "grad_norm": 0.23893538117408752,
      "learning_rate": 3.5951709450079916e-05,
      "loss": 0.631,
      "step": 8010
    },
    {
      "epoch": 1.94,
      "grad_norm": 0.23567819595336914,
      "learning_rate": 3.5915838383500445e-05,
      "loss": 0.6115,
      "step": 8020
    },
    {
      "epoch": 1.94,
      "grad_norm": 0.3273671865463257,
      "learning_rate": 3.5879939528407317e-05,
      "loss": 0.5974,
      "step": 8030
    },
    {
      "epoch": 1.94,
      "grad_norm": 0.22168773412704468,
      "learning_rate": 3.5844012976188454e-05,
      "loss": 0.5933,
      "step": 8040
    },
    {
      "epoch": 1.95,
      "grad_norm": 0.3699816167354584,
      "learning_rate": 3.580805881830231e-05,
      "loss": 0.6095,
      "step": 8050
    },
    {
      "epoch": 1.95,
      "grad_norm": 0.23126068711280823,
      "learning_rate": 3.577207714627756e-05,
      "loss": 0.6214,
      "step": 8060
    },
    {
      "epoch": 1.95,
      "grad_norm": 0.21293146908283234,
      "learning_rate": 3.5736068051712985e-05,
      "loss": 0.6085,
      "step": 8070
    },
    {
      "epoch": 1.95,
      "grad_norm": 0.22205476462841034,
      "learning_rate": 3.570003162627714e-05,
      "loss": 0.594,
      "step": 8080
    },
    {
      "epoch": 1.96,
      "grad_norm": 0.28807544708251953,
      "learning_rate": 3.5663967961708156e-05,
      "loss": 0.6233,
      "step": 8090
    },
    {
      "epoch": 1.96,
      "grad_norm": 0.2474154680967331,
      "learning_rate": 3.562787714981353e-05,
      "loss": 0.6122,
      "step": 8100
    },
    {
      "epoch": 1.96,
      "grad_norm": 0.3010067641735077,
      "learning_rate": 3.5591759282469836e-05,
      "loss": 0.6027,
      "step": 8110
    },
    {
      "epoch": 1.96,
      "grad_norm": 0.21274247765541077,
      "learning_rate": 3.555561445162255e-05,
      "loss": 0.6373,
      "step": 8120
    },
    {
      "epoch": 1.96,
      "grad_norm": 0.22347737848758698,
      "learning_rate": 3.551944274928578e-05,
      "loss": 0.6108,
      "step": 8130
    },
    {
      "epoch": 1.97,
      "grad_norm": 0.24143600463867188,
      "learning_rate": 3.548324426754204e-05,
      "loss": 0.6406,
      "step": 8140
    },
    {
      "epoch": 1.97,
      "grad_norm": 0.2388855516910553,
      "learning_rate": 3.544701909854201e-05,
      "loss": 0.6053,
      "step": 8150
    },
    {
      "epoch": 1.97,
      "grad_norm": 0.22766222059726715,
      "learning_rate": 3.541076733450433e-05,
      "loss": 0.5907,
      "step": 8160
    },
    {
      "epoch": 1.97,
      "grad_norm": 0.26293396949768066,
      "learning_rate": 3.537448906771532e-05,
      "loss": 0.6113,
      "step": 8170
    },
    {
      "epoch": 1.98,
      "grad_norm": 0.3174780607223511,
      "learning_rate": 3.5338184390528765e-05,
      "loss": 0.6083,
      "step": 8180
    },
    {
      "epoch": 1.98,
      "grad_norm": 0.2787187099456787,
      "learning_rate": 3.530185339536571e-05,
      "loss": 0.6206,
      "step": 8190
    },
    {
      "epoch": 1.98,
      "grad_norm": 0.25283700227737427,
      "learning_rate": 3.526549617471417e-05,
      "loss": 0.6078,
      "step": 8200
    },
    {
      "epoch": 1.98,
      "grad_norm": 0.2156975120306015,
      "learning_rate": 3.522911282112894e-05,
      "loss": 0.5923,
      "step": 8210
    },
    {
      "epoch": 1.99,
      "grad_norm": 0.253841757774353,
      "learning_rate": 3.5192703427231344e-05,
      "loss": 0.6046,
      "step": 8220
    },
    {
      "epoch": 1.99,
      "grad_norm": 0.2438536435365677,
      "learning_rate": 3.5156268085708974e-05,
      "loss": 0.6121,
      "step": 8230
    },
    {
      "epoch": 1.99,
      "grad_norm": 0.2552967667579651,
      "learning_rate": 3.511980688931551e-05,
      "loss": 0.6314,
      "step": 8240
    },
    {
      "epoch": 1.99,
      "grad_norm": 0.4905732572078705,
      "learning_rate": 3.5083319930870414e-05,
      "loss": 0.5972,
      "step": 8250
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.2145986407995224,
      "learning_rate": 3.504680730325876e-05,
      "loss": 0.6234,
      "step": 8260
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.24331778287887573,
      "learning_rate": 3.501026909943095e-05,
      "loss": 0.6076,
      "step": 8270
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.2849058210849762,
      "learning_rate": 3.4973705412402514e-05,
      "loss": 0.6067,
      "step": 8280
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.2842152714729309,
      "learning_rate": 3.4937116335253836e-05,
      "loss": 0.6028,
      "step": 8290
    },
    {
      "epoch": 2.01,
      "grad_norm": 0.24628552794456482,
      "learning_rate": 3.490050196112995e-05,
      "loss": 0.6119,
      "step": 8300
    },
    {
      "epoch": 2.01,
      "grad_norm": 0.237552210688591,
      "learning_rate": 3.486386238324029e-05,
      "loss": 0.5985,
      "step": 8310
    },
    {
      "epoch": 2.01,
      "grad_norm": 0.2442472279071808,
      "learning_rate": 3.482719769485841e-05,
      "loss": 0.6007,
      "step": 8320
    },
    {
      "epoch": 2.01,
      "grad_norm": 0.29659032821655273,
      "learning_rate": 3.479050798932186e-05,
      "loss": 0.6082,
      "step": 8330
    },
    {
      "epoch": 2.02,
      "grad_norm": 0.2106475681066513,
      "learning_rate": 3.475379336003183e-05,
      "loss": 0.6247,
      "step": 8340
    },
    {
      "epoch": 2.02,
      "grad_norm": 0.23286166787147522,
      "learning_rate": 3.471705390045295e-05,
      "loss": 0.5975,
      "step": 8350
    },
    {
      "epoch": 2.02,
      "grad_norm": 0.2842426002025604,
      "learning_rate": 3.468028970411308e-05,
      "loss": 0.6206,
      "step": 8360
    },
    {
      "epoch": 2.02,
      "grad_norm": 0.2059846669435501,
      "learning_rate": 3.4643500864603075e-05,
      "loss": 0.6074,
      "step": 8370
    },
    {
      "epoch": 2.03,
      "grad_norm": 0.24341388046741486,
      "learning_rate": 3.4606687475576473e-05,
      "loss": 0.6219,
      "step": 8380
    },
    {
      "epoch": 2.03,
      "grad_norm": 0.25778728723526,
      "learning_rate": 3.456984963074934e-05,
      "loss": 0.6136,
      "step": 8390
    },
    {
      "epoch": 2.03,
      "grad_norm": 0.2483309656381607,
      "learning_rate": 3.453298742389999e-05,
      "loss": 0.6033,
      "step": 8400
    },
    {
      "epoch": 2.03,
      "grad_norm": 0.26595956087112427,
      "learning_rate": 3.4496100948868774e-05,
      "loss": 0.6104,
      "step": 8410
    },
    {
      "epoch": 2.03,
      "grad_norm": 0.3210870623588562,
      "learning_rate": 3.445919029955779e-05,
      "loss": 0.5867,
      "step": 8420
    },
    {
      "epoch": 2.04,
      "grad_norm": 0.2549401819705963,
      "learning_rate": 3.442225556993071e-05,
      "loss": 0.6011,
      "step": 8430
    },
    {
      "epoch": 2.04,
      "grad_norm": 0.2380758821964264,
      "learning_rate": 3.438529685401248e-05,
      "loss": 0.613,
      "step": 8440
    },
    {
      "epoch": 2.04,
      "grad_norm": 0.1969003528356552,
      "learning_rate": 3.4348314245889114e-05,
      "loss": 0.5812,
      "step": 8450
    },
    {
      "epoch": 2.04,
      "grad_norm": 0.38220301270484924,
      "learning_rate": 3.431130783970747e-05,
      "loss": 0.6103,
      "step": 8460
    },
    {
      "epoch": 2.05,
      "grad_norm": 0.3186286687850952,
      "learning_rate": 3.427427772967496e-05,
      "loss": 0.5989,
      "step": 8470
    },
    {
      "epoch": 2.05,
      "grad_norm": 0.2749629020690918,
      "learning_rate": 3.423722401005936e-05,
      "loss": 0.6073,
      "step": 8480
    },
    {
      "epoch": 2.05,
      "grad_norm": 0.2259276658296585,
      "learning_rate": 3.420014677518854e-05,
      "loss": 0.627,
      "step": 8490
    },
    {
      "epoch": 2.05,
      "grad_norm": 0.22658781707286835,
      "learning_rate": 3.416304611945022e-05,
      "loss": 0.6126,
      "step": 8500
    },
    {
      "epoch": 2.05,
      "eval_loss": 0.6338043212890625,
      "eval_runtime": 67.2252,
      "eval_samples_per_second": 29.751,
      "eval_steps_per_second": 0.937,
      "step": 8500
    },
    {
      "epoch": 2.06,
      "grad_norm": 0.24936778843402863,
      "learning_rate": 3.4125922137291774e-05,
      "loss": 0.6181,
      "step": 8510
    },
    {
      "epoch": 2.06,
      "grad_norm": 0.25658896565437317,
      "learning_rate": 3.4088774923219936e-05,
      "loss": 0.572,
      "step": 8520
    },
    {
      "epoch": 2.06,
      "grad_norm": 0.2544236481189728,
      "learning_rate": 3.4051604571800584e-05,
      "loss": 0.5965,
      "step": 8530
    },
    {
      "epoch": 2.06,
      "grad_norm": 0.24977684020996094,
      "learning_rate": 3.4014411177658494e-05,
      "loss": 0.5839,
      "step": 8540
    },
    {
      "epoch": 2.07,
      "grad_norm": 0.29316794872283936,
      "learning_rate": 3.397719483547712e-05,
      "loss": 0.6268,
      "step": 8550
    },
    {
      "epoch": 2.07,
      "grad_norm": 0.31263527274131775,
      "learning_rate": 3.393995563999832e-05,
      "loss": 0.601,
      "step": 8560
    },
    {
      "epoch": 2.07,
      "grad_norm": 0.2534136474132538,
      "learning_rate": 3.390269368602212e-05,
      "loss": 0.6274,
      "step": 8570
    },
    {
      "epoch": 2.07,
      "grad_norm": 0.20804938673973083,
      "learning_rate": 3.38654090684065e-05,
      "loss": 0.6127,
      "step": 8580
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.3475063443183899,
      "learning_rate": 3.382810188206712e-05,
      "loss": 0.602,
      "step": 8590
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.21181778609752655,
      "learning_rate": 3.379077222197713e-05,
      "loss": 0.6136,
      "step": 8600
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.3004330098628998,
      "learning_rate": 3.3753420183166845e-05,
      "loss": 0.5797,
      "step": 8610
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.17146137356758118,
      "learning_rate": 3.3716045860723565e-05,
      "loss": 0.581,
      "step": 8620
    },
    {
      "epoch": 2.09,
      "grad_norm": 0.23555147647857666,
      "learning_rate": 3.367864934979133e-05,
      "loss": 0.6013,
      "step": 8630
    },
    {
      "epoch": 2.09,
      "grad_norm": 0.21328724920749664,
      "learning_rate": 3.364123074557066e-05,
      "loss": 0.6053,
      "step": 8640
    },
    {
      "epoch": 2.09,
      "grad_norm": 0.21844583749771118,
      "learning_rate": 3.36037901433183e-05,
      "loss": 0.6241,
      "step": 8650
    },
    {
      "epoch": 2.09,
      "grad_norm": 0.2694018483161926,
      "learning_rate": 3.356632763834702e-05,
      "loss": 0.5905,
      "step": 8660
    },
    {
      "epoch": 2.1,
      "grad_norm": 0.30904027819633484,
      "learning_rate": 3.3528843326025334e-05,
      "loss": 0.6002,
      "step": 8670
    },
    {
      "epoch": 2.1,
      "grad_norm": 0.35731956362724304,
      "learning_rate": 3.349133730177729e-05,
      "loss": 0.6118,
      "step": 8680
    },
    {
      "epoch": 2.1,
      "grad_norm": 0.33931079506874084,
      "learning_rate": 3.345380966108218e-05,
      "loss": 0.5879,
      "step": 8690
    },
    {
      "epoch": 2.1,
      "grad_norm": 0.21542824804782867,
      "learning_rate": 3.3416260499474334e-05,
      "loss": 0.5831,
      "step": 8700
    },
    {
      "epoch": 2.1,
      "grad_norm": 0.27431559562683105,
      "learning_rate": 3.3378689912542885e-05,
      "loss": 0.597,
      "step": 8710
    },
    {
      "epoch": 2.11,
      "grad_norm": 0.37733927369117737,
      "learning_rate": 3.3341097995931483e-05,
      "loss": 0.6167,
      "step": 8720
    },
    {
      "epoch": 2.11,
      "grad_norm": 0.23641188442707062,
      "learning_rate": 3.3303484845338095e-05,
      "loss": 0.5875,
      "step": 8730
    },
    {
      "epoch": 2.11,
      "grad_norm": 0.19541165232658386,
      "learning_rate": 3.326585055651475e-05,
      "loss": 0.583,
      "step": 8740
    },
    {
      "epoch": 2.11,
      "grad_norm": 0.22675809264183044,
      "learning_rate": 3.3228195225267255e-05,
      "loss": 0.5963,
      "step": 8750
    },
    {
      "epoch": 2.12,
      "grad_norm": 0.2038368433713913,
      "learning_rate": 3.319051894745503e-05,
      "loss": 0.5973,
      "step": 8760
    },
    {
      "epoch": 2.12,
      "grad_norm": 0.21011051535606384,
      "learning_rate": 3.3152821818990786e-05,
      "loss": 0.5982,
      "step": 8770
    },
    {
      "epoch": 2.12,
      "grad_norm": 0.26622429490089417,
      "learning_rate": 3.3115103935840326e-05,
      "loss": 0.6,
      "step": 8780
    },
    {
      "epoch": 2.12,
      "grad_norm": 0.25305575132369995,
      "learning_rate": 3.307736539402227e-05,
      "loss": 0.6024,
      "step": 8790
    },
    {
      "epoch": 2.13,
      "grad_norm": 0.2928634583950043,
      "learning_rate": 3.303960628960788e-05,
      "loss": 0.585,
      "step": 8800
    },
    {
      "epoch": 2.13,
      "grad_norm": 0.2512208819389343,
      "learning_rate": 3.3001826718720694e-05,
      "loss": 0.6148,
      "step": 8810
    },
    {
      "epoch": 2.13,
      "grad_norm": 0.22255320847034454,
      "learning_rate": 3.29640267775364e-05,
      "loss": 0.5983,
      "step": 8820
    },
    {
      "epoch": 2.13,
      "grad_norm": 0.30739256739616394,
      "learning_rate": 3.292620656228253e-05,
      "loss": 0.5949,
      "step": 8830
    },
    {
      "epoch": 2.14,
      "grad_norm": 0.25877636671066284,
      "learning_rate": 3.288836616923823e-05,
      "loss": 0.6027,
      "step": 8840
    },
    {
      "epoch": 2.14,
      "grad_norm": 0.2691468298435211,
      "learning_rate": 3.2850505694734005e-05,
      "loss": 0.5976,
      "step": 8850
    },
    {
      "epoch": 2.14,
      "grad_norm": 0.22831767797470093,
      "learning_rate": 3.281262523515149e-05,
      "loss": 0.6169,
      "step": 8860
    },
    {
      "epoch": 2.14,
      "grad_norm": 0.2406102865934372,
      "learning_rate": 3.277472488692319e-05,
      "loss": 0.5938,
      "step": 8870
    },
    {
      "epoch": 2.15,
      "grad_norm": 0.24443283677101135,
      "learning_rate": 3.273680474653224e-05,
      "loss": 0.6063,
      "step": 8880
    },
    {
      "epoch": 2.15,
      "grad_norm": 0.25821536779403687,
      "learning_rate": 3.269886491051217e-05,
      "loss": 0.6056,
      "step": 8890
    },
    {
      "epoch": 2.15,
      "grad_norm": 0.2244919389486313,
      "learning_rate": 3.266090547544664e-05,
      "loss": 0.5799,
      "step": 8900
    },
    {
      "epoch": 2.15,
      "grad_norm": 0.2633070647716522,
      "learning_rate": 3.2622926537969205e-05,
      "loss": 0.6082,
      "step": 8910
    },
    {
      "epoch": 2.16,
      "grad_norm": 0.2933361530303955,
      "learning_rate": 3.258492819476308e-05,
      "loss": 0.6109,
      "step": 8920
    },
    {
      "epoch": 2.16,
      "grad_norm": 0.23176881670951843,
      "learning_rate": 3.2546910542560855e-05,
      "loss": 0.5844,
      "step": 8930
    },
    {
      "epoch": 2.16,
      "grad_norm": 0.2534427046775818,
      "learning_rate": 3.250887367814429e-05,
      "loss": 0.6032,
      "step": 8940
    },
    {
      "epoch": 2.16,
      "grad_norm": 0.23787447810173035,
      "learning_rate": 3.247081769834406e-05,
      "loss": 0.5859,
      "step": 8950
    },
    {
      "epoch": 2.17,
      "grad_norm": 0.23194755613803864,
      "learning_rate": 3.243274270003949e-05,
      "loss": 0.6103,
      "step": 8960
    },
    {
      "epoch": 2.17,
      "grad_norm": 0.24576812982559204,
      "learning_rate": 3.239464878015833e-05,
      "loss": 0.5964,
      "step": 8970
    },
    {
      "epoch": 2.17,
      "grad_norm": 0.24313139915466309,
      "learning_rate": 3.2356536035676485e-05,
      "loss": 0.6119,
      "step": 8980
    },
    {
      "epoch": 2.17,
      "grad_norm": 0.2334582507610321,
      "learning_rate": 3.231840456361781e-05,
      "loss": 0.5947,
      "step": 8990
    },
    {
      "epoch": 2.17,
      "grad_norm": 0.31086283922195435,
      "learning_rate": 3.2280254461053813e-05,
      "loss": 0.5932,
      "step": 9000
    },
    {
      "epoch": 2.17,
      "eval_loss": 0.6343629956245422,
      "eval_runtime": 67.3178,
      "eval_samples_per_second": 29.71,
      "eval_steps_per_second": 0.936,
      "step": 9000
    },
    {
      "epoch": 2.18,
      "grad_norm": 0.3237190842628479,
      "learning_rate": 3.224208582510342e-05,
      "loss": 0.5944,
      "step": 9010
    },
    {
      "epoch": 2.18,
      "grad_norm": 0.22707690298557281,
      "learning_rate": 3.220389875293276e-05,
      "loss": 0.6017,
      "step": 9020
    },
    {
      "epoch": 2.18,
      "grad_norm": 0.24644169211387634,
      "learning_rate": 3.216569334175488e-05,
      "loss": 0.6058,
      "step": 9030
    },
    {
      "epoch": 2.18,
      "grad_norm": 0.2204134464263916,
      "learning_rate": 3.212746968882952e-05,
      "loss": 0.5961,
      "step": 9040
    },
    {
      "epoch": 2.19,
      "grad_norm": 0.280127614736557,
      "learning_rate": 3.208922789146287e-05,
      "loss": 0.5885,
      "step": 9050
    },
    {
      "epoch": 2.19,
      "grad_norm": 0.24572765827178955,
      "learning_rate": 3.205096804700729e-05,
      "loss": 0.6038,
      "step": 9060
    },
    {
      "epoch": 2.19,
      "grad_norm": 0.23512475192546844,
      "learning_rate": 3.2012690252861085e-05,
      "loss": 0.6248,
      "step": 9070
    },
    {
      "epoch": 2.19,
      "grad_norm": 0.2967652678489685,
      "learning_rate": 3.197439460646826e-05,
      "loss": 0.5897,
      "step": 9080
    },
    {
      "epoch": 2.2,
      "grad_norm": 0.8382120728492737,
      "learning_rate": 3.193608120531826e-05,
      "loss": 0.6157,
      "step": 9090
    },
    {
      "epoch": 2.2,
      "grad_norm": 0.26500606536865234,
      "learning_rate": 3.189775014694575e-05,
      "loss": 0.5879,
      "step": 9100
    },
    {
      "epoch": 2.2,
      "grad_norm": 0.22318434715270996,
      "learning_rate": 3.1859401528930325e-05,
      "loss": 0.6026,
      "step": 9110
    },
    {
      "epoch": 2.2,
      "grad_norm": 0.25416818261146545,
      "learning_rate": 3.182103544889628e-05,
      "loss": 0.5955,
      "step": 9120
    },
    {
      "epoch": 2.21,
      "grad_norm": 0.2655612826347351,
      "learning_rate": 3.1782652004512367e-05,
      "loss": 0.5644,
      "step": 9130
    },
    {
      "epoch": 2.21,
      "grad_norm": 0.25753074884414673,
      "learning_rate": 3.174425129349156e-05,
      "loss": 0.5813,
      "step": 9140
    },
    {
      "epoch": 2.21,
      "grad_norm": 0.4179852306842804,
      "learning_rate": 3.1705833413590766e-05,
      "loss": 0.5831,
      "step": 9150
    },
    {
      "epoch": 2.21,
      "grad_norm": 0.26823192834854126,
      "learning_rate": 3.16673984626106e-05,
      "loss": 0.609,
      "step": 9160
    },
    {
      "epoch": 2.22,
      "grad_norm": 0.23745331168174744,
      "learning_rate": 3.162894653839515e-05,
      "loss": 0.5846,
      "step": 9170
    },
    {
      "epoch": 2.22,
      "grad_norm": 0.23219923675060272,
      "learning_rate": 3.1590477738831704e-05,
      "loss": 0.6069,
      "step": 9180
    },
    {
      "epoch": 2.22,
      "grad_norm": 0.22652706503868103,
      "learning_rate": 3.155199216185051e-05,
      "loss": 0.5775,
      "step": 9190
    },
    {
      "epoch": 2.22,
      "grad_norm": 0.22930103540420532,
      "learning_rate": 3.151348990542452e-05,
      "loss": 0.5936,
      "step": 9200
    },
    {
      "epoch": 2.23,
      "grad_norm": 0.19906581938266754,
      "learning_rate": 3.1474971067569146e-05,
      "loss": 0.6061,
      "step": 9210
    },
    {
      "epoch": 2.23,
      "grad_norm": 0.24898895621299744,
      "learning_rate": 3.1436435746342046e-05,
      "loss": 0.5913,
      "step": 9220
    },
    {
      "epoch": 2.23,
      "grad_norm": 0.30313482880592346,
      "learning_rate": 3.139788403984279e-05,
      "loss": 0.5724,
      "step": 9230
    },
    {
      "epoch": 2.23,
      "grad_norm": 0.24632027745246887,
      "learning_rate": 3.135931604621271e-05,
      "loss": 0.5828,
      "step": 9240
    },
    {
      "epoch": 2.24,
      "grad_norm": 0.23896867036819458,
      "learning_rate": 3.1320731863634533e-05,
      "loss": 0.6245,
      "step": 9250
    },
    {
      "epoch": 2.24,
      "grad_norm": 0.20651116967201233,
      "learning_rate": 3.128213159033226e-05,
      "loss": 0.5987,
      "step": 9260
    },
    {
      "epoch": 2.24,
      "grad_norm": 0.24778035283088684,
      "learning_rate": 3.124351532457084e-05,
      "loss": 0.5993,
      "step": 9270
    },
    {
      "epoch": 2.24,
      "grad_norm": 0.3181811273097992,
      "learning_rate": 3.120488316465592e-05,
      "loss": 0.5981,
      "step": 9280
    },
    {
      "epoch": 2.25,
      "grad_norm": 0.33533188700675964,
      "learning_rate": 3.116623520893361e-05,
      "loss": 0.6235,
      "step": 9290
    },
    {
      "epoch": 2.25,
      "grad_norm": 0.3419838845729828,
      "learning_rate": 3.1127571555790234e-05,
      "loss": 0.6245,
      "step": 9300
    },
    {
      "epoch": 2.25,
      "grad_norm": 0.34298205375671387,
      "learning_rate": 3.10888923036521e-05,
      "loss": 0.6195,
      "step": 9310
    },
    {
      "epoch": 2.25,
      "grad_norm": 0.27872177958488464,
      "learning_rate": 3.105019755098519e-05,
      "loss": 0.5843,
      "step": 9320
    },
    {
      "epoch": 2.25,
      "grad_norm": 0.35872554779052734,
      "learning_rate": 3.1011487396294966e-05,
      "loss": 0.6106,
      "step": 9330
    },
    {
      "epoch": 2.26,
      "grad_norm": 0.29521918296813965,
      "learning_rate": 3.0972761938126085e-05,
      "loss": 0.5885,
      "step": 9340
    },
    {
      "epoch": 2.26,
      "grad_norm": 0.20505386590957642,
      "learning_rate": 3.0934021275062196e-05,
      "loss": 0.5906,
      "step": 9350
    },
    {
      "epoch": 2.26,
      "grad_norm": 0.2266906499862671,
      "learning_rate": 3.089526550572561e-05,
      "loss": 0.6021,
      "step": 9360
    },
    {
      "epoch": 2.26,
      "grad_norm": 0.3935636281967163,
      "learning_rate": 3.085649472877713e-05,
      "loss": 0.6167,
      "step": 9370
    },
    {
      "epoch": 2.27,
      "grad_norm": 0.25832274556159973,
      "learning_rate": 3.0817709042915744e-05,
      "loss": 0.6253,
      "step": 9380
    },
    {
      "epoch": 2.27,
      "grad_norm": 0.3040432333946228,
      "learning_rate": 3.077890854687839e-05,
      "loss": 0.6024,
      "step": 9390
    },
    {
      "epoch": 2.27,
      "grad_norm": 0.33817920088768005,
      "learning_rate": 3.074009333943974e-05,
      "loss": 0.5923,
      "step": 9400
    },
    {
      "epoch": 2.27,
      "grad_norm": 0.31015434861183167,
      "learning_rate": 3.070126351941188e-05,
      "loss": 0.5873,
      "step": 9410
    },
    {
      "epoch": 2.28,
      "grad_norm": 0.3499102294445038,
      "learning_rate": 3.0662419185644115e-05,
      "loss": 0.6071,
      "step": 9420
    },
    {
      "epoch": 2.28,
      "grad_norm": 0.42582130432128906,
      "learning_rate": 3.062356043702268e-05,
      "loss": 0.5789,
      "step": 9430
    },
    {
      "epoch": 2.28,
      "grad_norm": 0.22721824049949646,
      "learning_rate": 3.058468737247054e-05,
      "loss": 0.5973,
      "step": 9440
    },
    {
      "epoch": 2.28,
      "grad_norm": 0.28947141766548157,
      "learning_rate": 3.054580009094706e-05,
      "loss": 0.6056,
      "step": 9450
    },
    {
      "epoch": 2.29,
      "grad_norm": 0.2677144408226013,
      "learning_rate": 3.0506898691447844e-05,
      "loss": 0.5715,
      "step": 9460
    },
    {
      "epoch": 2.29,
      "grad_norm": 0.2382638156414032,
      "learning_rate": 3.0467983273004398e-05,
      "loss": 0.5976,
      "step": 9470
    },
    {
      "epoch": 2.29,
      "grad_norm": 0.29905447363853455,
      "learning_rate": 3.0429053934683928e-05,
      "loss": 0.6101,
      "step": 9480
    },
    {
      "epoch": 2.29,
      "grad_norm": 0.27092835307121277,
      "learning_rate": 3.0390110775589086e-05,
      "loss": 0.5924,
      "step": 9490
    },
    {
      "epoch": 2.3,
      "grad_norm": 0.2604258954524994,
      "learning_rate": 3.0351153894857698e-05,
      "loss": 0.5927,
      "step": 9500
    },
    {
      "epoch": 2.3,
      "eval_loss": 0.633247435092926,
      "eval_runtime": 67.1714,
      "eval_samples_per_second": 29.775,
      "eval_steps_per_second": 0.938,
      "step": 9500
    },
    {
      "epoch": 2.3,
      "grad_norm": 0.4717111885547638,
      "learning_rate": 3.0312183391662523e-05,
      "loss": 0.6141,
      "step": 9510
    },
    {
      "epoch": 2.3,
      "grad_norm": 0.29630929231643677,
      "learning_rate": 3.0273199365210992e-05,
      "loss": 0.6167,
      "step": 9520
    },
    {
      "epoch": 2.3,
      "grad_norm": 0.2589368224143982,
      "learning_rate": 3.023420191474498e-05,
      "loss": 0.5991,
      "step": 9530
    },
    {
      "epoch": 2.31,
      "grad_norm": 0.28849098086357117,
      "learning_rate": 3.019519113954053e-05,
      "loss": 0.6362,
      "step": 9540
    },
    {
      "epoch": 2.31,
      "grad_norm": 0.2815060317516327,
      "learning_rate": 3.0156167138907583e-05,
      "loss": 0.5963,
      "step": 9550
    },
    {
      "epoch": 2.31,
      "grad_norm": 0.273151695728302,
      "learning_rate": 3.0117130012189775e-05,
      "loss": 0.5971,
      "step": 9560
    },
    {
      "epoch": 2.31,
      "grad_norm": 0.4395759403705597,
      "learning_rate": 3.0078079858764145e-05,
      "loss": 0.5809,
      "step": 9570
    },
    {
      "epoch": 2.32,
      "grad_norm": 0.2783837616443634,
      "learning_rate": 3.0039016778040906e-05,
      "loss": 0.6155,
      "step": 9580
    },
    {
      "epoch": 2.32,
      "grad_norm": 0.26403704285621643,
      "learning_rate": 2.9999940869463156e-05,
      "loss": 0.5951,
      "step": 9590
    },
    {
      "epoch": 2.32,
      "grad_norm": 0.2775236666202545,
      "learning_rate": 2.9960852232506677e-05,
      "loss": 0.5948,
      "step": 9600
    },
    {
      "epoch": 2.32,
      "grad_norm": 0.2422449290752411,
      "learning_rate": 2.992175096667964e-05,
      "loss": 0.6123,
      "step": 9610
    },
    {
      "epoch": 2.32,
      "grad_norm": 0.3329131007194519,
      "learning_rate": 2.988263717152236e-05,
      "loss": 0.621,
      "step": 9620
    },
    {
      "epoch": 2.33,
      "grad_norm": 0.2817908227443695,
      "learning_rate": 2.9843510946607058e-05,
      "loss": 0.5881,
      "step": 9630
    },
    {
      "epoch": 2.33,
      "grad_norm": 0.35110849142074585,
      "learning_rate": 2.9804372391537587e-05,
      "loss": 0.6114,
      "step": 9640
    },
    {
      "epoch": 2.33,
      "grad_norm": 0.2615245580673218,
      "learning_rate": 2.9765221605949212e-05,
      "loss": 0.5923,
      "step": 9650
    },
    {
      "epoch": 2.33,
      "grad_norm": 0.2712041139602661,
      "learning_rate": 2.97260586895083e-05,
      "loss": 0.6016,
      "step": 9660
    },
    {
      "epoch": 2.34,
      "grad_norm": 0.24361445009708405,
      "learning_rate": 2.9686883741912125e-05,
      "loss": 0.6129,
      "step": 9670
    },
    {
      "epoch": 2.34,
      "grad_norm": 0.31295427680015564,
      "learning_rate": 2.964769686288858e-05,
      "loss": 0.5826,
      "step": 9680
    },
    {
      "epoch": 2.34,
      "grad_norm": 0.31069615483283997,
      "learning_rate": 2.9608498152195928e-05,
      "loss": 0.5942,
      "step": 9690
    },
    {
      "epoch": 2.34,
      "grad_norm": 0.2735722064971924,
      "learning_rate": 2.9569287709622563e-05,
      "loss": 0.6068,
      "step": 9700
    },
    {
      "epoch": 2.35,
      "grad_norm": 0.2219727486371994,
      "learning_rate": 2.953006563498673e-05,
      "loss": 0.5923,
      "step": 9710
    },
    {
      "epoch": 2.35,
      "grad_norm": 0.2567574977874756,
      "learning_rate": 2.94908320281363e-05,
      "loss": 0.5893,
      "step": 9720
    },
    {
      "epoch": 2.35,
      "grad_norm": 0.2815946936607361,
      "learning_rate": 2.9451586988948492e-05,
      "loss": 0.5921,
      "step": 9730
    },
    {
      "epoch": 2.35,
      "grad_norm": 0.31510668992996216,
      "learning_rate": 2.941233061732963e-05,
      "loss": 0.6057,
      "step": 9740
    },
    {
      "epoch": 2.36,
      "grad_norm": 0.26856091618537903,
      "learning_rate": 2.9373063013214897e-05,
      "loss": 0.5932,
      "step": 9750
    },
    {
      "epoch": 2.36,
      "grad_norm": 0.2610081434249878,
      "learning_rate": 2.933378427656806e-05,
      "loss": 0.6147,
      "step": 9760
    },
    {
      "epoch": 2.36,
      "grad_norm": 0.607680082321167,
      "learning_rate": 2.9294494507381225e-05,
      "loss": 0.5894,
      "step": 9770
    },
    {
      "epoch": 2.36,
      "grad_norm": 0.22400206327438354,
      "learning_rate": 2.925519380567459e-05,
      "loss": 0.5905,
      "step": 9780
    },
    {
      "epoch": 2.37,
      "grad_norm": 0.23891247808933258,
      "learning_rate": 2.9215882271496187e-05,
      "loss": 0.6151,
      "step": 9790
    },
    {
      "epoch": 2.37,
      "grad_norm": 0.26922714710235596,
      "learning_rate": 2.9176560004921605e-05,
      "loss": 0.5743,
      "step": 9800
    },
    {
      "epoch": 2.37,
      "grad_norm": 0.23906663060188293,
      "learning_rate": 2.9137227106053773e-05,
      "loss": 0.5635,
      "step": 9810
    },
    {
      "epoch": 2.37,
      "grad_norm": 0.328803688287735,
      "learning_rate": 2.9097883675022685e-05,
      "loss": 0.6124,
      "step": 9820
    },
    {
      "epoch": 2.38,
      "grad_norm": 0.23414617776870728,
      "learning_rate": 2.9058529811985145e-05,
      "loss": 0.5927,
      "step": 9830
    },
    {
      "epoch": 2.38,
      "grad_norm": 0.2562233507633209,
      "learning_rate": 2.9019165617124515e-05,
      "loss": 0.5924,
      "step": 9840
    },
    {
      "epoch": 2.38,
      "grad_norm": 0.2643386125564575,
      "learning_rate": 2.8979791190650445e-05,
      "loss": 0.6075,
      "step": 9850
    },
    {
      "epoch": 2.38,
      "grad_norm": 0.2533874809741974,
      "learning_rate": 2.8940406632798655e-05,
      "loss": 0.602,
      "step": 9860
    },
    {
      "epoch": 2.39,
      "grad_norm": 0.24434837698936462,
      "learning_rate": 2.8901012043830637e-05,
      "loss": 0.6164,
      "step": 9870
    },
    {
      "epoch": 2.39,
      "grad_norm": 0.2734259068965912,
      "learning_rate": 2.886160752403343e-05,
      "loss": 0.6107,
      "step": 9880
    },
    {
      "epoch": 2.39,
      "grad_norm": 0.28115034103393555,
      "learning_rate": 2.8822193173719347e-05,
      "loss": 0.6327,
      "step": 9890
    },
    {
      "epoch": 2.39,
      "grad_norm": 0.3063926696777344,
      "learning_rate": 2.8782769093225727e-05,
      "loss": 0.6123,
      "step": 9900
    },
    {
      "epoch": 2.39,
      "grad_norm": 0.37148985266685486,
      "learning_rate": 2.87433353829147e-05,
      "loss": 0.5871,
      "step": 9910
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.2425338476896286,
      "learning_rate": 2.8703892143172874e-05,
      "loss": 0.6112,
      "step": 9920
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.20781515538692474,
      "learning_rate": 2.8664439474411142e-05,
      "loss": 0.6139,
      "step": 9930
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.2456682324409485,
      "learning_rate": 2.8624977477064397e-05,
      "loss": 0.5967,
      "step": 9940
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.42635104060173035,
      "learning_rate": 2.858550625159127e-05,
      "loss": 0.5918,
      "step": 9950
    },
    {
      "epoch": 2.41,
      "grad_norm": 0.28098204731941223,
      "learning_rate": 2.8546025898473898e-05,
      "loss": 0.6112,
      "step": 9960
    },
    {
      "epoch": 2.41,
      "grad_norm": 0.2819831073284149,
      "learning_rate": 2.8506536518217634e-05,
      "loss": 0.5838,
      "step": 9970
    },
    {
      "epoch": 2.41,
      "grad_norm": 0.2633509933948517,
      "learning_rate": 2.8467038211350845e-05,
      "loss": 0.5969,
      "step": 9980
    },
    {
      "epoch": 2.41,
      "grad_norm": 0.411672979593277,
      "learning_rate": 2.8427531078424585e-05,
      "loss": 0.6103,
      "step": 9990
    },
    {
      "epoch": 2.42,
      "grad_norm": 0.35813358426094055,
      "learning_rate": 2.8388015220012404e-05,
      "loss": 0.5883,
      "step": 10000
    },
    {
      "epoch": 2.42,
      "eval_loss": 0.6317408680915833,
      "eval_runtime": 67.2696,
      "eval_samples_per_second": 29.731,
      "eval_steps_per_second": 0.937,
      "step": 10000
    },
    {
      "epoch": 2.42,
      "grad_norm": 0.30284374952316284,
      "learning_rate": 2.8348490736710047e-05,
      "loss": 0.5987,
      "step": 10010
    },
    {
      "epoch": 2.42,
      "grad_norm": 0.24432061612606049,
      "learning_rate": 2.8308957729135226e-05,
      "loss": 0.6017,
      "step": 10020
    },
    {
      "epoch": 2.42,
      "grad_norm": 0.3139924108982086,
      "learning_rate": 2.8269416297927354e-05,
      "loss": 0.5923,
      "step": 10030
    },
    {
      "epoch": 2.43,
      "grad_norm": 0.2257777899503708,
      "learning_rate": 2.8229866543747263e-05,
      "loss": 0.5962,
      "step": 10040
    },
    {
      "epoch": 2.43,
      "grad_norm": 0.27044352889060974,
      "learning_rate": 2.8190308567277018e-05,
      "loss": 0.6025,
      "step": 10050
    },
    {
      "epoch": 2.43,
      "grad_norm": 0.24292701482772827,
      "learning_rate": 2.8150742469219583e-05,
      "loss": 0.6209,
      "step": 10060
    },
    {
      "epoch": 2.43,
      "grad_norm": 0.2715989947319031,
      "learning_rate": 2.8111168350298605e-05,
      "loss": 0.584,
      "step": 10070
    },
    {
      "epoch": 2.44,
      "grad_norm": 0.27733975648880005,
      "learning_rate": 2.8071586311258146e-05,
      "loss": 0.6006,
      "step": 10080
    },
    {
      "epoch": 2.44,
      "grad_norm": 0.22047729790210724,
      "learning_rate": 2.8031996452862443e-05,
      "loss": 0.6166,
      "step": 10090
    },
    {
      "epoch": 2.44,
      "grad_norm": 0.29286694526672363,
      "learning_rate": 2.7992398875895626e-05,
      "loss": 0.5847,
      "step": 10100
    },
    {
      "epoch": 2.44,
      "grad_norm": 0.3046290874481201,
      "learning_rate": 2.7952793681161472e-05,
      "loss": 0.5896,
      "step": 10110
    },
    {
      "epoch": 2.45,
      "grad_norm": 0.35147929191589355,
      "learning_rate": 2.7913180969483165e-05,
      "loss": 0.5912,
      "step": 10120
    },
    {
      "epoch": 2.45,
      "grad_norm": 0.3445433974266052,
      "learning_rate": 2.787356084170301e-05,
      "loss": 0.5868,
      "step": 10130
    },
    {
      "epoch": 2.45,
      "grad_norm": 0.2513396143913269,
      "learning_rate": 2.783393339868221e-05,
      "loss": 0.5864,
      "step": 10140
    },
    {
      "epoch": 2.45,
      "grad_norm": 0.2578549385070801,
      "learning_rate": 2.7794298741300566e-05,
      "loss": 0.6003,
      "step": 10150
    },
    {
      "epoch": 2.46,
      "grad_norm": 0.24509668350219727,
      "learning_rate": 2.7754656970456276e-05,
      "loss": 0.5681,
      "step": 10160
    },
    {
      "epoch": 2.46,
      "grad_norm": 0.2571702003479004,
      "learning_rate": 2.7715008187065607e-05,
      "loss": 0.5718,
      "step": 10170
    },
    {
      "epoch": 2.46,
      "grad_norm": 0.21920159459114075,
      "learning_rate": 2.7675352492062708e-05,
      "loss": 0.5838,
      "step": 10180
    },
    {
      "epoch": 2.46,
      "grad_norm": 0.4092113673686981,
      "learning_rate": 2.7635689986399303e-05,
      "loss": 0.5871,
      "step": 10190
    },
    {
      "epoch": 2.46,
      "grad_norm": 0.23578539490699768,
      "learning_rate": 2.759602077104448e-05,
      "loss": 0.6008,
      "step": 10200
    },
    {
      "epoch": 2.47,
      "grad_norm": 0.3018145263195038,
      "learning_rate": 2.7556344946984393e-05,
      "loss": 0.5997,
      "step": 10210
    },
    {
      "epoch": 2.47,
      "grad_norm": 0.2498631775379181,
      "learning_rate": 2.7516662615222005e-05,
      "loss": 0.5932,
      "step": 10220
    },
    {
      "epoch": 2.47,
      "grad_norm": 0.26965656876564026,
      "learning_rate": 2.7476973876776873e-05,
      "loss": 0.5662,
      "step": 10230
    },
    {
      "epoch": 2.47,
      "grad_norm": 0.263055682182312,
      "learning_rate": 2.743727883268485e-05,
      "loss": 0.6244,
      "step": 10240
    },
    {
      "epoch": 2.48,
      "grad_norm": 0.24924887716770172,
      "learning_rate": 2.7397577583997837e-05,
      "loss": 0.6451,
      "step": 10250
    },
    {
      "epoch": 2.48,
      "grad_norm": 0.24575600028038025,
      "learning_rate": 2.7357870231783535e-05,
      "loss": 0.6254,
      "step": 10260
    },
    {
      "epoch": 2.48,
      "grad_norm": 0.23808762431144714,
      "learning_rate": 2.7318156877125184e-05,
      "loss": 0.5941,
      "step": 10270
    },
    {
      "epoch": 2.48,
      "grad_norm": 0.25935983657836914,
      "learning_rate": 2.7278437621121306e-05,
      "loss": 0.5965,
      "step": 10280
    },
    {
      "epoch": 2.49,
      "grad_norm": 0.23470312356948853,
      "learning_rate": 2.723871256488544e-05,
      "loss": 0.5981,
      "step": 10290
    },
    {
      "epoch": 2.49,
      "grad_norm": 0.282685786485672,
      "learning_rate": 2.719898180954589e-05,
      "loss": 0.6036,
      "step": 10300
    },
    {
      "epoch": 2.49,
      "grad_norm": 0.2658996284008026,
      "learning_rate": 2.715924545624549e-05,
      "loss": 0.6023,
      "step": 10310
    },
    {
      "epoch": 2.49,
      "grad_norm": 0.22930899262428284,
      "learning_rate": 2.711950360614129e-05,
      "loss": 0.5816,
      "step": 10320
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.27134084701538086,
      "learning_rate": 2.7079756360404358e-05,
      "loss": 0.5873,
      "step": 10330
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.23598235845565796,
      "learning_rate": 2.7040003820219483e-05,
      "loss": 0.5954,
      "step": 10340
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.23500126600265503,
      "learning_rate": 2.700024608678494e-05,
      "loss": 0.5989,
      "step": 10350
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.21222622692584991,
      "learning_rate": 2.696048326131223e-05,
      "loss": 0.5756,
      "step": 10360
    },
    {
      "epoch": 2.51,
      "grad_norm": 0.27849477529525757,
      "learning_rate": 2.6920715445025814e-05,
      "loss": 0.6158,
      "step": 10370
    },
    {
      "epoch": 2.51,
      "grad_norm": 0.26395028829574585,
      "learning_rate": 2.6880942739162844e-05,
      "loss": 0.5845,
      "step": 10380
    },
    {
      "epoch": 2.51,
      "grad_norm": 0.27106571197509766,
      "learning_rate": 2.6841165244972937e-05,
      "loss": 0.5866,
      "step": 10390
    },
    {
      "epoch": 2.51,
      "grad_norm": 0.2842698395252228,
      "learning_rate": 2.68013830637179e-05,
      "loss": 0.5813,
      "step": 10400
    },
    {
      "epoch": 2.52,
      "grad_norm": 0.3903340995311737,
      "learning_rate": 2.6761596296671444e-05,
      "loss": 0.5909,
      "step": 10410
    },
    {
      "epoch": 2.52,
      "grad_norm": 0.21686884760856628,
      "learning_rate": 2.6721805045118985e-05,
      "loss": 0.6051,
      "step": 10420
    },
    {
      "epoch": 2.52,
      "grad_norm": 0.2636258602142334,
      "learning_rate": 2.6682009410357343e-05,
      "loss": 0.6066,
      "step": 10430
    },
    {
      "epoch": 2.52,
      "grad_norm": 0.27400442957878113,
      "learning_rate": 2.664220949369451e-05,
      "loss": 0.6107,
      "step": 10440
    },
    {
      "epoch": 2.53,
      "grad_norm": 0.2363210916519165,
      "learning_rate": 2.6602405396449354e-05,
      "loss": 0.6099,
      "step": 10450
    },
    {
      "epoch": 2.53,
      "grad_norm": 0.2572976350784302,
      "learning_rate": 2.65625972199514e-05,
      "loss": 0.6098,
      "step": 10460
    },
    {
      "epoch": 2.53,
      "grad_norm": 0.24827510118484497,
      "learning_rate": 2.652278506554057e-05,
      "loss": 0.599,
      "step": 10470
    },
    {
      "epoch": 2.53,
      "grad_norm": 0.25139114260673523,
      "learning_rate": 2.6482969034566875e-05,
      "loss": 0.5938,
      "step": 10480
    },
    {
      "epoch": 2.54,
      "grad_norm": 0.2517510950565338,
      "learning_rate": 2.644314922839023e-05,
      "loss": 0.6115,
      "step": 10490
    },
    {
      "epoch": 2.54,
      "grad_norm": 0.2278544306755066,
      "learning_rate": 2.640332574838015e-05,
      "loss": 0.6023,
      "step": 10500
    },
    {
      "epoch": 2.54,
      "eval_loss": 0.6308007836341858,
      "eval_runtime": 67.2785,
      "eval_samples_per_second": 29.727,
      "eval_steps_per_second": 0.936,
      "step": 10500
    },
    {
      "epoch": 2.54,
      "grad_norm": 0.19135211408138275,
      "learning_rate": 2.63634986959155e-05,
      "loss": 0.5877,
      "step": 10510
    },
    {
      "epoch": 2.54,
      "grad_norm": 0.25551867485046387,
      "learning_rate": 2.632366817238424e-05,
      "loss": 0.6095,
      "step": 10520
    },
    {
      "epoch": 2.54,
      "grad_norm": 0.21738716959953308,
      "learning_rate": 2.628383427918317e-05,
      "loss": 0.5754,
      "step": 10530
    },
    {
      "epoch": 2.55,
      "grad_norm": 0.261692613363266,
      "learning_rate": 2.624399711771766e-05,
      "loss": 0.6184,
      "step": 10540
    },
    {
      "epoch": 2.55,
      "grad_norm": 0.26923197507858276,
      "learning_rate": 2.6204156789401407e-05,
      "loss": 0.5911,
      "step": 10550
    },
    {
      "epoch": 2.55,
      "grad_norm": 0.27664855122566223,
      "learning_rate": 2.616431339565617e-05,
      "loss": 0.5788,
      "step": 10560
    },
    {
      "epoch": 2.55,
      "grad_norm": 0.24067947268486023,
      "learning_rate": 2.6124467037911504e-05,
      "loss": 0.6112,
      "step": 10570
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.2814284563064575,
      "learning_rate": 2.6084617817604516e-05,
      "loss": 0.6022,
      "step": 10580
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.2756374776363373,
      "learning_rate": 2.6044765836179602e-05,
      "loss": 0.6149,
      "step": 10590
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.25635477900505066,
      "learning_rate": 2.6004911195088187e-05,
      "loss": 0.5833,
      "step": 10600
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.2209189385175705,
      "learning_rate": 2.5965053995788468e-05,
      "loss": 0.596,
      "step": 10610
    },
    {
      "epoch": 2.57,
      "grad_norm": 0.7075193524360657,
      "learning_rate": 2.5925194339745136e-05,
      "loss": 0.5932,
      "step": 10620
    },
    {
      "epoch": 2.57,
      "grad_norm": 0.28784462809562683,
      "learning_rate": 2.5885332328429163e-05,
      "loss": 0.6118,
      "step": 10630
    },
    {
      "epoch": 2.57,
      "grad_norm": 0.26419249176979065,
      "learning_rate": 2.5845468063317496e-05,
      "loss": 0.618,
      "step": 10640
    },
    {
      "epoch": 2.57,
      "grad_norm": 0.23297163844108582,
      "learning_rate": 2.580560164589284e-05,
      "loss": 0.5814,
      "step": 10650
    },
    {
      "epoch": 2.58,
      "grad_norm": 0.2898586094379425,
      "learning_rate": 2.5765733177643355e-05,
      "loss": 0.5962,
      "step": 10660
    },
    {
      "epoch": 2.58,
      "grad_norm": 0.2931255102157593,
      "learning_rate": 2.5725862760062437e-05,
      "loss": 0.6289,
      "step": 10670
    },
    {
      "epoch": 2.58,
      "grad_norm": 0.27760767936706543,
      "learning_rate": 2.5685990494648444e-05,
      "loss": 0.5976,
      "step": 10680
    },
    {
      "epoch": 2.58,
      "grad_norm": 0.2815667688846588,
      "learning_rate": 2.5646116482904432e-05,
      "loss": 0.5649,
      "step": 10690
    },
    {
      "epoch": 2.59,
      "grad_norm": 0.22442054748535156,
      "learning_rate": 2.5606240826337908e-05,
      "loss": 0.5873,
      "step": 10700
    },
    {
      "epoch": 2.59,
      "grad_norm": 0.26230359077453613,
      "learning_rate": 2.5566363626460565e-05,
      "loss": 0.5881,
      "step": 10710
    },
    {
      "epoch": 2.59,
      "grad_norm": 0.3227124512195587,
      "learning_rate": 2.5526484984788023e-05,
      "loss": 0.5871,
      "step": 10720
    },
    {
      "epoch": 2.59,
      "grad_norm": 0.2664228677749634,
      "learning_rate": 2.5486605002839574e-05,
      "loss": 0.6118,
      "step": 10730
    },
    {
      "epoch": 2.6,
      "grad_norm": 0.3071444034576416,
      "learning_rate": 2.5446723782137926e-05,
      "loss": 0.6102,
      "step": 10740
    },
    {
      "epoch": 2.6,
      "grad_norm": 0.3024210035800934,
      "learning_rate": 2.540684142420892e-05,
      "loss": 0.6361,
      "step": 10750
    },
    {
      "epoch": 2.6,
      "grad_norm": 0.2285085916519165,
      "learning_rate": 2.536695803058133e-05,
      "loss": 0.5904,
      "step": 10760
    },
    {
      "epoch": 2.6,
      "grad_norm": 0.2410888373851776,
      "learning_rate": 2.5327073702786536e-05,
      "loss": 0.6251,
      "step": 10770
    },
    {
      "epoch": 2.61,
      "grad_norm": 0.27157413959503174,
      "learning_rate": 2.5287188542358298e-05,
      "loss": 0.6109,
      "step": 10780
    },
    {
      "epoch": 2.61,
      "grad_norm": 0.22924353182315826,
      "learning_rate": 2.524730265083251e-05,
      "loss": 0.614,
      "step": 10790
    },
    {
      "epoch": 2.61,
      "grad_norm": 0.26990529894828796,
      "learning_rate": 2.5207416129746913e-05,
      "loss": 0.6039,
      "step": 10800
    },
    {
      "epoch": 2.61,
      "grad_norm": 0.2284127026796341,
      "learning_rate": 2.5167529080640862e-05,
      "loss": 0.584,
      "step": 10810
    },
    {
      "epoch": 2.61,
      "grad_norm": 0.23105378448963165,
      "learning_rate": 2.512764160505504e-05,
      "loss": 0.6111,
      "step": 10820
    },
    {
      "epoch": 2.62,
      "grad_norm": 0.2851148247718811,
      "learning_rate": 2.508775380453125e-05,
      "loss": 0.5742,
      "step": 10830
    },
    {
      "epoch": 2.62,
      "grad_norm": 0.2170216143131256,
      "learning_rate": 2.5047865780612078e-05,
      "loss": 0.6303,
      "step": 10840
    },
    {
      "epoch": 2.62,
      "grad_norm": 0.3306771516799927,
      "learning_rate": 2.5007977634840713e-05,
      "loss": 0.6104,
      "step": 10850
    },
    {
      "epoch": 2.62,
      "grad_norm": 0.27079054713249207,
      "learning_rate": 2.4968089468760643e-05,
      "loss": 0.5814,
      "step": 10860
    },
    {
      "epoch": 2.63,
      "grad_norm": 0.3222564160823822,
      "learning_rate": 2.4928201383915392e-05,
      "loss": 0.601,
      "step": 10870
    },
    {
      "epoch": 2.63,
      "grad_norm": 0.21210065484046936,
      "learning_rate": 2.4888313481848298e-05,
      "loss": 0.5942,
      "step": 10880
    },
    {
      "epoch": 2.63,
      "grad_norm": 0.20585991442203522,
      "learning_rate": 2.484842586410222e-05,
      "loss": 0.588,
      "step": 10890
    },
    {
      "epoch": 2.63,
      "grad_norm": 0.22469113767147064,
      "learning_rate": 2.4808538632219316e-05,
      "loss": 0.6187,
      "step": 10900
    },
    {
      "epoch": 2.64,
      "grad_norm": 0.28604716062545776,
      "learning_rate": 2.4768651887740736e-05,
      "loss": 0.6074,
      "step": 10910
    },
    {
      "epoch": 2.64,
      "grad_norm": 0.2785724699497223,
      "learning_rate": 2.4728765732206393e-05,
      "loss": 0.6013,
      "step": 10920
    },
    {
      "epoch": 2.64,
      "grad_norm": 0.25105738639831543,
      "learning_rate": 2.4688880267154722e-05,
      "loss": 0.6189,
      "step": 10930
    },
    {
      "epoch": 2.64,
      "grad_norm": 0.25772786140441895,
      "learning_rate": 2.4648995594122367e-05,
      "loss": 0.6039,
      "step": 10940
    },
    {
      "epoch": 2.65,
      "grad_norm": 0.2568160593509674,
      "learning_rate": 2.4609111814643988e-05,
      "loss": 0.6091,
      "step": 10950
    },
    {
      "epoch": 2.65,
      "grad_norm": 0.2306566685438156,
      "learning_rate": 2.456922903025195e-05,
      "loss": 0.594,
      "step": 10960
    },
    {
      "epoch": 2.65,
      "grad_norm": 0.256623774766922,
      "learning_rate": 2.4529347342476082e-05,
      "loss": 0.5884,
      "step": 10970
    },
    {
      "epoch": 2.65,
      "grad_norm": 0.30249103903770447,
      "learning_rate": 2.4489466852843444e-05,
      "loss": 0.5972,
      "step": 10980
    },
    {
      "epoch": 2.66,
      "grad_norm": 0.31607022881507874,
      "learning_rate": 2.4449587662878032e-05,
      "loss": 0.572,
      "step": 10990
    },
    {
      "epoch": 2.66,
      "grad_norm": 0.3034932613372803,
      "learning_rate": 2.4409709874100524e-05,
      "loss": 0.5898,
      "step": 11000
    },
    {
      "epoch": 2.66,
      "eval_loss": 0.6311084032058716,
      "eval_runtime": 67.1728,
      "eval_samples_per_second": 29.774,
      "eval_steps_per_second": 0.938,
      "step": 11000
    },
    {
      "epoch": 2.66,
      "grad_norm": 0.27423152327537537,
      "learning_rate": 2.4369833588028045e-05,
      "loss": 0.6076,
      "step": 11010
    },
    {
      "epoch": 2.66,
      "grad_norm": 0.281901478767395,
      "learning_rate": 2.432995890617389e-05,
      "loss": 0.6151,
      "step": 11020
    },
    {
      "epoch": 2.67,
      "grad_norm": 0.2870057225227356,
      "learning_rate": 2.4290085930047267e-05,
      "loss": 0.5811,
      "step": 11030
    },
    {
      "epoch": 2.67,
      "grad_norm": 0.22503837943077087,
      "learning_rate": 2.4250214761153044e-05,
      "loss": 0.5827,
      "step": 11040
    },
    {
      "epoch": 2.67,
      "grad_norm": 0.2397720068693161,
      "learning_rate": 2.4210345500991495e-05,
      "loss": 0.6094,
      "step": 11050
    },
    {
      "epoch": 2.67,
      "grad_norm": 0.3159298598766327,
      "learning_rate": 2.417047825105802e-05,
      "loss": 0.6024,
      "step": 11060
    },
    {
      "epoch": 2.68,
      "grad_norm": 0.2685520648956299,
      "learning_rate": 2.4130613112842923e-05,
      "loss": 0.596,
      "step": 11070
    },
    {
      "epoch": 2.68,
      "grad_norm": 0.3091444671154022,
      "learning_rate": 2.409075018783112e-05,
      "loss": 0.5859,
      "step": 11080
    },
    {
      "epoch": 2.68,
      "grad_norm": 0.28575173020362854,
      "learning_rate": 2.405088957750188e-05,
      "loss": 0.593,
      "step": 11090
    },
    {
      "epoch": 2.68,
      "grad_norm": 0.19359290599822998,
      "learning_rate": 2.401103138332859e-05,
      "loss": 0.5569,
      "step": 11100
    },
    {
      "epoch": 2.68,
      "grad_norm": 0.2809715270996094,
      "learning_rate": 2.3971175706778494e-05,
      "loss": 0.5919,
      "step": 11110
    },
    {
      "epoch": 2.69,
      "grad_norm": 0.2378830909729004,
      "learning_rate": 2.3931322649312427e-05,
      "loss": 0.5889,
      "step": 11120
    },
    {
      "epoch": 2.69,
      "grad_norm": 0.2940937578678131,
      "learning_rate": 2.3891472312384548e-05,
      "loss": 0.5938,
      "step": 11130
    },
    {
      "epoch": 2.69,
      "grad_norm": 0.27026721835136414,
      "learning_rate": 2.3851624797442088e-05,
      "loss": 0.5832,
      "step": 11140
    },
    {
      "epoch": 2.69,
      "grad_norm": 0.2958199679851532,
      "learning_rate": 2.3811780205925104e-05,
      "loss": 0.6124,
      "step": 11150
    },
    {
      "epoch": 2.7,
      "grad_norm": 0.3658323585987091,
      "learning_rate": 2.37719386392662e-05,
      "loss": 0.5711,
      "step": 11160
    },
    {
      "epoch": 2.7,
      "grad_norm": 0.29097726941108704,
      "learning_rate": 2.373210019889029e-05,
      "loss": 0.5889,
      "step": 11170
    },
    {
      "epoch": 2.7,
      "grad_norm": 0.2615368366241455,
      "learning_rate": 2.3692264986214326e-05,
      "loss": 0.6016,
      "step": 11180
    },
    {
      "epoch": 2.7,
      "grad_norm": 0.29355910420417786,
      "learning_rate": 2.365243310264704e-05,
      "loss": 0.5766,
      "step": 11190
    },
    {
      "epoch": 2.71,
      "grad_norm": 0.5139521360397339,
      "learning_rate": 2.361260464958869e-05,
      "loss": 0.6089,
      "step": 11200
    },
    {
      "epoch": 2.71,
      "grad_norm": 0.2786054313182831,
      "learning_rate": 2.35727797284308e-05,
      "loss": 0.5833,
      "step": 11210
    },
    {
      "epoch": 2.71,
      "grad_norm": 0.3363838791847229,
      "learning_rate": 2.3532958440555912e-05,
      "loss": 0.6096,
      "step": 11220
    },
    {
      "epoch": 2.71,
      "grad_norm": 0.2659907042980194,
      "learning_rate": 2.349314088733731e-05,
      "loss": 0.615,
      "step": 11230
    },
    {
      "epoch": 2.72,
      "grad_norm": 0.5291584730148315,
      "learning_rate": 2.3453327170138767e-05,
      "loss": 0.6248,
      "step": 11240
    },
    {
      "epoch": 2.72,
      "grad_norm": 0.24188485741615295,
      "learning_rate": 2.34135173903143e-05,
      "loss": 0.599,
      "step": 11250
    },
    {
      "epoch": 2.72,
      "grad_norm": 0.2817278802394867,
      "learning_rate": 2.3373711649207894e-05,
      "loss": 0.5902,
      "step": 11260
    },
    {
      "epoch": 2.72,
      "grad_norm": 0.23381677269935608,
      "learning_rate": 2.333391004815326e-05,
      "loss": 0.5901,
      "step": 11270
    },
    {
      "epoch": 2.73,
      "grad_norm": 0.23642730712890625,
      "learning_rate": 2.3294112688473574e-05,
      "loss": 0.5952,
      "step": 11280
    },
    {
      "epoch": 2.73,
      "grad_norm": 0.26222020387649536,
      "learning_rate": 2.325431967148119e-05,
      "loss": 0.5755,
      "step": 11290
    },
    {
      "epoch": 2.73,
      "grad_norm": 0.24193081259727478,
      "learning_rate": 2.3214531098477444e-05,
      "loss": 0.5885,
      "step": 11300
    },
    {
      "epoch": 2.73,
      "grad_norm": 0.25730979442596436,
      "learning_rate": 2.3174747070752325e-05,
      "loss": 0.6022,
      "step": 11310
    },
    {
      "epoch": 2.74,
      "grad_norm": 0.2317025512456894,
      "learning_rate": 2.3134967689584282e-05,
      "loss": 0.6031,
      "step": 11320
    },
    {
      "epoch": 2.74,
      "grad_norm": 0.3505846858024597,
      "learning_rate": 2.30951930562399e-05,
      "loss": 0.5984,
      "step": 11330
    },
    {
      "epoch": 2.74,
      "grad_norm": 0.2722209393978119,
      "learning_rate": 2.3055423271973693e-05,
      "loss": 0.5938,
      "step": 11340
    },
    {
      "epoch": 2.74,
      "grad_norm": 0.26406392455101013,
      "learning_rate": 2.3015658438027848e-05,
      "loss": 0.6124,
      "step": 11350
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.23661866784095764,
      "learning_rate": 2.2975898655631934e-05,
      "loss": 0.6019,
      "step": 11360
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.2529833912849426,
      "learning_rate": 2.2936144026002654e-05,
      "loss": 0.5768,
      "step": 11370
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.28207114338874817,
      "learning_rate": 2.289639465034361e-05,
      "loss": 0.6138,
      "step": 11380
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.23660141229629517,
      "learning_rate": 2.2856650629845013e-05,
      "loss": 0.6115,
      "step": 11390
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.29152390360832214,
      "learning_rate": 2.2816912065683452e-05,
      "loss": 0.6052,
      "step": 11400
    },
    {
      "epoch": 2.76,
      "grad_norm": 0.22973516583442688,
      "learning_rate": 2.2777179059021624e-05,
      "loss": 0.6114,
      "step": 11410
    },
    {
      "epoch": 2.76,
      "grad_norm": 0.2583330571651459,
      "learning_rate": 2.273745171100807e-05,
      "loss": 0.5999,
      "step": 11420
    },
    {
      "epoch": 2.76,
      "grad_norm": 0.2971131503582001,
      "learning_rate": 2.2697730122776934e-05,
      "loss": 0.5824,
      "step": 11430
    },
    {
      "epoch": 2.76,
      "grad_norm": 0.25637754797935486,
      "learning_rate": 2.2658014395447703e-05,
      "loss": 0.6169,
      "step": 11440
    },
    {
      "epoch": 2.77,
      "grad_norm": 0.23559825122356415,
      "learning_rate": 2.261830463012493e-05,
      "loss": 0.5988,
      "step": 11450
    },
    {
      "epoch": 2.77,
      "grad_norm": 0.25648581981658936,
      "learning_rate": 2.2578600927898e-05,
      "loss": 0.5789,
      "step": 11460
    },
    {
      "epoch": 2.77,
      "grad_norm": 0.26241737604141235,
      "learning_rate": 2.2538903389840863e-05,
      "loss": 0.5919,
      "step": 11470
    },
    {
      "epoch": 2.77,
      "grad_norm": 0.4651870131492615,
      "learning_rate": 2.2499212117011763e-05,
      "loss": 0.5575,
      "step": 11480
    },
    {
      "epoch": 2.78,
      "grad_norm": 0.22167663276195526,
      "learning_rate": 2.2459527210453018e-05,
      "loss": 0.5879,
      "step": 11490
    },
    {
      "epoch": 2.78,
      "grad_norm": 0.258618026971817,
      "learning_rate": 2.2419848771190718e-05,
      "loss": 0.576,
      "step": 11500
    },
    {
      "epoch": 2.78,
      "eval_loss": 0.6290733814239502,
      "eval_runtime": 67.2009,
      "eval_samples_per_second": 29.762,
      "eval_steps_per_second": 0.937,
      "step": 11500
    },
    {
      "epoch": 2.78,
      "grad_norm": 0.23681069910526276,
      "learning_rate": 2.2380176900234505e-05,
      "loss": 0.5946,
      "step": 11510
    },
    {
      "epoch": 2.78,
      "grad_norm": 0.2815132439136505,
      "learning_rate": 2.2340511698577278e-05,
      "loss": 0.6174,
      "step": 11520
    },
    {
      "epoch": 2.79,
      "grad_norm": 0.2809601128101349,
      "learning_rate": 2.2300853267194995e-05,
      "loss": 0.6148,
      "step": 11530
    },
    {
      "epoch": 2.79,
      "grad_norm": 0.2977219223976135,
      "learning_rate": 2.2261201707046345e-05,
      "loss": 0.5943,
      "step": 11540
    },
    {
      "epoch": 2.79,
      "grad_norm": 0.37512487173080444,
      "learning_rate": 2.222155711907254e-05,
      "loss": 0.5865,
      "step": 11550
    },
    {
      "epoch": 2.79,
      "grad_norm": 0.2732716500759125,
      "learning_rate": 2.218191960419703e-05,
      "loss": 0.5988,
      "step": 11560
    },
    {
      "epoch": 2.8,
      "grad_norm": 0.25176912546157837,
      "learning_rate": 2.2142289263325287e-05,
      "loss": 0.5866,
      "step": 11570
    },
    {
      "epoch": 2.8,
      "grad_norm": 0.3007689416408539,
      "learning_rate": 2.210266619734449e-05,
      "loss": 0.5757,
      "step": 11580
    },
    {
      "epoch": 2.8,
      "grad_norm": 0.22966600954532623,
      "learning_rate": 2.2063050507123303e-05,
      "loss": 0.6029,
      "step": 11590
    },
    {
      "epoch": 2.8,
      "grad_norm": 0.2383899837732315,
      "learning_rate": 2.2023442293511632e-05,
      "loss": 0.6104,
      "step": 11600
    },
    {
      "epoch": 2.81,
      "grad_norm": 0.3035461902618408,
      "learning_rate": 2.1983841657340335e-05,
      "loss": 0.5959,
      "step": 11610
    },
    {
      "epoch": 2.81,
      "grad_norm": 0.35818007588386536,
      "learning_rate": 2.194424869942098e-05,
      "loss": 0.5975,
      "step": 11620
    },
    {
      "epoch": 2.81,
      "grad_norm": 0.2527376413345337,
      "learning_rate": 2.1904663520545598e-05,
      "loss": 0.5946,
      "step": 11630
    },
    {
      "epoch": 2.81,
      "grad_norm": 0.4648072421550751,
      "learning_rate": 2.1865086221486404e-05,
      "loss": 0.6145,
      "step": 11640
    },
    {
      "epoch": 2.82,
      "grad_norm": 0.23881538212299347,
      "learning_rate": 2.1825516902995564e-05,
      "loss": 0.5863,
      "step": 11650
    },
    {
      "epoch": 2.82,
      "grad_norm": 0.3298749327659607,
      "learning_rate": 2.1785955665804917e-05,
      "loss": 0.5972,
      "step": 11660
    },
    {
      "epoch": 2.82,
      "grad_norm": 0.2054242491722107,
      "learning_rate": 2.1746402610625748e-05,
      "loss": 0.5892,
      "step": 11670
    },
    {
      "epoch": 2.82,
      "grad_norm": 0.234252467751503,
      "learning_rate": 2.1706857838148493e-05,
      "loss": 0.6132,
      "step": 11680
    },
    {
      "epoch": 2.83,
      "grad_norm": 0.271100252866745,
      "learning_rate": 2.1667321449042512e-05,
      "loss": 0.6064,
      "step": 11690
    },
    {
      "epoch": 2.83,
      "grad_norm": 0.2767755687236786,
      "learning_rate": 2.1627793543955825e-05,
      "loss": 0.5936,
      "step": 11700
    },
    {
      "epoch": 2.83,
      "grad_norm": 0.24367476999759674,
      "learning_rate": 2.1588274223514844e-05,
      "loss": 0.6025,
      "step": 11710
    },
    {
      "epoch": 2.83,
      "grad_norm": 0.26168349385261536,
      "learning_rate": 2.1548763588324143e-05,
      "loss": 0.5963,
      "step": 11720
    },
    {
      "epoch": 2.83,
      "grad_norm": 0.21627385914325714,
      "learning_rate": 2.1509261738966174e-05,
      "loss": 0.5925,
      "step": 11730
    },
    {
      "epoch": 2.84,
      "grad_norm": 0.27272096276283264,
      "learning_rate": 2.1469768776001023e-05,
      "loss": 0.5767,
      "step": 11740
    },
    {
      "epoch": 2.84,
      "grad_norm": 0.2876264750957489,
      "learning_rate": 2.1430284799966155e-05,
      "loss": 0.5666,
      "step": 11750
    },
    {
      "epoch": 2.84,
      "grad_norm": 0.24279695749282837,
      "learning_rate": 2.1390809911376165e-05,
      "loss": 0.5988,
      "step": 11760
    },
    {
      "epoch": 2.84,
      "grad_norm": 0.26712995767593384,
      "learning_rate": 2.1351344210722512e-05,
      "loss": 0.5845,
      "step": 11770
    },
    {
      "epoch": 2.85,
      "grad_norm": 0.27297675609588623,
      "learning_rate": 2.1311887798473253e-05,
      "loss": 0.5906,
      "step": 11780
    },
    {
      "epoch": 2.85,
      "grad_norm": 0.20993754267692566,
      "learning_rate": 2.1272440775072812e-05,
      "loss": 0.5805,
      "step": 11790
    },
    {
      "epoch": 2.85,
      "grad_norm": 0.26127463579177856,
      "learning_rate": 2.12330032409417e-05,
      "loss": 0.5836,
      "step": 11800
    },
    {
      "epoch": 2.85,
      "grad_norm": 0.6268289089202881,
      "learning_rate": 2.119357529647629e-05,
      "loss": 0.5689,
      "step": 11810
    },
    {
      "epoch": 2.86,
      "grad_norm": 0.2731999456882477,
      "learning_rate": 2.1154157042048522e-05,
      "loss": 0.6072,
      "step": 11820
    },
    {
      "epoch": 2.86,
      "grad_norm": 0.31480905413627625,
      "learning_rate": 2.111474857800568e-05,
      "loss": 0.59,
      "step": 11830
    },
    {
      "epoch": 2.86,
      "grad_norm": 1.1808747053146362,
      "learning_rate": 2.1075350004670118e-05,
      "loss": 0.6054,
      "step": 11840
    },
    {
      "epoch": 2.86,
      "grad_norm": 0.397026389837265,
      "learning_rate": 2.103596142233902e-05,
      "loss": 0.6116,
      "step": 11850
    },
    {
      "epoch": 2.87,
      "grad_norm": 0.27153530716896057,
      "learning_rate": 2.099658293128413e-05,
      "loss": 0.5955,
      "step": 11860
    },
    {
      "epoch": 2.87,
      "grad_norm": 0.29511383175849915,
      "learning_rate": 2.09572146317515e-05,
      "loss": 0.5935,
      "step": 11870
    },
    {
      "epoch": 2.87,
      "grad_norm": 0.25342148542404175,
      "learning_rate": 2.091785662396124e-05,
      "loss": 0.5909,
      "step": 11880
    },
    {
      "epoch": 2.87,
      "grad_norm": 0.2613728940486908,
      "learning_rate": 2.0878509008107255e-05,
      "loss": 0.5765,
      "step": 11890
    },
    {
      "epoch": 2.88,
      "grad_norm": 0.3336595296859741,
      "learning_rate": 2.083917188435702e-05,
      "loss": 0.5875,
      "step": 11900
    },
    {
      "epoch": 2.88,
      "grad_norm": 0.24153658747673035,
      "learning_rate": 2.079984535285127e-05,
      "loss": 0.5963,
      "step": 11910
    },
    {
      "epoch": 2.88,
      "grad_norm": 0.20768964290618896,
      "learning_rate": 2.0760529513703783e-05,
      "loss": 0.5892,
      "step": 11920
    },
    {
      "epoch": 2.88,
      "grad_norm": 0.31084755063056946,
      "learning_rate": 2.0721224467001126e-05,
      "loss": 0.6118,
      "step": 11930
    },
    {
      "epoch": 2.89,
      "grad_norm": 0.2713968753814697,
      "learning_rate": 2.0681930312802384e-05,
      "loss": 0.6273,
      "step": 11940
    },
    {
      "epoch": 2.89,
      "grad_norm": 0.22585509717464447,
      "learning_rate": 2.0642647151138914e-05,
      "loss": 0.5808,
      "step": 11950
    },
    {
      "epoch": 2.89,
      "grad_norm": 0.22664512693881989,
      "learning_rate": 2.0603375082014095e-05,
      "loss": 0.5849,
      "step": 11960
    },
    {
      "epoch": 2.89,
      "grad_norm": 0.2538776695728302,
      "learning_rate": 2.056411420540306e-05,
      "loss": 0.5719,
      "step": 11970
    },
    {
      "epoch": 2.9,
      "grad_norm": 0.30300286412239075,
      "learning_rate": 2.0524864621252448e-05,
      "loss": 0.6011,
      "step": 11980
    },
    {
      "epoch": 2.9,
      "grad_norm": 0.21366974711418152,
      "learning_rate": 2.0485626429480165e-05,
      "loss": 0.5993,
      "step": 11990
    },
    {
      "epoch": 2.9,
      "grad_norm": 0.35079991817474365,
      "learning_rate": 2.04463997299751e-05,
      "loss": 0.5699,
      "step": 12000
    },
    {
      "epoch": 2.9,
      "eval_loss": 0.6291141510009766,
      "eval_runtime": 67.26,
      "eval_samples_per_second": 29.735,
      "eval_steps_per_second": 0.937,
      "step": 12000
    },
    {
      "epoch": 2.9,
      "grad_norm": 0.26222503185272217,
      "learning_rate": 2.0407184622596893e-05,
      "loss": 0.6118,
      "step": 12010
    },
    {
      "epoch": 2.9,
      "grad_norm": 0.2841508388519287,
      "learning_rate": 2.036798120717567e-05,
      "loss": 0.5965,
      "step": 12020
    },
    {
      "epoch": 2.91,
      "grad_norm": 0.23934020102024078,
      "learning_rate": 2.03287895835118e-05,
      "loss": 0.5796,
      "step": 12030
    },
    {
      "epoch": 2.91,
      "grad_norm": 0.296139657497406,
      "learning_rate": 2.0289609851375622e-05,
      "loss": 0.5819,
      "step": 12040
    },
    {
      "epoch": 2.91,
      "grad_norm": 0.2427491545677185,
      "learning_rate": 2.025044211050721e-05,
      "loss": 0.5946,
      "step": 12050
    },
    {
      "epoch": 2.91,
      "grad_norm": 0.24599017202854156,
      "learning_rate": 2.0211286460616126e-05,
      "loss": 0.5966,
      "step": 12060
    },
    {
      "epoch": 2.92,
      "grad_norm": 0.23710612952709198,
      "learning_rate": 2.0172143001381115e-05,
      "loss": 0.6212,
      "step": 12070
    },
    {
      "epoch": 2.92,
      "grad_norm": 0.22220022976398468,
      "learning_rate": 2.013301183244992e-05,
      "loss": 0.5821,
      "step": 12080
    },
    {
      "epoch": 2.92,
      "grad_norm": 0.49431440234184265,
      "learning_rate": 2.0093893053438985e-05,
      "loss": 0.5618,
      "step": 12090
    },
    {
      "epoch": 2.92,
      "grad_norm": 0.2873918414115906,
      "learning_rate": 2.005478676393322e-05,
      "loss": 0.5902,
      "step": 12100
    },
    {
      "epoch": 2.93,
      "grad_norm": 0.25227466225624084,
      "learning_rate": 2.001569306348573e-05,
      "loss": 0.6166,
      "step": 12110
    },
    {
      "epoch": 2.93,
      "grad_norm": 0.25766241550445557,
      "learning_rate": 1.9976612051617575e-05,
      "loss": 0.5913,
      "step": 12120
    },
    {
      "epoch": 2.93,
      "grad_norm": 0.2500627636909485,
      "learning_rate": 1.993754382781752e-05,
      "loss": 0.5989,
      "step": 12130
    },
    {
      "epoch": 2.93,
      "grad_norm": 0.2150602489709854,
      "learning_rate": 1.9898488491541778e-05,
      "loss": 0.5943,
      "step": 12140
    },
    {
      "epoch": 2.94,
      "grad_norm": 0.358582079410553,
      "learning_rate": 1.9859446142213733e-05,
      "loss": 0.5906,
      "step": 12150
    },
    {
      "epoch": 2.94,
      "grad_norm": 0.2866133153438568,
      "learning_rate": 1.982041687922373e-05,
      "loss": 0.5908,
      "step": 12160
    },
    {
      "epoch": 2.94,
      "grad_norm": 0.39618533849716187,
      "learning_rate": 1.9781400801928792e-05,
      "loss": 0.5931,
      "step": 12170
    },
    {
      "epoch": 2.94,
      "grad_norm": 0.2567703425884247,
      "learning_rate": 1.9742398009652375e-05,
      "loss": 0.5765,
      "step": 12180
    },
    {
      "epoch": 2.95,
      "grad_norm": 0.45456215739250183,
      "learning_rate": 1.9703408601684108e-05,
      "loss": 0.6154,
      "step": 12190
    },
    {
      "epoch": 2.95,
      "grad_norm": 0.41257551312446594,
      "learning_rate": 1.9664432677279567e-05,
      "loss": 0.6084,
      "step": 12200
    },
    {
      "epoch": 2.95,
      "grad_norm": 0.23691555857658386,
      "learning_rate": 1.9625470335659982e-05,
      "loss": 0.6152,
      "step": 12210
    },
    {
      "epoch": 2.95,
      "grad_norm": 0.36675411462783813,
      "learning_rate": 1.9586521676012022e-05,
      "loss": 0.6038,
      "step": 12220
    },
    {
      "epoch": 2.96,
      "grad_norm": 0.28038734197616577,
      "learning_rate": 1.954758679748751e-05,
      "loss": 0.5973,
      "step": 12230
    },
    {
      "epoch": 2.96,
      "grad_norm": 0.27618926763534546,
      "learning_rate": 1.9508665799203195e-05,
      "loss": 0.597,
      "step": 12240
    },
    {
      "epoch": 2.96,
      "grad_norm": 0.27861592173576355,
      "learning_rate": 1.94697587802405e-05,
      "loss": 0.6154,
      "step": 12250
    },
    {
      "epoch": 2.96,
      "grad_norm": 0.26619014143943787,
      "learning_rate": 1.943086583964524e-05,
      "loss": 0.5895,
      "step": 12260
    },
    {
      "epoch": 2.97,
      "grad_norm": 0.45826488733291626,
      "learning_rate": 1.9391987076427404e-05,
      "loss": 0.5932,
      "step": 12270
    },
    {
      "epoch": 2.97,
      "grad_norm": 0.26050978899002075,
      "learning_rate": 1.935312258956089e-05,
      "loss": 0.5946,
      "step": 12280
    },
    {
      "epoch": 2.97,
      "grad_norm": 0.31627464294433594,
      "learning_rate": 1.9314272477983243e-05,
      "loss": 0.5753,
      "step": 12290
    },
    {
      "epoch": 2.97,
      "grad_norm": 0.2674327492713928,
      "learning_rate": 1.9275436840595426e-05,
      "loss": 0.5921,
      "step": 12300
    },
    {
      "epoch": 2.97,
      "grad_norm": 0.2414204329252243,
      "learning_rate": 1.923661577626156e-05,
      "loss": 0.5923,
      "step": 12310
    },
    {
      "epoch": 2.98,
      "grad_norm": 0.2296002209186554,
      "learning_rate": 1.919780938380863e-05,
      "loss": 0.5895,
      "step": 12320
    },
    {
      "epoch": 2.98,
      "grad_norm": 0.23779046535491943,
      "learning_rate": 1.915901776202631e-05,
      "loss": 0.5567,
      "step": 12330
    },
    {
      "epoch": 2.98,
      "grad_norm": 0.2853798568248749,
      "learning_rate": 1.9120241009666658e-05,
      "loss": 0.5711,
      "step": 12340
    },
    {
      "epoch": 2.98,
      "grad_norm": 0.26367607712745667,
      "learning_rate": 1.908147922544387e-05,
      "loss": 0.6071,
      "step": 12350
    },
    {
      "epoch": 2.99,
      "grad_norm": 0.23959624767303467,
      "learning_rate": 1.9042732508034062e-05,
      "loss": 0.5891,
      "step": 12360
    },
    {
      "epoch": 2.99,
      "grad_norm": 0.21563242375850677,
      "learning_rate": 1.900400095607497e-05,
      "loss": 0.5867,
      "step": 12370
    },
    {
      "epoch": 2.99,
      "grad_norm": 0.27703240513801575,
      "learning_rate": 1.896528466816574e-05,
      "loss": 0.6282,
      "step": 12380
    },
    {
      "epoch": 2.99,
      "grad_norm": 0.21136893332004547,
      "learning_rate": 1.8926583742866648e-05,
      "loss": 0.589,
      "step": 12390
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.23731322586536407,
      "learning_rate": 1.888789827869886e-05,
      "loss": 0.5715,
      "step": 12400
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.23349808156490326,
      "learning_rate": 1.88492283741442e-05,
      "loss": 0.5869,
      "step": 12410
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.29223474860191345,
      "learning_rate": 1.8810574127644865e-05,
      "loss": 0.5895,
      "step": 12420
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.24479955434799194,
      "learning_rate": 1.877193563760319e-05,
      "loss": 0.5704,
      "step": 12430
    },
    {
      "epoch": 3.01,
      "grad_norm": 0.3249976933002472,
      "learning_rate": 1.8733313002381417e-05,
      "loss": 0.5865,
      "step": 12440
    },
    {
      "epoch": 3.01,
      "grad_norm": 0.2533276081085205,
      "learning_rate": 1.8694706320301408e-05,
      "loss": 0.5764,
      "step": 12450
    },
    {
      "epoch": 3.01,
      "grad_norm": 0.21968787908554077,
      "learning_rate": 1.865611568964442e-05,
      "loss": 0.6075,
      "step": 12460
    },
    {
      "epoch": 3.01,
      "grad_norm": 0.26140934228897095,
      "learning_rate": 1.861754120865085e-05,
      "loss": 0.5914,
      "step": 12470
    },
    {
      "epoch": 3.02,
      "grad_norm": 0.2797393500804901,
      "learning_rate": 1.857898297551998e-05,
      "loss": 0.5894,
      "step": 12480
    },
    {
      "epoch": 3.02,
      "grad_norm": 0.21738211810588837,
      "learning_rate": 1.8540441088409723e-05,
      "loss": 0.5926,
      "step": 12490
    },
    {
      "epoch": 3.02,
      "grad_norm": 0.2682536840438843,
      "learning_rate": 1.8501915645436396e-05,
      "loss": 0.6093,
      "step": 12500
    },
    {
      "epoch": 3.02,
      "eval_loss": 0.6290040016174316,
      "eval_runtime": 67.129,
      "eval_samples_per_second": 29.793,
      "eval_steps_per_second": 0.938,
      "step": 12500
    },
    {
      "epoch": 3.02,
      "grad_norm": 0.244039848446846,
      "learning_rate": 1.846340674467444e-05,
      "loss": 0.6037,
      "step": 12510
    },
    {
      "epoch": 3.03,
      "grad_norm": 0.25075218081474304,
      "learning_rate": 1.8424914484156178e-05,
      "loss": 0.5869,
      "step": 12520
    },
    {
      "epoch": 3.03,
      "grad_norm": 0.21912485361099243,
      "learning_rate": 1.8386438961871602e-05,
      "loss": 0.5982,
      "step": 12530
    },
    {
      "epoch": 3.03,
      "grad_norm": 0.2603309750556946,
      "learning_rate": 1.8347980275768066e-05,
      "loss": 0.5728,
      "step": 12540
    },
    {
      "epoch": 3.03,
      "grad_norm": 0.22281397879123688,
      "learning_rate": 1.8309538523750075e-05,
      "loss": 0.5919,
      "step": 12550
    },
    {
      "epoch": 3.04,
      "grad_norm": 0.25083571672439575,
      "learning_rate": 1.827111380367903e-05,
      "loss": 0.5794,
      "step": 12560
    },
    {
      "epoch": 3.04,
      "grad_norm": 0.2255573868751526,
      "learning_rate": 1.8232706213372935e-05,
      "loss": 0.5907,
      "step": 12570
    },
    {
      "epoch": 3.04,
      "grad_norm": 0.24321115016937256,
      "learning_rate": 1.8194315850606248e-05,
      "loss": 0.5837,
      "step": 12580
    },
    {
      "epoch": 3.04,
      "grad_norm": 0.294193297624588,
      "learning_rate": 1.8155942813109523e-05,
      "loss": 0.5927,
      "step": 12590
    },
    {
      "epoch": 3.04,
      "grad_norm": 0.290958970785141,
      "learning_rate": 1.811758719856924e-05,
      "loss": 0.5821,
      "step": 12600
    },
    {
      "epoch": 3.05,
      "grad_norm": 0.2639531195163727,
      "learning_rate": 1.80792491046275e-05,
      "loss": 0.6157,
      "step": 12610
    },
    {
      "epoch": 3.05,
      "grad_norm": 0.2535398304462433,
      "learning_rate": 1.804092862888182e-05,
      "loss": 0.6162,
      "step": 12620
    },
    {
      "epoch": 3.05,
      "grad_norm": 0.3162219226360321,
      "learning_rate": 1.800262586888486e-05,
      "loss": 0.5951,
      "step": 12630
    },
    {
      "epoch": 3.05,
      "grad_norm": 0.22566238045692444,
      "learning_rate": 1.7964340922144173e-05,
      "loss": 0.5748,
      "step": 12640
    },
    {
      "epoch": 3.06,
      "grad_norm": 0.3351154923439026,
      "learning_rate": 1.7926073886121973e-05,
      "loss": 0.6037,
      "step": 12650
    },
    {
      "epoch": 3.06,
      "grad_norm": 0.2531121075153351,
      "learning_rate": 1.788782485823488e-05,
      "loss": 0.5783,
      "step": 12660
    },
    {
      "epoch": 3.06,
      "grad_norm": 0.31655964255332947,
      "learning_rate": 1.7849593935853676e-05,
      "loss": 0.5867,
      "step": 12670
    },
    {
      "epoch": 3.06,
      "grad_norm": 0.3081534802913666,
      "learning_rate": 1.7811381216303034e-05,
      "loss": 0.5773,
      "step": 12680
    },
    {
      "epoch": 3.07,
      "grad_norm": 0.3251687288284302,
      "learning_rate": 1.7773186796861298e-05,
      "loss": 0.5811,
      "step": 12690
    },
    {
      "epoch": 3.07,
      "grad_norm": 0.2518479526042938,
      "learning_rate": 1.773501077476023e-05,
      "loss": 0.5979,
      "step": 12700
    },
    {
      "epoch": 3.07,
      "grad_norm": 0.26604709029197693,
      "learning_rate": 1.7696853247184754e-05,
      "loss": 0.5809,
      "step": 12710
    },
    {
      "epoch": 3.07,
      "grad_norm": 0.47998151183128357,
      "learning_rate": 1.7658714311272705e-05,
      "loss": 0.5646,
      "step": 12720
    },
    {
      "epoch": 3.08,
      "grad_norm": 0.3014504015445709,
      "learning_rate": 1.7620594064114596e-05,
      "loss": 0.6015,
      "step": 12730
    },
    {
      "epoch": 3.08,
      "grad_norm": 0.3193497955799103,
      "learning_rate": 1.7582492602753366e-05,
      "loss": 0.5847,
      "step": 12740
    },
    {
      "epoch": 3.08,
      "grad_norm": 0.3089773952960968,
      "learning_rate": 1.7544410024184124e-05,
      "loss": 0.5835,
      "step": 12750
    },
    {
      "epoch": 3.08,
      "grad_norm": 0.35424813628196716,
      "learning_rate": 1.750634642535392e-05,
      "loss": 0.592,
      "step": 12760
    },
    {
      "epoch": 3.09,
      "grad_norm": 0.28449276089668274,
      "learning_rate": 1.746830190316148e-05,
      "loss": 0.5647,
      "step": 12770
    },
    {
      "epoch": 3.09,
      "grad_norm": 0.2374422550201416,
      "learning_rate": 1.743027655445696e-05,
      "loss": 0.5622,
      "step": 12780
    },
    {
      "epoch": 3.09,
      "grad_norm": 0.29000285267829895,
      "learning_rate": 1.7392270476041723e-05,
      "loss": 0.5957,
      "step": 12790
    },
    {
      "epoch": 3.09,
      "grad_norm": 0.35042545199394226,
      "learning_rate": 1.7354283764668057e-05,
      "loss": 0.609,
      "step": 12800
    },
    {
      "epoch": 3.1,
      "grad_norm": 0.23474256694316864,
      "learning_rate": 1.7316316517038956e-05,
      "loss": 0.5988,
      "step": 12810
    },
    {
      "epoch": 3.1,
      "grad_norm": 0.2591429352760315,
      "learning_rate": 1.7278368829807864e-05,
      "loss": 0.5941,
      "step": 12820
    },
    {
      "epoch": 3.1,
      "grad_norm": 0.2101234495639801,
      "learning_rate": 1.7240440799578432e-05,
      "loss": 0.5852,
      "step": 12830
    },
    {
      "epoch": 3.1,
      "grad_norm": 0.28370776772499084,
      "learning_rate": 1.720253252290427e-05,
      "loss": 0.5792,
      "step": 12840
    },
    {
      "epoch": 3.11,
      "grad_norm": 0.3017008900642395,
      "learning_rate": 1.71646440962887e-05,
      "loss": 0.5818,
      "step": 12850
    },
    {
      "epoch": 3.11,
      "grad_norm": 0.2230721116065979,
      "learning_rate": 1.712677561618451e-05,
      "loss": 0.5679,
      "step": 12860
    },
    {
      "epoch": 3.11,
      "grad_norm": 0.274613618850708,
      "learning_rate": 1.7088927178993707e-05,
      "loss": 0.5818,
      "step": 12870
    },
    {
      "epoch": 3.11,
      "grad_norm": 0.3069445788860321,
      "learning_rate": 1.7051098881067285e-05,
      "loss": 0.5928,
      "step": 12880
    },
    {
      "epoch": 3.12,
      "grad_norm": 0.2262873351573944,
      "learning_rate": 1.7013290818704958e-05,
      "loss": 0.603,
      "step": 12890
    },
    {
      "epoch": 3.12,
      "grad_norm": 0.2648998200893402,
      "learning_rate": 1.6975503088154942e-05,
      "loss": 0.597,
      "step": 12900
    },
    {
      "epoch": 3.12,
      "grad_norm": 0.3107318580150604,
      "learning_rate": 1.6937735785613675e-05,
      "loss": 0.5764,
      "step": 12910
    },
    {
      "epoch": 3.12,
      "grad_norm": 0.22244475781917572,
      "learning_rate": 1.6899989007225602e-05,
      "loss": 0.595,
      "step": 12920
    },
    {
      "epoch": 3.12,
      "grad_norm": 0.2974041700363159,
      "learning_rate": 1.6862262849082918e-05,
      "loss": 0.5847,
      "step": 12930
    },
    {
      "epoch": 3.13,
      "grad_norm": 0.31232744455337524,
      "learning_rate": 1.6824557407225327e-05,
      "loss": 0.5647,
      "step": 12940
    },
    {
      "epoch": 3.13,
      "grad_norm": 0.2458878457546234,
      "learning_rate": 1.678687277763979e-05,
      "loss": 0.5689,
      "step": 12950
    },
    {
      "epoch": 3.13,
      "grad_norm": 0.2889362573623657,
      "learning_rate": 1.6749209056260288e-05,
      "loss": 0.5877,
      "step": 12960
    },
    {
      "epoch": 3.13,
      "grad_norm": 0.3089177906513214,
      "learning_rate": 1.6711566338967577e-05,
      "loss": 0.5906,
      "step": 12970
    },
    {
      "epoch": 3.14,
      "grad_norm": 0.2911435067653656,
      "learning_rate": 1.667394472158894e-05,
      "loss": 0.6117,
      "step": 12980
    },
    {
      "epoch": 3.14,
      "grad_norm": 0.26806724071502686,
      "learning_rate": 1.6636344299897963e-05,
      "loss": 0.5781,
      "step": 12990
    },
    {
      "epoch": 3.14,
      "grad_norm": 0.2905066907405853,
      "learning_rate": 1.6598765169614243e-05,
      "loss": 0.5754,
      "step": 13000
    },
    {
      "epoch": 3.14,
      "eval_loss": 0.6291921138763428,
      "eval_runtime": 67.1137,
      "eval_samples_per_second": 29.8,
      "eval_steps_per_second": 0.939,
      "step": 13000
    },
    {
      "epoch": 3.14,
      "grad_norm": 0.22024208307266235,
      "learning_rate": 1.65612074264032e-05,
      "loss": 0.5669,
      "step": 13010
    },
    {
      "epoch": 3.15,
      "grad_norm": 0.31273871660232544,
      "learning_rate": 1.6523671165875803e-05,
      "loss": 0.5747,
      "step": 13020
    },
    {
      "epoch": 3.15,
      "grad_norm": 0.28565514087677,
      "learning_rate": 1.6486156483588326e-05,
      "loss": 0.5926,
      "step": 13030
    },
    {
      "epoch": 3.15,
      "grad_norm": 0.2939855456352234,
      "learning_rate": 1.644866347504211e-05,
      "loss": 0.6168,
      "step": 13040
    },
    {
      "epoch": 3.15,
      "grad_norm": 0.2832067310810089,
      "learning_rate": 1.641119223568334e-05,
      "loss": 0.563,
      "step": 13050
    },
    {
      "epoch": 3.16,
      "grad_norm": 0.24646145105361938,
      "learning_rate": 1.6373742860902756e-05,
      "loss": 0.6023,
      "step": 13060
    },
    {
      "epoch": 3.16,
      "grad_norm": 0.2672120928764343,
      "learning_rate": 1.6336315446035454e-05,
      "loss": 0.6038,
      "step": 13070
    },
    {
      "epoch": 3.16,
      "grad_norm": 0.25014764070510864,
      "learning_rate": 1.6298910086360628e-05,
      "loss": 0.5727,
      "step": 13080
    },
    {
      "epoch": 3.16,
      "grad_norm": 0.2431202381849289,
      "learning_rate": 1.6261526877101314e-05,
      "loss": 0.573,
      "step": 13090
    },
    {
      "epoch": 3.17,
      "grad_norm": 0.26143622398376465,
      "learning_rate": 1.6224165913424173e-05,
      "loss": 0.598,
      "step": 13100
    },
    {
      "epoch": 3.17,
      "grad_norm": 0.31623631715774536,
      "learning_rate": 1.6186827290439225e-05,
      "loss": 0.5963,
      "step": 13110
    },
    {
      "epoch": 3.17,
      "grad_norm": 0.24909652769565582,
      "learning_rate": 1.6149511103199616e-05,
      "loss": 0.6004,
      "step": 13120
    },
    {
      "epoch": 3.17,
      "grad_norm": 0.2736775279045105,
      "learning_rate": 1.6112217446701393e-05,
      "loss": 0.5938,
      "step": 13130
    },
    {
      "epoch": 3.18,
      "grad_norm": 0.24586784839630127,
      "learning_rate": 1.6074946415883236e-05,
      "loss": 0.5964,
      "step": 13140
    },
    {
      "epoch": 3.18,
      "grad_norm": 0.23617498576641083,
      "learning_rate": 1.6037698105626213e-05,
      "loss": 0.598,
      "step": 13150
    },
    {
      "epoch": 3.18,
      "grad_norm": 0.23995892703533173,
      "learning_rate": 1.600047261075358e-05,
      "loss": 0.5928,
      "step": 13160
    },
    {
      "epoch": 3.18,
      "grad_norm": 0.27234625816345215,
      "learning_rate": 1.5963270026030493e-05,
      "loss": 0.5989,
      "step": 13170
    },
    {
      "epoch": 3.19,
      "grad_norm": 0.2851603031158447,
      "learning_rate": 1.592609044616378e-05,
      "loss": 0.584,
      "step": 13180
    },
    {
      "epoch": 3.19,
      "grad_norm": 0.24323895573616028,
      "learning_rate": 1.588893396580173e-05,
      "loss": 0.5933,
      "step": 13190
    },
    {
      "epoch": 3.19,
      "grad_norm": 0.4213368594646454,
      "learning_rate": 1.5851800679533794e-05,
      "loss": 0.5801,
      "step": 13200
    },
    {
      "epoch": 3.19,
      "grad_norm": 0.2636274993419647,
      "learning_rate": 1.5814690681890415e-05,
      "loss": 0.5868,
      "step": 13210
    },
    {
      "epoch": 3.19,
      "grad_norm": 0.2891892194747925,
      "learning_rate": 1.577760406734272e-05,
      "loss": 0.5865,
      "step": 13220
    },
    {
      "epoch": 3.2,
      "grad_norm": 0.2736075222492218,
      "learning_rate": 1.574054093030233e-05,
      "loss": 0.5988,
      "step": 13230
    },
    {
      "epoch": 3.2,
      "grad_norm": 0.2928101420402527,
      "learning_rate": 1.5703501365121083e-05,
      "loss": 0.5951,
      "step": 13240
    },
    {
      "epoch": 3.2,
      "grad_norm": 0.2634028196334839,
      "learning_rate": 1.5666485466090815e-05,
      "loss": 0.5918,
      "step": 13250
    },
    {
      "epoch": 3.2,
      "grad_norm": 0.24428558349609375,
      "learning_rate": 1.5629493327443118e-05,
      "loss": 0.5926,
      "step": 13260
    },
    {
      "epoch": 3.21,
      "grad_norm": 0.24596397578716278,
      "learning_rate": 1.5592525043349102e-05,
      "loss": 0.6074,
      "step": 13270
    },
    {
      "epoch": 3.21,
      "grad_norm": 0.24957533180713654,
      "learning_rate": 1.555558070791914e-05,
      "loss": 0.5818,
      "step": 13280
    },
    {
      "epoch": 3.21,
      "grad_norm": 0.23625266551971436,
      "learning_rate": 1.5518660415202647e-05,
      "loss": 0.5797,
      "step": 13290
    },
    {
      "epoch": 3.21,
      "grad_norm": 0.3125057518482208,
      "learning_rate": 1.5481764259187816e-05,
      "loss": 0.6017,
      "step": 13300
    },
    {
      "epoch": 3.22,
      "grad_norm": 0.290134459733963,
      "learning_rate": 1.5444892333801414e-05,
      "loss": 0.5903,
      "step": 13310
    },
    {
      "epoch": 3.22,
      "grad_norm": 0.3861146867275238,
      "learning_rate": 1.540804473290852e-05,
      "loss": 0.5916,
      "step": 13320
    },
    {
      "epoch": 3.22,
      "grad_norm": 0.3101816475391388,
      "learning_rate": 1.5371221550312288e-05,
      "loss": 0.6039,
      "step": 13330
    },
    {
      "epoch": 3.22,
      "grad_norm": 0.24264681339263916,
      "learning_rate": 1.533442287975371e-05,
      "loss": 0.5609,
      "step": 13340
    },
    {
      "epoch": 3.23,
      "grad_norm": 0.27712488174438477,
      "learning_rate": 1.5297648814911363e-05,
      "loss": 0.5543,
      "step": 13350
    },
    {
      "epoch": 3.23,
      "grad_norm": 0.2841449975967407,
      "learning_rate": 1.5260899449401217e-05,
      "loss": 0.5915,
      "step": 13360
    },
    {
      "epoch": 3.23,
      "grad_norm": 0.24764423072338104,
      "learning_rate": 1.522417487677634e-05,
      "loss": 0.5787,
      "step": 13370
    },
    {
      "epoch": 3.23,
      "grad_norm": 0.27273863554000854,
      "learning_rate": 1.518747519052669e-05,
      "loss": 0.5971,
      "step": 13380
    },
    {
      "epoch": 3.24,
      "grad_norm": 0.28492897748947144,
      "learning_rate": 1.5150800484078878e-05,
      "loss": 0.6063,
      "step": 13390
    },
    {
      "epoch": 3.24,
      "grad_norm": 0.32788246870040894,
      "learning_rate": 1.5114150850795913e-05,
      "loss": 0.5725,
      "step": 13400
    },
    {
      "epoch": 3.24,
      "grad_norm": 0.47013843059539795,
      "learning_rate": 1.5077526383976986e-05,
      "loss": 0.5684,
      "step": 13410
    },
    {
      "epoch": 3.24,
      "grad_norm": 0.312014102935791,
      "learning_rate": 1.504092717685721e-05,
      "loss": 0.608,
      "step": 13420
    },
    {
      "epoch": 3.25,
      "grad_norm": 0.22890594601631165,
      "learning_rate": 1.5004353322607407e-05,
      "loss": 0.5558,
      "step": 13430
    },
    {
      "epoch": 3.25,
      "grad_norm": 0.24974258244037628,
      "learning_rate": 1.4967804914333844e-05,
      "loss": 0.5752,
      "step": 13440
    },
    {
      "epoch": 3.25,
      "grad_norm": 0.25919726490974426,
      "learning_rate": 1.4931282045078027e-05,
      "loss": 0.5665,
      "step": 13450
    },
    {
      "epoch": 3.25,
      "grad_norm": 0.91816645860672,
      "learning_rate": 1.4894784807816436e-05,
      "loss": 0.6129,
      "step": 13460
    },
    {
      "epoch": 3.26,
      "grad_norm": 0.2525247633457184,
      "learning_rate": 1.48583132954603e-05,
      "loss": 0.5571,
      "step": 13470
    },
    {
      "epoch": 3.26,
      "grad_norm": 0.27164921164512634,
      "learning_rate": 1.482186760085536e-05,
      "loss": 0.6146,
      "step": 13480
    },
    {
      "epoch": 3.26,
      "grad_norm": 0.3456345796585083,
      "learning_rate": 1.4785447816781636e-05,
      "loss": 0.5815,
      "step": 13490
    },
    {
      "epoch": 3.26,
      "grad_norm": 0.319261372089386,
      "learning_rate": 1.4749054035953186e-05,
      "loss": 0.6294,
      "step": 13500
    },
    {
      "epoch": 3.26,
      "eval_loss": 0.628188967704773,
      "eval_runtime": 67.1584,
      "eval_samples_per_second": 29.78,
      "eval_steps_per_second": 0.938,
      "step": 13500
    },
    {
      "epoch": 3.26,
      "grad_norm": 0.26717182993888855,
      "learning_rate": 1.4712686351017869e-05,
      "loss": 0.5856,
      "step": 13510
    },
    {
      "epoch": 3.27,
      "grad_norm": 0.2683553397655487,
      "learning_rate": 1.4676344854557111e-05,
      "loss": 0.596,
      "step": 13520
    },
    {
      "epoch": 3.27,
      "grad_norm": 0.2221207320690155,
      "learning_rate": 1.4640029639085672e-05,
      "loss": 0.5712,
      "step": 13530
    },
    {
      "epoch": 3.27,
      "grad_norm": 0.27545279264450073,
      "learning_rate": 1.4603740797051427e-05,
      "loss": 0.5711,
      "step": 13540
    },
    {
      "epoch": 3.27,
      "grad_norm": 0.25706854462623596,
      "learning_rate": 1.4567478420835063e-05,
      "loss": 0.5571,
      "step": 13550
    },
    {
      "epoch": 3.28,
      "grad_norm": 0.27466848492622375,
      "learning_rate": 1.4531242602749945e-05,
      "loss": 0.6243,
      "step": 13560
    },
    {
      "epoch": 3.28,
      "grad_norm": 0.27467504143714905,
      "learning_rate": 1.44950334350418e-05,
      "loss": 0.5736,
      "step": 13570
    },
    {
      "epoch": 3.28,
      "grad_norm": 0.2699827253818512,
      "learning_rate": 1.445885100988852e-05,
      "loss": 0.5682,
      "step": 13580
    },
    {
      "epoch": 3.28,
      "grad_norm": 0.3584638237953186,
      "learning_rate": 1.4422695419399912e-05,
      "loss": 0.6159,
      "step": 13590
    },
    {
      "epoch": 3.29,
      "grad_norm": 0.26607200503349304,
      "learning_rate": 1.4386566755617481e-05,
      "loss": 0.5933,
      "step": 13600
    },
    {
      "epoch": 3.29,
      "grad_norm": 0.3185410797595978,
      "learning_rate": 1.4350465110514173e-05,
      "loss": 0.5867,
      "step": 13610
    },
    {
      "epoch": 3.29,
      "grad_norm": 0.3076455891132355,
      "learning_rate": 1.4314390575994158e-05,
      "loss": 0.6021,
      "step": 13620
    },
    {
      "epoch": 3.29,
      "grad_norm": 0.2944042384624481,
      "learning_rate": 1.4278343243892591e-05,
      "loss": 0.5724,
      "step": 13630
    },
    {
      "epoch": 3.3,
      "grad_norm": 0.31576016545295715,
      "learning_rate": 1.4242323205975377e-05,
      "loss": 0.5957,
      "step": 13640
    },
    {
      "epoch": 3.3,
      "grad_norm": 0.29806095361709595,
      "learning_rate": 1.4206330553938934e-05,
      "loss": 0.5885,
      "step": 13650
    },
    {
      "epoch": 3.3,
      "grad_norm": 0.2612191438674927,
      "learning_rate": 1.4170365379409972e-05,
      "loss": 0.5932,
      "step": 13660
    },
    {
      "epoch": 3.3,
      "grad_norm": 0.2951468825340271,
      "learning_rate": 1.4134427773945241e-05,
      "loss": 0.5865,
      "step": 13670
    },
    {
      "epoch": 3.31,
      "grad_norm": 0.3061927855014801,
      "learning_rate": 1.4098517829031316e-05,
      "loss": 0.57,
      "step": 13680
    },
    {
      "epoch": 3.31,
      "grad_norm": 0.304347425699234,
      "learning_rate": 1.4062635636084354e-05,
      "loss": 0.5793,
      "step": 13690
    },
    {
      "epoch": 3.31,
      "grad_norm": 0.2383164018392563,
      "learning_rate": 1.4026781286449864e-05,
      "loss": 0.5777,
      "step": 13700
    },
    {
      "epoch": 3.31,
      "grad_norm": 0.28097665309906006,
      "learning_rate": 1.3990954871402462e-05,
      "loss": 0.5898,
      "step": 13710
    },
    {
      "epoch": 3.32,
      "grad_norm": 0.291500449180603,
      "learning_rate": 1.3955156482145686e-05,
      "loss": 0.5624,
      "step": 13720
    },
    {
      "epoch": 3.32,
      "grad_norm": 0.29882997274398804,
      "learning_rate": 1.3919386209811692e-05,
      "loss": 0.5985,
      "step": 13730
    },
    {
      "epoch": 3.32,
      "grad_norm": 0.2620564103126526,
      "learning_rate": 1.3883644145461069e-05,
      "loss": 0.5614,
      "step": 13740
    },
    {
      "epoch": 3.32,
      "grad_norm": 0.2394556701183319,
      "learning_rate": 1.3847930380082605e-05,
      "loss": 0.5542,
      "step": 13750
    },
    {
      "epoch": 3.33,
      "grad_norm": 0.2630850076675415,
      "learning_rate": 1.3812245004593039e-05,
      "loss": 0.6184,
      "step": 13760
    },
    {
      "epoch": 3.33,
      "grad_norm": 0.22769545018672943,
      "learning_rate": 1.3776588109836841e-05,
      "loss": 0.5724,
      "step": 13770
    },
    {
      "epoch": 3.33,
      "grad_norm": 0.31552964448928833,
      "learning_rate": 1.3740959786585971e-05,
      "loss": 0.5861,
      "step": 13780
    },
    {
      "epoch": 3.33,
      "grad_norm": 0.2552676498889923,
      "learning_rate": 1.370536012553968e-05,
      "loss": 0.5954,
      "step": 13790
    },
    {
      "epoch": 3.33,
      "grad_norm": 0.25912633538246155,
      "learning_rate": 1.3669789217324208e-05,
      "loss": 0.5921,
      "step": 13800
    },
    {
      "epoch": 3.34,
      "grad_norm": 0.2176179140806198,
      "learning_rate": 1.3634247152492635e-05,
      "loss": 0.5917,
      "step": 13810
    },
    {
      "epoch": 3.34,
      "grad_norm": 0.2581143081188202,
      "learning_rate": 1.35987340215246e-05,
      "loss": 0.5953,
      "step": 13820
    },
    {
      "epoch": 3.34,
      "grad_norm": 0.28065741062164307,
      "learning_rate": 1.3563249914826095e-05,
      "loss": 0.5938,
      "step": 13830
    },
    {
      "epoch": 3.34,
      "grad_norm": 0.24440771341323853,
      "learning_rate": 1.3527794922729215e-05,
      "loss": 0.5786,
      "step": 13840
    },
    {
      "epoch": 3.35,
      "grad_norm": 0.3171332776546478,
      "learning_rate": 1.3492369135491927e-05,
      "loss": 0.5856,
      "step": 13850
    },
    {
      "epoch": 3.35,
      "grad_norm": 0.3420179486274719,
      "learning_rate": 1.3456972643297888e-05,
      "loss": 0.5964,
      "step": 13860
    },
    {
      "epoch": 3.35,
      "grad_norm": 0.3031647503376007,
      "learning_rate": 1.3421605536256148e-05,
      "loss": 0.575,
      "step": 13870
    },
    {
      "epoch": 3.35,
      "grad_norm": 0.25836315751075745,
      "learning_rate": 1.3386267904400956e-05,
      "loss": 0.5705,
      "step": 13880
    },
    {
      "epoch": 3.36,
      "grad_norm": 0.26373711228370667,
      "learning_rate": 1.3350959837691528e-05,
      "loss": 0.589,
      "step": 13890
    },
    {
      "epoch": 3.36,
      "grad_norm": 0.29463452100753784,
      "learning_rate": 1.3315681426011817e-05,
      "loss": 0.5992,
      "step": 13900
    },
    {
      "epoch": 3.36,
      "grad_norm": 0.3104293942451477,
      "learning_rate": 1.328043275917028e-05,
      "loss": 0.5658,
      "step": 13910
    },
    {
      "epoch": 3.36,
      "grad_norm": 0.2526361346244812,
      "learning_rate": 1.3245213926899657e-05,
      "loss": 0.5952,
      "step": 13920
    },
    {
      "epoch": 3.37,
      "grad_norm": 0.24315546452999115,
      "learning_rate": 1.3210025018856736e-05,
      "loss": 0.5706,
      "step": 13930
    },
    {
      "epoch": 3.37,
      "grad_norm": 0.2698476016521454,
      "learning_rate": 1.3174866124622121e-05,
      "loss": 0.6062,
      "step": 13940
    },
    {
      "epoch": 3.37,
      "grad_norm": 0.29000622034072876,
      "learning_rate": 1.3139737333700019e-05,
      "loss": 0.5738,
      "step": 13950
    },
    {
      "epoch": 3.37,
      "grad_norm": 0.2777455449104309,
      "learning_rate": 1.3104638735517999e-05,
      "loss": 0.5948,
      "step": 13960
    },
    {
      "epoch": 3.38,
      "grad_norm": 0.23992811143398285,
      "learning_rate": 1.3069570419426769e-05,
      "loss": 0.5875,
      "step": 13970
    },
    {
      "epoch": 3.38,
      "grad_norm": 0.33731457591056824,
      "learning_rate": 1.303453247469994e-05,
      "loss": 0.5963,
      "step": 13980
    },
    {
      "epoch": 3.38,
      "grad_norm": 0.2974693477153778,
      "learning_rate": 1.2999524990533817e-05,
      "loss": 0.5743,
      "step": 13990
    },
    {
      "epoch": 3.38,
      "grad_norm": 0.26297178864479065,
      "learning_rate": 1.296454805604716e-05,
      "loss": 0.591,
      "step": 14000
    },
    {
      "epoch": 3.38,
      "eval_loss": 0.6283172369003296,
      "eval_runtime": 67.2721,
      "eval_samples_per_second": 29.73,
      "eval_steps_per_second": 0.936,
      "step": 14000
    },
    {
      "epoch": 3.39,
      "grad_norm": 0.33907148241996765,
      "learning_rate": 1.2929601760280951e-05,
      "loss": 0.5744,
      "step": 14010
    },
    {
      "epoch": 3.39,
      "grad_norm": 0.23848184943199158,
      "learning_rate": 1.2894686192198185e-05,
      "loss": 0.5837,
      "step": 14020
    },
    {
      "epoch": 3.39,
      "grad_norm": 0.23010163009166718,
      "learning_rate": 1.2859801440683622e-05,
      "loss": 0.5789,
      "step": 14030
    },
    {
      "epoch": 3.39,
      "grad_norm": 0.27091971039772034,
      "learning_rate": 1.2824947594543585e-05,
      "loss": 0.5973,
      "step": 14040
    },
    {
      "epoch": 3.4,
      "grad_norm": 0.2591380774974823,
      "learning_rate": 1.2790124742505707e-05,
      "loss": 0.5959,
      "step": 14050
    },
    {
      "epoch": 3.4,
      "grad_norm": 0.28662291169166565,
      "learning_rate": 1.275533297321873e-05,
      "loss": 0.5831,
      "step": 14060
    },
    {
      "epoch": 3.4,
      "grad_norm": 0.3045969009399414,
      "learning_rate": 1.2720572375252265e-05,
      "loss": 0.5661,
      "step": 14070
    },
    {
      "epoch": 3.4,
      "grad_norm": 0.3080662488937378,
      "learning_rate": 1.2685843037096568e-05,
      "loss": 0.6112,
      "step": 14080
    },
    {
      "epoch": 3.41,
      "grad_norm": 0.2610645294189453,
      "learning_rate": 1.2651145047162318e-05,
      "loss": 0.5864,
      "step": 14090
    },
    {
      "epoch": 3.41,
      "grad_norm": 0.33825016021728516,
      "learning_rate": 1.2616478493780393e-05,
      "loss": 0.602,
      "step": 14100
    },
    {
      "epoch": 3.41,
      "grad_norm": 0.30778974294662476,
      "learning_rate": 1.2581843465201643e-05,
      "loss": 0.5965,
      "step": 14110
    },
    {
      "epoch": 3.41,
      "grad_norm": 0.31114304065704346,
      "learning_rate": 1.2547240049596656e-05,
      "loss": 0.5883,
      "step": 14120
    },
    {
      "epoch": 3.41,
      "grad_norm": 0.22209766507148743,
      "learning_rate": 1.2512668335055555e-05,
      "loss": 0.5779,
      "step": 14130
    },
    {
      "epoch": 3.42,
      "grad_norm": 0.4704256057739258,
      "learning_rate": 1.2478128409587755e-05,
      "loss": 0.5939,
      "step": 14140
    },
    {
      "epoch": 3.42,
      "grad_norm": 0.4344974458217621,
      "learning_rate": 1.2443620361121743e-05,
      "loss": 0.6017,
      "step": 14150
    },
    {
      "epoch": 3.42,
      "grad_norm": 0.2912081182003021,
      "learning_rate": 1.2409144277504866e-05,
      "loss": 0.6152,
      "step": 14160
    },
    {
      "epoch": 3.42,
      "grad_norm": 0.3525172770023346,
      "learning_rate": 1.2374700246503081e-05,
      "loss": 0.5971,
      "step": 14170
    },
    {
      "epoch": 3.43,
      "grad_norm": 0.32317399978637695,
      "learning_rate": 1.2340288355800775e-05,
      "loss": 0.5869,
      "step": 14180
    },
    {
      "epoch": 3.43,
      "grad_norm": 0.2769618630409241,
      "learning_rate": 1.2305908693000489e-05,
      "loss": 0.5793,
      "step": 14190
    },
    {
      "epoch": 3.43,
      "grad_norm": 0.2443193644285202,
      "learning_rate": 1.2271561345622739e-05,
      "loss": 0.5792,
      "step": 14200
    },
    {
      "epoch": 3.43,
      "grad_norm": 0.22904929518699646,
      "learning_rate": 1.2237246401105762e-05,
      "loss": 0.5795,
      "step": 14210
    },
    {
      "epoch": 3.44,
      "grad_norm": 0.27904483675956726,
      "learning_rate": 1.2202963946805313e-05,
      "loss": 0.5659,
      "step": 14220
    },
    {
      "epoch": 3.44,
      "grad_norm": 0.27308809757232666,
      "learning_rate": 1.2168714069994438e-05,
      "loss": 0.5938,
      "step": 14230
    },
    {
      "epoch": 3.44,
      "grad_norm": 0.263809472322464,
      "learning_rate": 1.2134496857863254e-05,
      "loss": 0.5655,
      "step": 14240
    },
    {
      "epoch": 3.44,
      "grad_norm": 0.2016855776309967,
      "learning_rate": 1.210031239751871e-05,
      "loss": 0.5856,
      "step": 14250
    },
    {
      "epoch": 3.45,
      "grad_norm": 0.294119656085968,
      "learning_rate": 1.206616077598439e-05,
      "loss": 0.5685,
      "step": 14260
    },
    {
      "epoch": 3.45,
      "grad_norm": 0.2778434157371521,
      "learning_rate": 1.2032042080200275e-05,
      "loss": 0.5639,
      "step": 14270
    },
    {
      "epoch": 3.45,
      "grad_norm": 0.27407506108283997,
      "learning_rate": 1.1997956397022542e-05,
      "loss": 0.599,
      "step": 14280
    },
    {
      "epoch": 3.45,
      "grad_norm": 0.28835251927375793,
      "learning_rate": 1.1963903813223292e-05,
      "loss": 0.5834,
      "step": 14290
    },
    {
      "epoch": 3.46,
      "grad_norm": 0.33038949966430664,
      "learning_rate": 1.19298844154904e-05,
      "loss": 0.5913,
      "step": 14300
    },
    {
      "epoch": 3.46,
      "grad_norm": 0.24437056481838226,
      "learning_rate": 1.1895898290427232e-05,
      "loss": 0.5783,
      "step": 14310
    },
    {
      "epoch": 3.46,
      "grad_norm": 0.233577698469162,
      "learning_rate": 1.186194552455249e-05,
      "loss": 0.5903,
      "step": 14320
    },
    {
      "epoch": 3.46,
      "grad_norm": 0.26180973649024963,
      "learning_rate": 1.182802620429991e-05,
      "loss": 0.5893,
      "step": 14330
    },
    {
      "epoch": 3.47,
      "grad_norm": 0.31134265661239624,
      "learning_rate": 1.1794140416018112e-05,
      "loss": 0.6027,
      "step": 14340
    },
    {
      "epoch": 3.47,
      "grad_norm": 0.2642928957939148,
      "learning_rate": 1.1760288245970343e-05,
      "loss": 0.6103,
      "step": 14350
    },
    {
      "epoch": 3.47,
      "grad_norm": 0.29957324266433716,
      "learning_rate": 1.1726469780334273e-05,
      "loss": 0.6005,
      "step": 14360
    },
    {
      "epoch": 3.47,
      "grad_norm": 0.2773592472076416,
      "learning_rate": 1.1692685105201764e-05,
      "loss": 0.5912,
      "step": 14370
    },
    {
      "epoch": 3.48,
      "grad_norm": 0.2868300676345825,
      "learning_rate": 1.1658934306578662e-05,
      "loss": 0.5984,
      "step": 14380
    },
    {
      "epoch": 3.48,
      "grad_norm": 0.3231727182865143,
      "learning_rate": 1.1625217470384575e-05,
      "loss": 0.5791,
      "step": 14390
    },
    {
      "epoch": 3.48,
      "grad_norm": 0.24512945115566254,
      "learning_rate": 1.1591534682452645e-05,
      "loss": 0.5964,
      "step": 14400
    },
    {
      "epoch": 3.48,
      "grad_norm": 0.25115954875946045,
      "learning_rate": 1.1557886028529349e-05,
      "loss": 0.5926,
      "step": 14410
    },
    {
      "epoch": 3.48,
      "grad_norm": 0.24561603367328644,
      "learning_rate": 1.152427159427426e-05,
      "loss": 0.6006,
      "step": 14420
    },
    {
      "epoch": 3.49,
      "grad_norm": 0.32604101300239563,
      "learning_rate": 1.1490691465259836e-05,
      "loss": 0.5947,
      "step": 14430
    },
    {
      "epoch": 3.49,
      "grad_norm": 0.3201538622379303,
      "learning_rate": 1.1457145726971213e-05,
      "loss": 0.5837,
      "step": 14440
    },
    {
      "epoch": 3.49,
      "grad_norm": 0.30608269572257996,
      "learning_rate": 1.1423634464805972e-05,
      "loss": 0.585,
      "step": 14450
    },
    {
      "epoch": 3.49,
      "grad_norm": 0.2931601107120514,
      "learning_rate": 1.1390157764073928e-05,
      "loss": 0.5629,
      "step": 14460
    },
    {
      "epoch": 3.5,
      "grad_norm": 0.2878846228122711,
      "learning_rate": 1.1356715709996915e-05,
      "loss": 0.5809,
      "step": 14470
    },
    {
      "epoch": 3.5,
      "grad_norm": 0.29207083582878113,
      "learning_rate": 1.1323308387708556e-05,
      "loss": 0.579,
      "step": 14480
    },
    {
      "epoch": 3.5,
      "grad_norm": 0.32907265424728394,
      "learning_rate": 1.1289935882254086e-05,
      "loss": 0.6063,
      "step": 14490
    },
    {
      "epoch": 3.5,
      "grad_norm": 0.22724080085754395,
      "learning_rate": 1.1256598278590074e-05,
      "loss": 0.599,
      "step": 14500
    },
    {
      "epoch": 3.5,
      "eval_loss": 0.6272611618041992,
      "eval_runtime": 67.1443,
      "eval_samples_per_second": 29.787,
      "eval_steps_per_second": 0.938,
      "step": 14500
    },
    {
      "epoch": 3.51,
      "grad_norm": 0.278616338968277,
      "learning_rate": 1.122329566158426e-05,
      "loss": 0.5815,
      "step": 14510
    },
    {
      "epoch": 3.51,
      "grad_norm": 0.2185838669538498,
      "learning_rate": 1.1190028116015317e-05,
      "loss": 0.5939,
      "step": 14520
    },
    {
      "epoch": 3.51,
      "grad_norm": 0.2838265299797058,
      "learning_rate": 1.1156795726572612e-05,
      "loss": 0.5988,
      "step": 14530
    },
    {
      "epoch": 3.51,
      "grad_norm": 0.3511171340942383,
      "learning_rate": 1.112359857785604e-05,
      "loss": 0.6013,
      "step": 14540
    },
    {
      "epoch": 3.52,
      "grad_norm": 0.2815764546394348,
      "learning_rate": 1.109043675437578e-05,
      "loss": 0.5695,
      "step": 14550
    },
    {
      "epoch": 3.52,
      "grad_norm": 0.2923150062561035,
      "learning_rate": 1.1057310340552077e-05,
      "loss": 0.5783,
      "step": 14560
    },
    {
      "epoch": 3.52,
      "grad_norm": 0.35399025678634644,
      "learning_rate": 1.102421942071504e-05,
      "loss": 0.6142,
      "step": 14570
    },
    {
      "epoch": 3.52,
      "grad_norm": 0.28528544306755066,
      "learning_rate": 1.0991164079104418e-05,
      "loss": 0.5913,
      "step": 14580
    },
    {
      "epoch": 3.53,
      "grad_norm": 0.27019065618515015,
      "learning_rate": 1.0958144399869384e-05,
      "loss": 0.5773,
      "step": 14590
    },
    {
      "epoch": 3.53,
      "grad_norm": 0.33911919593811035,
      "learning_rate": 1.092516046706833e-05,
      "loss": 0.5882,
      "step": 14600
    },
    {
      "epoch": 3.53,
      "grad_norm": 0.2976870834827423,
      "learning_rate": 1.0892212364668646e-05,
      "loss": 0.594,
      "step": 14610
    },
    {
      "epoch": 3.53,
      "grad_norm": 0.30538877844810486,
      "learning_rate": 1.0859300176546502e-05,
      "loss": 0.5762,
      "step": 14620
    },
    {
      "epoch": 3.54,
      "grad_norm": 0.2815379202365875,
      "learning_rate": 1.0826423986486666e-05,
      "loss": 0.5885,
      "step": 14630
    },
    {
      "epoch": 3.54,
      "grad_norm": 0.21779607236385345,
      "learning_rate": 1.0793583878182236e-05,
      "loss": 0.5976,
      "step": 14640
    },
    {
      "epoch": 3.54,
      "grad_norm": 0.5334603190422058,
      "learning_rate": 1.0760779935234469e-05,
      "loss": 0.5805,
      "step": 14650
    },
    {
      "epoch": 3.54,
      "grad_norm": 0.2569158971309662,
      "learning_rate": 1.0728012241152558e-05,
      "loss": 0.5743,
      "step": 14660
    },
    {
      "epoch": 3.55,
      "grad_norm": 0.23342689871788025,
      "learning_rate": 1.0695280879353406e-05,
      "loss": 0.5792,
      "step": 14670
    },
    {
      "epoch": 3.55,
      "grad_norm": 0.22866085171699524,
      "learning_rate": 1.066258593316144e-05,
      "loss": 0.5812,
      "step": 14680
    },
    {
      "epoch": 3.55,
      "grad_norm": 0.25173404812812805,
      "learning_rate": 1.0629927485808371e-05,
      "loss": 0.5651,
      "step": 14690
    },
    {
      "epoch": 3.55,
      "grad_norm": 0.2520875930786133,
      "learning_rate": 1.0597305620433e-05,
      "loss": 0.5925,
      "step": 14700
    },
    {
      "epoch": 3.55,
      "grad_norm": 0.2767471969127655,
      "learning_rate": 1.0564720420081007e-05,
      "loss": 0.6022,
      "step": 14710
    },
    {
      "epoch": 3.56,
      "grad_norm": 0.25810670852661133,
      "learning_rate": 1.0532171967704718e-05,
      "loss": 0.5805,
      "step": 14720
    },
    {
      "epoch": 3.56,
      "grad_norm": 0.3143045902252197,
      "learning_rate": 1.0499660346162924e-05,
      "loss": 0.6024,
      "step": 14730
    },
    {
      "epoch": 3.56,
      "grad_norm": 0.24170958995819092,
      "learning_rate": 1.0467185638220648e-05,
      "loss": 0.5847,
      "step": 14740
    },
    {
      "epoch": 3.56,
      "grad_norm": 0.22440150380134583,
      "learning_rate": 1.0434747926548943e-05,
      "loss": 0.5754,
      "step": 14750
    },
    {
      "epoch": 3.57,
      "grad_norm": 0.3215000331401825,
      "learning_rate": 1.0402347293724687e-05,
      "loss": 0.5925,
      "step": 14760
    },
    {
      "epoch": 3.57,
      "grad_norm": 0.23354408144950867,
      "learning_rate": 1.0369983822230356e-05,
      "loss": 0.5953,
      "step": 14770
    },
    {
      "epoch": 3.57,
      "grad_norm": 0.29122018814086914,
      "learning_rate": 1.0337657594453828e-05,
      "loss": 0.5995,
      "step": 14780
    },
    {
      "epoch": 3.57,
      "grad_norm": 0.3059288561344147,
      "learning_rate": 1.0305368692688174e-05,
      "loss": 0.5848,
      "step": 14790
    },
    {
      "epoch": 3.58,
      "grad_norm": 0.28843456506729126,
      "learning_rate": 1.0273117199131439e-05,
      "loss": 0.5822,
      "step": 14800
    },
    {
      "epoch": 3.58,
      "grad_norm": 0.3051983118057251,
      "learning_rate": 1.0240903195886437e-05,
      "loss": 0.5738,
      "step": 14810
    },
    {
      "epoch": 3.58,
      "grad_norm": 0.29712456464767456,
      "learning_rate": 1.0208726764960544e-05,
      "loss": 0.564,
      "step": 14820
    },
    {
      "epoch": 3.58,
      "grad_norm": 0.2549232840538025,
      "learning_rate": 1.0176587988265492e-05,
      "loss": 0.5588,
      "step": 14830
    },
    {
      "epoch": 3.59,
      "grad_norm": 0.29335421323776245,
      "learning_rate": 1.0144486947617152e-05,
      "loss": 0.5834,
      "step": 14840
    },
    {
      "epoch": 3.59,
      "grad_norm": 0.28511741757392883,
      "learning_rate": 1.0112423724735328e-05,
      "loss": 0.5969,
      "step": 14850
    },
    {
      "epoch": 3.59,
      "grad_norm": 0.26911091804504395,
      "learning_rate": 1.008039840124356e-05,
      "loss": 0.5845,
      "step": 14860
    },
    {
      "epoch": 3.59,
      "grad_norm": 0.23659975826740265,
      "learning_rate": 1.0048411058668894e-05,
      "loss": 0.5677,
      "step": 14870
    },
    {
      "epoch": 3.6,
      "grad_norm": 0.28330573439598083,
      "learning_rate": 1.0016461778441701e-05,
      "loss": 0.5861,
      "step": 14880
    },
    {
      "epoch": 3.6,
      "grad_norm": 0.2796567380428314,
      "learning_rate": 9.98455064189545e-06,
      "loss": 0.5745,
      "step": 14890
    },
    {
      "epoch": 3.6,
      "grad_norm": 0.28617990016937256,
      "learning_rate": 9.952677730266505e-06,
      "loss": 0.5893,
      "step": 14900
    },
    {
      "epoch": 3.6,
      "grad_norm": 0.25992870330810547,
      "learning_rate": 9.920843124693927e-06,
      "loss": 0.5923,
      "step": 14910
    },
    {
      "epoch": 3.61,
      "grad_norm": 0.32537975907325745,
      "learning_rate": 9.889046906219255e-06,
      "loss": 0.5799,
      "step": 14920
    },
    {
      "epoch": 3.61,
      "grad_norm": 0.3050180673599243,
      "learning_rate": 9.857289155786306e-06,
      "loss": 0.6,
      "step": 14930
    },
    {
      "epoch": 3.61,
      "grad_norm": 0.30784061551094055,
      "learning_rate": 9.82556995424097e-06,
      "loss": 0.591,
      "step": 14940
    },
    {
      "epoch": 3.61,
      "grad_norm": 0.29528340697288513,
      "learning_rate": 9.793889382331017e-06,
      "loss": 0.5743,
      "step": 14950
    },
    {
      "epoch": 3.62,
      "grad_norm": 0.22412078082561493,
      "learning_rate": 9.762247520705855e-06,
      "loss": 0.5812,
      "step": 14960
    },
    {
      "epoch": 3.62,
      "grad_norm": 0.2795025408267975,
      "learning_rate": 9.73064444991636e-06,
      "loss": 0.5836,
      "step": 14970
    },
    {
      "epoch": 3.62,
      "grad_norm": 0.2767314314842224,
      "learning_rate": 9.699080250414649e-06,
      "loss": 0.5802,
      "step": 14980
    },
    {
      "epoch": 3.62,
      "grad_norm": 0.32638996839523315,
      "learning_rate": 9.667555002553893e-06,
      "loss": 0.5749,
      "step": 14990
    },
    {
      "epoch": 3.62,
      "grad_norm": 0.26170921325683594,
      "learning_rate": 9.636068786588098e-06,
      "loss": 0.5933,
      "step": 15000
    },
    {
      "epoch": 3.62,
      "eval_loss": 0.6281116008758545,
      "eval_runtime": 67.2955,
      "eval_samples_per_second": 29.72,
      "eval_steps_per_second": 0.936,
      "step": 15000
    },
    {
      "epoch": 3.63,
      "grad_norm": 0.33843374252319336,
      "learning_rate": 9.60462168267191e-06,
      "loss": 0.6,
      "step": 15010
    },
    {
      "epoch": 3.63,
      "grad_norm": 0.26133570075035095,
      "learning_rate": 9.573213770860415e-06,
      "loss": 0.5586,
      "step": 15020
    },
    {
      "epoch": 3.63,
      "grad_norm": 0.2854001224040985,
      "learning_rate": 9.541845131108898e-06,
      "loss": 0.5921,
      "step": 15030
    },
    {
      "epoch": 3.63,
      "grad_norm": 0.2590722143650055,
      "learning_rate": 9.510515843272696e-06,
      "loss": 0.5583,
      "step": 15040
    },
    {
      "epoch": 3.64,
      "grad_norm": 0.28067246079444885,
      "learning_rate": 9.479225987106966e-06,
      "loss": 0.5833,
      "step": 15050
    },
    {
      "epoch": 3.64,
      "grad_norm": 0.2886711359024048,
      "learning_rate": 9.447975642266474e-06,
      "loss": 0.6027,
      "step": 15060
    },
    {
      "epoch": 3.64,
      "grad_norm": 0.2666650712490082,
      "learning_rate": 9.416764888305408e-06,
      "loss": 0.5906,
      "step": 15070
    },
    {
      "epoch": 3.64,
      "grad_norm": 0.28997379541397095,
      "learning_rate": 9.385593804677154e-06,
      "loss": 0.5905,
      "step": 15080
    },
    {
      "epoch": 3.65,
      "grad_norm": 0.35694199800491333,
      "learning_rate": 9.354462470734143e-06,
      "loss": 0.5904,
      "step": 15090
    },
    {
      "epoch": 3.65,
      "grad_norm": 0.29599496722221375,
      "learning_rate": 9.323370965727585e-06,
      "loss": 0.5971,
      "step": 15100
    },
    {
      "epoch": 3.65,
      "grad_norm": 0.2581184208393097,
      "learning_rate": 9.292319368807306e-06,
      "loss": 0.5988,
      "step": 15110
    },
    {
      "epoch": 3.65,
      "grad_norm": 0.26546022295951843,
      "learning_rate": 9.261307759021536e-06,
      "loss": 0.6109,
      "step": 15120
    },
    {
      "epoch": 3.66,
      "grad_norm": 0.2711322009563446,
      "learning_rate": 9.230336215316714e-06,
      "loss": 0.5809,
      "step": 15130
    },
    {
      "epoch": 3.66,
      "grad_norm": 0.47347956895828247,
      "learning_rate": 9.199404816537282e-06,
      "loss": 0.5703,
      "step": 15140
    },
    {
      "epoch": 3.66,
      "grad_norm": 0.28618571162223816,
      "learning_rate": 9.16851364142548e-06,
      "loss": 0.5756,
      "step": 15150
    },
    {
      "epoch": 3.66,
      "grad_norm": 0.23602980375289917,
      "learning_rate": 9.137662768621152e-06,
      "loss": 0.5825,
      "step": 15160
    },
    {
      "epoch": 3.67,
      "grad_norm": 0.28499022126197815,
      "learning_rate": 9.106852276661548e-06,
      "loss": 0.6218,
      "step": 15170
    },
    {
      "epoch": 3.67,
      "grad_norm": 0.24658203125,
      "learning_rate": 9.076082243981115e-06,
      "loss": 0.5873,
      "step": 15180
    },
    {
      "epoch": 3.67,
      "grad_norm": 0.26833969354629517,
      "learning_rate": 9.045352748911309e-06,
      "loss": 0.5773,
      "step": 15190
    },
    {
      "epoch": 3.67,
      "grad_norm": 0.2373725175857544,
      "learning_rate": 9.014663869680385e-06,
      "loss": 0.5727,
      "step": 15200
    },
    {
      "epoch": 3.68,
      "grad_norm": 0.28897523880004883,
      "learning_rate": 8.9840156844132e-06,
      "loss": 0.5838,
      "step": 15210
    },
    {
      "epoch": 3.68,
      "grad_norm": 0.2560650408267975,
      "learning_rate": 8.953408271131016e-06,
      "loss": 0.5655,
      "step": 15220
    },
    {
      "epoch": 3.68,
      "grad_norm": 0.3112213611602783,
      "learning_rate": 8.92284170775131e-06,
      "loss": 0.5905,
      "step": 15230
    },
    {
      "epoch": 3.68,
      "grad_norm": 0.24210262298583984,
      "learning_rate": 8.892316072087562e-06,
      "loss": 0.5687,
      "step": 15240
    },
    {
      "epoch": 3.69,
      "grad_norm": 0.2916972041130066,
      "learning_rate": 8.861831441849054e-06,
      "loss": 0.607,
      "step": 15250
    },
    {
      "epoch": 3.69,
      "grad_norm": 0.28263208270072937,
      "learning_rate": 8.831387894640694e-06,
      "loss": 0.5783,
      "step": 15260
    },
    {
      "epoch": 3.69,
      "grad_norm": 0.3277551829814911,
      "learning_rate": 8.800985507962794e-06,
      "loss": 0.5811,
      "step": 15270
    },
    {
      "epoch": 3.69,
      "grad_norm": 0.23930184543132782,
      "learning_rate": 8.770624359210894e-06,
      "loss": 0.5888,
      "step": 15280
    },
    {
      "epoch": 3.7,
      "grad_norm": 0.23394325375556946,
      "learning_rate": 8.740304525675541e-06,
      "loss": 0.583,
      "step": 15290
    },
    {
      "epoch": 3.7,
      "grad_norm": 0.26787030696868896,
      "learning_rate": 8.710026084542117e-06,
      "loss": 0.5751,
      "step": 15300
    },
    {
      "epoch": 3.7,
      "grad_norm": 0.2613442838191986,
      "learning_rate": 8.679789112890624e-06,
      "loss": 0.5668,
      "step": 15310
    },
    {
      "epoch": 3.7,
      "grad_norm": 0.23228463530540466,
      "learning_rate": 8.649593687695498e-06,
      "loss": 0.5987,
      "step": 15320
    },
    {
      "epoch": 3.7,
      "grad_norm": 0.24326367676258087,
      "learning_rate": 8.619439885825414e-06,
      "loss": 0.5697,
      "step": 15330
    },
    {
      "epoch": 3.71,
      "grad_norm": 0.3632679879665375,
      "learning_rate": 8.589327784043078e-06,
      "loss": 0.5861,
      "step": 15340
    },
    {
      "epoch": 3.71,
      "grad_norm": 0.24837125837802887,
      "learning_rate": 8.559257459005044e-06,
      "loss": 0.5795,
      "step": 15350
    },
    {
      "epoch": 3.71,
      "grad_norm": 0.2757419943809509,
      "learning_rate": 8.529228987261514e-06,
      "loss": 0.6007,
      "step": 15360
    },
    {
      "epoch": 3.71,
      "grad_norm": 0.31828778982162476,
      "learning_rate": 8.499242445256145e-06,
      "loss": 0.572,
      "step": 15370
    },
    {
      "epoch": 3.72,
      "grad_norm": 0.3094756007194519,
      "learning_rate": 8.469297909325855e-06,
      "loss": 0.574,
      "step": 15380
    },
    {
      "epoch": 3.72,
      "grad_norm": 0.27057239413261414,
      "learning_rate": 8.43939545570062e-06,
      "loss": 0.5675,
      "step": 15390
    },
    {
      "epoch": 3.72,
      "grad_norm": 0.2364446073770523,
      "learning_rate": 8.409535160503284e-06,
      "loss": 0.5931,
      "step": 15400
    },
    {
      "epoch": 3.72,
      "grad_norm": 0.29072749614715576,
      "learning_rate": 8.379717099749394e-06,
      "loss": 0.5932,
      "step": 15410
    },
    {
      "epoch": 3.73,
      "grad_norm": 0.24592497944831848,
      "learning_rate": 8.349941349346952e-06,
      "loss": 0.5749,
      "step": 15420
    },
    {
      "epoch": 3.73,
      "grad_norm": 0.276470422744751,
      "learning_rate": 8.32020798509626e-06,
      "loss": 0.5911,
      "step": 15430
    },
    {
      "epoch": 3.73,
      "grad_norm": 0.36373811960220337,
      "learning_rate": 8.290517082689717e-06,
      "loss": 0.5989,
      "step": 15440
    },
    {
      "epoch": 3.73,
      "grad_norm": 0.2627997398376465,
      "learning_rate": 8.260868717711628e-06,
      "loss": 0.5996,
      "step": 15450
    },
    {
      "epoch": 3.74,
      "grad_norm": 0.30971255898475647,
      "learning_rate": 8.231262965638009e-06,
      "loss": 0.5988,
      "step": 15460
    },
    {
      "epoch": 3.74,
      "grad_norm": 0.29322072863578796,
      "learning_rate": 8.2016999018364e-06,
      "loss": 0.5727,
      "step": 15470
    },
    {
      "epoch": 3.74,
      "grad_norm": 0.345941960811615,
      "learning_rate": 8.172179601565656e-06,
      "loss": 0.6011,
      "step": 15480
    },
    {
      "epoch": 3.74,
      "grad_norm": 0.4636392593383789,
      "learning_rate": 8.142702139975786e-06,
      "loss": 0.6416,
      "step": 15490
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.2717054486274719,
      "learning_rate": 8.113267592107735e-06,
      "loss": 0.565,
      "step": 15500
    },
    {
      "epoch": 3.75,
      "eval_loss": 0.626787006855011,
      "eval_runtime": 67.2655,
      "eval_samples_per_second": 29.733,
      "eval_steps_per_second": 0.937,
      "step": 15500
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.23242239654064178,
      "learning_rate": 8.083876032893208e-06,
      "loss": 0.5883,
      "step": 15510
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.23897923529148102,
      "learning_rate": 8.05452753715446e-06,
      "loss": 0.5833,
      "step": 15520
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.29634857177734375,
      "learning_rate": 8.025222179604133e-06,
      "loss": 0.6021,
      "step": 15530
    },
    {
      "epoch": 3.76,
      "grad_norm": 0.25519996881484985,
      "learning_rate": 7.995960034845038e-06,
      "loss": 0.5629,
      "step": 15540
    },
    {
      "epoch": 3.76,
      "grad_norm": 0.3006519675254822,
      "learning_rate": 7.966741177370005e-06,
      "loss": 0.551,
      "step": 15550
    },
    {
      "epoch": 3.76,
      "grad_norm": 0.2494523823261261,
      "learning_rate": 7.937565681561643e-06,
      "loss": 0.6096,
      "step": 15560
    },
    {
      "epoch": 3.76,
      "grad_norm": 0.28488144278526306,
      "learning_rate": 7.90843362169218e-06,
      "loss": 0.5816,
      "step": 15570
    },
    {
      "epoch": 3.77,
      "grad_norm": 0.2764374017715454,
      "learning_rate": 7.879345071923274e-06,
      "loss": 0.5933,
      "step": 15580
    },
    {
      "epoch": 3.77,
      "grad_norm": 0.34692734479904175,
      "learning_rate": 7.85030010630582e-06,
      "loss": 0.5806,
      "step": 15590
    },
    {
      "epoch": 3.77,
      "grad_norm": 0.2502826750278473,
      "learning_rate": 7.82129879877975e-06,
      "loss": 0.5894,
      "step": 15600
    },
    {
      "epoch": 3.77,
      "grad_norm": 0.27148333191871643,
      "learning_rate": 7.792341223173863e-06,
      "loss": 0.5806,
      "step": 15610
    },
    {
      "epoch": 3.77,
      "grad_norm": 0.2680436670780182,
      "learning_rate": 7.763427453205632e-06,
      "loss": 0.567,
      "step": 15620
    },
    {
      "epoch": 3.78,
      "grad_norm": 0.27015039324760437,
      "learning_rate": 7.73455756248101e-06,
      "loss": 0.5792,
      "step": 15630
    },
    {
      "epoch": 3.78,
      "grad_norm": 0.3366507291793823,
      "learning_rate": 7.705731624494244e-06,
      "loss": 0.5806,
      "step": 15640
    },
    {
      "epoch": 3.78,
      "grad_norm": 0.261496901512146,
      "learning_rate": 7.676949712627699e-06,
      "loss": 0.5841,
      "step": 15650
    },
    {
      "epoch": 3.78,
      "grad_norm": 0.23413531482219696,
      "learning_rate": 7.64821190015165e-06,
      "loss": 0.587,
      "step": 15660
    },
    {
      "epoch": 3.79,
      "grad_norm": 0.3006034791469574,
      "learning_rate": 7.619518260224115e-06,
      "loss": 0.617,
      "step": 15670
    },
    {
      "epoch": 3.79,
      "grad_norm": 0.3840187191963196,
      "learning_rate": 7.590868865890666e-06,
      "loss": 0.5906,
      "step": 15680
    },
    {
      "epoch": 3.79,
      "grad_norm": 0.33729347586631775,
      "learning_rate": 7.562263790084229e-06,
      "loss": 0.5841,
      "step": 15690
    },
    {
      "epoch": 3.79,
      "grad_norm": 0.2606573700904846,
      "learning_rate": 7.533703105624915e-06,
      "loss": 0.5996,
      "step": 15700
    },
    {
      "epoch": 3.8,
      "grad_norm": 0.252876877784729,
      "learning_rate": 7.505186885219817e-06,
      "loss": 0.5687,
      "step": 15710
    },
    {
      "epoch": 3.8,
      "grad_norm": 0.24049414694309235,
      "learning_rate": 7.4767152014628615e-06,
      "loss": 0.5716,
      "step": 15720
    },
    {
      "epoch": 3.8,
      "grad_norm": 0.27121061086654663,
      "learning_rate": 7.448288126834571e-06,
      "loss": 0.5862,
      "step": 15730
    },
    {
      "epoch": 3.8,
      "grad_norm": 0.218089759349823,
      "learning_rate": 7.419905733701915e-06,
      "loss": 0.5783,
      "step": 15740
    },
    {
      "epoch": 3.81,
      "grad_norm": 0.24585284292697906,
      "learning_rate": 7.391568094318127e-06,
      "loss": 0.5764,
      "step": 15750
    },
    {
      "epoch": 3.81,
      "grad_norm": 0.2675657868385315,
      "learning_rate": 7.363275280822487e-06,
      "loss": 0.5772,
      "step": 15760
    },
    {
      "epoch": 3.81,
      "grad_norm": 0.23988156020641327,
      "learning_rate": 7.33502736524018e-06,
      "loss": 0.5769,
      "step": 15770
    },
    {
      "epoch": 3.81,
      "grad_norm": 0.2960077226161957,
      "learning_rate": 7.3068244194820946e-06,
      "loss": 0.5942,
      "step": 15780
    },
    {
      "epoch": 3.82,
      "grad_norm": 0.27055710554122925,
      "learning_rate": 7.2786665153446315e-06,
      "loss": 0.5907,
      "step": 15790
    },
    {
      "epoch": 3.82,
      "grad_norm": 0.4000237286090851,
      "learning_rate": 7.250553724509532e-06,
      "loss": 0.5729,
      "step": 15800
    },
    {
      "epoch": 3.82,
      "grad_norm": 0.23245370388031006,
      "learning_rate": 7.222486118543695e-06,
      "loss": 0.5827,
      "step": 15810
    },
    {
      "epoch": 3.82,
      "grad_norm": 0.29327449202537537,
      "learning_rate": 7.194463768898988e-06,
      "loss": 0.61,
      "step": 15820
    },
    {
      "epoch": 3.83,
      "grad_norm": 0.36466142535209656,
      "learning_rate": 7.166486746912068e-06,
      "loss": 0.5692,
      "step": 15830
    },
    {
      "epoch": 3.83,
      "grad_norm": 0.26802125573158264,
      "learning_rate": 7.138555123804208e-06,
      "loss": 0.5797,
      "step": 15840
    },
    {
      "epoch": 3.83,
      "grad_norm": 0.21287329494953156,
      "learning_rate": 7.110668970681092e-06,
      "loss": 0.5932,
      "step": 15850
    },
    {
      "epoch": 3.83,
      "grad_norm": 0.27712956070899963,
      "learning_rate": 7.082828358532684e-06,
      "loss": 0.5529,
      "step": 15860
    },
    {
      "epoch": 3.84,
      "grad_norm": 0.27336981892585754,
      "learning_rate": 7.055033358232979e-06,
      "loss": 0.5775,
      "step": 15870
    },
    {
      "epoch": 3.84,
      "grad_norm": 0.2662564218044281,
      "learning_rate": 7.027284040539878e-06,
      "loss": 0.5868,
      "step": 15880
    },
    {
      "epoch": 3.84,
      "grad_norm": 0.23860903084278107,
      "learning_rate": 6.99958047609498e-06,
      "loss": 0.5968,
      "step": 15890
    },
    {
      "epoch": 3.84,
      "grad_norm": 0.32224351167678833,
      "learning_rate": 6.971922735423408e-06,
      "loss": 0.5897,
      "step": 15900
    },
    {
      "epoch": 3.84,
      "grad_norm": 0.27302199602127075,
      "learning_rate": 6.944310888933636e-06,
      "loss": 0.6104,
      "step": 15910
    },
    {
      "epoch": 3.85,
      "grad_norm": 0.2775563597679138,
      "learning_rate": 6.916745006917305e-06,
      "loss": 0.6005,
      "step": 15920
    },
    {
      "epoch": 3.85,
      "grad_norm": 0.2953178882598877,
      "learning_rate": 6.8892251595490395e-06,
      "loss": 0.5822,
      "step": 15930
    },
    {
      "epoch": 3.85,
      "grad_norm": 0.2623131573200226,
      "learning_rate": 6.86175141688628e-06,
      "loss": 0.5948,
      "step": 15940
    },
    {
      "epoch": 3.85,
      "grad_norm": 0.3807394504547119,
      "learning_rate": 6.83432384886909e-06,
      "loss": 0.5917,
      "step": 15950
    },
    {
      "epoch": 3.86,
      "grad_norm": 0.2432747483253479,
      "learning_rate": 6.8069425253199896e-06,
      "loss": 0.5891,
      "step": 15960
    },
    {
      "epoch": 3.86,
      "grad_norm": 0.3248459994792938,
      "learning_rate": 6.779607515943778e-06,
      "loss": 0.5789,
      "step": 15970
    },
    {
      "epoch": 3.86,
      "grad_norm": 0.26909446716308594,
      "learning_rate": 6.752318890327345e-06,
      "loss": 0.5724,
      "step": 15980
    },
    {
      "epoch": 3.86,
      "grad_norm": 0.28674009442329407,
      "learning_rate": 6.725076717939507e-06,
      "loss": 0.6008,
      "step": 15990
    },
    {
      "epoch": 3.87,
      "grad_norm": 0.2869904339313507,
      "learning_rate": 6.697881068130821e-06,
      "loss": 0.5884,
      "step": 16000
    },
    {
      "epoch": 3.87,
      "eval_loss": 0.6267058849334717,
      "eval_runtime": 67.187,
      "eval_samples_per_second": 29.768,
      "eval_steps_per_second": 0.938,
      "step": 16000
    },
    {
      "epoch": 3.87,
      "grad_norm": 0.28786352276802063,
      "learning_rate": 6.670732010133412e-06,
      "loss": 0.6029,
      "step": 16010
    },
    {
      "epoch": 3.87,
      "grad_norm": 0.236564040184021,
      "learning_rate": 6.643629613060795e-06,
      "loss": 0.5998,
      "step": 16020
    },
    {
      "epoch": 3.87,
      "grad_norm": 0.34414246678352356,
      "learning_rate": 6.6165739459077e-06,
      "loss": 0.5833,
      "step": 16030
    },
    {
      "epoch": 3.88,
      "grad_norm": 0.25065818428993225,
      "learning_rate": 6.5895650775499e-06,
      "loss": 0.5889,
      "step": 16040
    },
    {
      "epoch": 3.88,
      "grad_norm": 0.2535193860530853,
      "learning_rate": 6.562603076744028e-06,
      "loss": 0.5838,
      "step": 16050
    },
    {
      "epoch": 3.88,
      "grad_norm": 0.2876463234424591,
      "learning_rate": 6.535688012127406e-06,
      "loss": 0.6073,
      "step": 16060
    },
    {
      "epoch": 3.88,
      "grad_norm": 0.28166264295578003,
      "learning_rate": 6.508819952217874e-06,
      "loss": 0.5833,
      "step": 16070
    },
    {
      "epoch": 3.89,
      "grad_norm": 0.3278672397136688,
      "learning_rate": 6.481998965413607e-06,
      "loss": 0.5992,
      "step": 16080
    },
    {
      "epoch": 3.89,
      "grad_norm": 0.250623494386673,
      "learning_rate": 6.455225119992949e-06,
      "loss": 0.5702,
      "step": 16090
    },
    {
      "epoch": 3.89,
      "grad_norm": 0.31878694891929626,
      "learning_rate": 6.428498484114235e-06,
      "loss": 0.5905,
      "step": 16100
    },
    {
      "epoch": 3.89,
      "grad_norm": 0.3199431002140045,
      "learning_rate": 6.40181912581562e-06,
      "loss": 0.5803,
      "step": 16110
    },
    {
      "epoch": 3.9,
      "grad_norm": 0.27698907256126404,
      "learning_rate": 6.375187113014899e-06,
      "loss": 0.5988,
      "step": 16120
    },
    {
      "epoch": 3.9,
      "grad_norm": 0.30214032530784607,
      "learning_rate": 6.348602513509344e-06,
      "loss": 0.5935,
      "step": 16130
    },
    {
      "epoch": 3.9,
      "grad_norm": 0.2558833658695221,
      "learning_rate": 6.322065394975527e-06,
      "loss": 0.5926,
      "step": 16140
    },
    {
      "epoch": 3.9,
      "grad_norm": 0.3078318238258362,
      "learning_rate": 6.295575824969146e-06,
      "loss": 0.5773,
      "step": 16150
    },
    {
      "epoch": 3.91,
      "grad_norm": 0.32137954235076904,
      "learning_rate": 6.269133870924853e-06,
      "loss": 0.6126,
      "step": 16160
    },
    {
      "epoch": 3.91,
      "grad_norm": 0.2796170115470886,
      "learning_rate": 6.24273960015608e-06,
      "loss": 0.5871,
      "step": 16170
    },
    {
      "epoch": 3.91,
      "grad_norm": 0.3090955317020416,
      "learning_rate": 6.216393079854888e-06,
      "loss": 0.5726,
      "step": 16180
    },
    {
      "epoch": 3.91,
      "grad_norm": 0.284445583820343,
      "learning_rate": 6.190094377091762e-06,
      "loss": 0.5795,
      "step": 16190
    },
    {
      "epoch": 3.91,
      "grad_norm": 0.32952529191970825,
      "learning_rate": 6.163843558815466e-06,
      "loss": 0.606,
      "step": 16200
    },
    {
      "epoch": 3.92,
      "grad_norm": 0.23674383759498596,
      "learning_rate": 6.137640691852859e-06,
      "loss": 0.6043,
      "step": 16210
    },
    {
      "epoch": 3.92,
      "grad_norm": 0.2780101001262665,
      "learning_rate": 6.111485842908732e-06,
      "loss": 0.5905,
      "step": 16220
    },
    {
      "epoch": 3.92,
      "grad_norm": 0.2556968033313751,
      "learning_rate": 6.085379078565637e-06,
      "loss": 0.5911,
      "step": 16230
    },
    {
      "epoch": 3.92,
      "grad_norm": 0.31288766860961914,
      "learning_rate": 6.059320465283722e-06,
      "loss": 0.5941,
      "step": 16240
    },
    {
      "epoch": 3.93,
      "grad_norm": 0.32016289234161377,
      "learning_rate": 6.033310069400552e-06,
      "loss": 0.589,
      "step": 16250
    },
    {
      "epoch": 3.93,
      "grad_norm": 0.24031643569469452,
      "learning_rate": 6.007347957130932e-06,
      "loss": 0.5834,
      "step": 16260
    },
    {
      "epoch": 3.93,
      "grad_norm": 0.3101363778114319,
      "learning_rate": 5.98143419456677e-06,
      "loss": 0.5961,
      "step": 16270
    },
    {
      "epoch": 3.93,
      "grad_norm": 0.3162418305873871,
      "learning_rate": 5.955568847676887e-06,
      "loss": 0.5783,
      "step": 16280
    },
    {
      "epoch": 3.94,
      "grad_norm": 0.29130181670188904,
      "learning_rate": 5.929751982306841e-06,
      "loss": 0.6111,
      "step": 16290
    },
    {
      "epoch": 3.94,
      "grad_norm": 0.2524430751800537,
      "learning_rate": 5.903983664178783e-06,
      "loss": 0.5893,
      "step": 16300
    },
    {
      "epoch": 3.94,
      "grad_norm": 0.2583165764808655,
      "learning_rate": 5.878263958891262e-06,
      "loss": 0.59,
      "step": 16310
    },
    {
      "epoch": 3.94,
      "grad_norm": 0.30341488122940063,
      "learning_rate": 5.8525929319190955e-06,
      "loss": 0.6,
      "step": 16320
    },
    {
      "epoch": 3.95,
      "grad_norm": 0.3176259696483612,
      "learning_rate": 5.826970648613164e-06,
      "loss": 0.554,
      "step": 16330
    },
    {
      "epoch": 3.95,
      "grad_norm": 0.3036240339279175,
      "learning_rate": 5.801397174200257e-06,
      "loss": 0.5785,
      "step": 16340
    },
    {
      "epoch": 3.95,
      "grad_norm": 0.2815450429916382,
      "learning_rate": 5.775872573782926e-06,
      "loss": 0.566,
      "step": 16350
    },
    {
      "epoch": 3.95,
      "grad_norm": 0.3002425730228424,
      "learning_rate": 5.750396912339287e-06,
      "loss": 0.5899,
      "step": 16360
    },
    {
      "epoch": 3.96,
      "grad_norm": 0.2630454897880554,
      "learning_rate": 5.7249702547228885e-06,
      "loss": 0.6125,
      "step": 16370
    },
    {
      "epoch": 3.96,
      "grad_norm": 0.32869815826416016,
      "learning_rate": 5.699592665662517e-06,
      "loss": 0.5733,
      "step": 16380
    },
    {
      "epoch": 3.96,
      "grad_norm": 0.3272031545639038,
      "learning_rate": 5.674264209762051e-06,
      "loss": 0.5931,
      "step": 16390
    },
    {
      "epoch": 3.96,
      "grad_norm": 0.25317224860191345,
      "learning_rate": 5.648984951500289e-06,
      "loss": 0.5771,
      "step": 16400
    },
    {
      "epoch": 3.97,
      "grad_norm": 0.3111300766468048,
      "learning_rate": 5.6237549552307864e-06,
      "loss": 0.5695,
      "step": 16410
    },
    {
      "epoch": 3.97,
      "grad_norm": 0.2719234526157379,
      "learning_rate": 5.598574285181693e-06,
      "loss": 0.5884,
      "step": 16420
    },
    {
      "epoch": 3.97,
      "grad_norm": 0.2526122033596039,
      "learning_rate": 5.573443005455587e-06,
      "loss": 0.5536,
      "step": 16430
    },
    {
      "epoch": 3.97,
      "grad_norm": 0.3456088900566101,
      "learning_rate": 5.548361180029313e-06,
      "loss": 0.595,
      "step": 16440
    },
    {
      "epoch": 3.98,
      "grad_norm": 0.292865514755249,
      "learning_rate": 5.523328872753822e-06,
      "loss": 0.5794,
      "step": 16450
    },
    {
      "epoch": 3.98,
      "grad_norm": 0.2841857969760895,
      "learning_rate": 5.498346147354008e-06,
      "loss": 0.5806,
      "step": 16460
    },
    {
      "epoch": 3.98,
      "grad_norm": 0.24631422758102417,
      "learning_rate": 5.473413067428537e-06,
      "loss": 0.61,
      "step": 16470
    },
    {
      "epoch": 3.98,
      "grad_norm": 0.28376200795173645,
      "learning_rate": 5.448529696449697e-06,
      "loss": 0.5785,
      "step": 16480
    },
    {
      "epoch": 3.99,
      "grad_norm": 0.2724916338920593,
      "learning_rate": 5.423696097763234e-06,
      "loss": 0.6048,
      "step": 16490
    },
    {
      "epoch": 3.99,
      "grad_norm": 0.33017095923423767,
      "learning_rate": 5.398912334588186e-06,
      "loss": 0.5809,
      "step": 16500
    },
    {
      "epoch": 3.99,
      "eval_loss": 0.6265645623207092,
      "eval_runtime": 67.1906,
      "eval_samples_per_second": 29.766,
      "eval_steps_per_second": 0.938,
      "step": 16500
    },
    {
      "epoch": 3.99,
      "grad_norm": 0.23196405172348022,
      "learning_rate": 5.374178470016719e-06,
      "loss": 0.5879,
      "step": 16510
    },
    {
      "epoch": 3.99,
      "grad_norm": 0.2956921458244324,
      "learning_rate": 5.34949456701398e-06,
      "loss": 0.5713,
      "step": 16520
    },
    {
      "epoch": 3.99,
      "grad_norm": 0.271476149559021,
      "learning_rate": 5.324860688417924e-06,
      "loss": 0.5779,
      "step": 16530
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.22633521258831024,
      "learning_rate": 5.30027689693916e-06,
      "loss": 0.5803,
      "step": 16540
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.397101491689682,
      "learning_rate": 5.275743255160792e-06,
      "loss": 0.5718,
      "step": 16550
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.2787100374698639,
      "learning_rate": 5.251259825538254e-06,
      "loss": 0.5845,
      "step": 16560
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.24712499976158142,
      "learning_rate": 5.226826670399157e-06,
      "loss": 0.604,
      "step": 16570
    },
    {
      "epoch": 4.01,
      "grad_norm": 0.26938384771347046,
      "learning_rate": 5.202443851943126e-06,
      "loss": 0.568,
      "step": 16580
    },
    {
      "epoch": 4.01,
      "grad_norm": 0.34945255517959595,
      "learning_rate": 5.1781114322416445e-06,
      "loss": 0.5979,
      "step": 16590
    },
    {
      "epoch": 4.01,
      "grad_norm": 0.27607110142707825,
      "learning_rate": 5.153829473237898e-06,
      "loss": 0.5587,
      "step": 16600
    },
    {
      "epoch": 4.01,
      "grad_norm": 0.2416377067565918,
      "learning_rate": 5.129598036746608e-06,
      "loss": 0.5914,
      "step": 16610
    },
    {
      "epoch": 4.02,
      "grad_norm": 0.3223896324634552,
      "learning_rate": 5.105417184453884e-06,
      "loss": 0.5792,
      "step": 16620
    },
    {
      "epoch": 4.02,
      "grad_norm": 0.26762667298316956,
      "learning_rate": 5.08128697791706e-06,
      "loss": 0.5602,
      "step": 16630
    },
    {
      "epoch": 4.02,
      "grad_norm": 0.24923330545425415,
      "learning_rate": 5.0572074785645575e-06,
      "loss": 0.5911,
      "step": 16640
    },
    {
      "epoch": 4.02,
      "grad_norm": 0.2844977378845215,
      "learning_rate": 5.033178747695688e-06,
      "loss": 0.618,
      "step": 16650
    },
    {
      "epoch": 4.03,
      "grad_norm": 0.2711544334888458,
      "learning_rate": 5.0092008464805355e-06,
      "loss": 0.6036,
      "step": 16660
    },
    {
      "epoch": 4.03,
      "grad_norm": 0.3091453015804291,
      "learning_rate": 4.985273835959784e-06,
      "loss": 0.5894,
      "step": 16670
    },
    {
      "epoch": 4.03,
      "grad_norm": 0.3095366954803467,
      "learning_rate": 4.961397777044563e-06,
      "loss": 0.5825,
      "step": 16680
    },
    {
      "epoch": 4.03,
      "grad_norm": 0.3563496172428131,
      "learning_rate": 4.937572730516294e-06,
      "loss": 0.5767,
      "step": 16690
    },
    {
      "epoch": 4.04,
      "grad_norm": 0.24569082260131836,
      "learning_rate": 4.913798757026538e-06,
      "loss": 0.5684,
      "step": 16700
    },
    {
      "epoch": 4.04,
      "grad_norm": 0.346018522977829,
      "learning_rate": 4.890075917096843e-06,
      "loss": 0.5919,
      "step": 16710
    },
    {
      "epoch": 4.04,
      "grad_norm": 0.23421183228492737,
      "learning_rate": 4.866404271118574e-06,
      "loss": 0.572,
      "step": 16720
    },
    {
      "epoch": 4.04,
      "grad_norm": 0.2508642077445984,
      "learning_rate": 4.842783879352786e-06,
      "loss": 0.5764,
      "step": 16730
    },
    {
      "epoch": 4.05,
      "grad_norm": 0.3773396909236908,
      "learning_rate": 4.8192148019300474e-06,
      "loss": 0.5764,
      "step": 16740
    },
    {
      "epoch": 4.05,
      "grad_norm": 0.32823699712753296,
      "learning_rate": 4.795697098850291e-06,
      "loss": 0.5711,
      "step": 16750
    },
    {
      "epoch": 4.05,
      "grad_norm": 0.24900804460048676,
      "learning_rate": 4.7722308299826715e-06,
      "loss": 0.5825,
      "step": 16760
    },
    {
      "epoch": 4.05,
      "grad_norm": 0.3153171241283417,
      "learning_rate": 4.748816055065405e-06,
      "loss": 0.5643,
      "step": 16770
    },
    {
      "epoch": 4.06,
      "grad_norm": 0.2724463939666748,
      "learning_rate": 4.725452833705632e-06,
      "loss": 0.6106,
      "step": 16780
    },
    {
      "epoch": 4.06,
      "grad_norm": 0.39892250299453735,
      "learning_rate": 4.702141225379234e-06,
      "loss": 0.5816,
      "step": 16790
    },
    {
      "epoch": 4.06,
      "grad_norm": 0.33181869983673096,
      "learning_rate": 4.678881289430712e-06,
      "loss": 0.5901,
      "step": 16800
    },
    {
      "epoch": 4.06,
      "grad_norm": 0.32668840885162354,
      "learning_rate": 4.6556730850730175e-06,
      "loss": 0.5777,
      "step": 16810
    },
    {
      "epoch": 4.06,
      "grad_norm": 0.31462836265563965,
      "learning_rate": 4.632516671387416e-06,
      "loss": 0.5736,
      "step": 16820
    },
    {
      "epoch": 4.07,
      "grad_norm": 0.31752124428749084,
      "learning_rate": 4.609412107323325e-06,
      "loss": 0.5975,
      "step": 16830
    },
    {
      "epoch": 4.07,
      "grad_norm": 0.32917672395706177,
      "learning_rate": 4.586359451698164e-06,
      "loss": 0.5866,
      "step": 16840
    },
    {
      "epoch": 4.07,
      "grad_norm": 0.2878721058368683,
      "learning_rate": 4.5633587631972145e-06,
      "loss": 0.5887,
      "step": 16850
    },
    {
      "epoch": 4.07,
      "grad_norm": 0.2430756390094757,
      "learning_rate": 4.540410100373468e-06,
      "loss": 0.5681,
      "step": 16860
    },
    {
      "epoch": 4.08,
      "grad_norm": 0.2615582048892975,
      "learning_rate": 4.5175135216474675e-06,
      "loss": 0.567,
      "step": 16870
    },
    {
      "epoch": 4.08,
      "grad_norm": 0.2856476306915283,
      "learning_rate": 4.494669085307165e-06,
      "loss": 0.592,
      "step": 16880
    },
    {
      "epoch": 4.08,
      "grad_norm": 0.2746788561344147,
      "learning_rate": 4.471876849507781e-06,
      "loss": 0.5963,
      "step": 16890
    },
    {
      "epoch": 4.08,
      "grad_norm": 0.2498600333929062,
      "learning_rate": 4.4491368722716395e-06,
      "loss": 0.5999,
      "step": 16900
    },
    {
      "epoch": 4.09,
      "grad_norm": 0.2452557235956192,
      "learning_rate": 4.426449211488032e-06,
      "loss": 0.576,
      "step": 16910
    },
    {
      "epoch": 4.09,
      "grad_norm": 0.27374720573425293,
      "learning_rate": 4.403813924913075e-06,
      "loss": 0.5595,
      "step": 16920
    },
    {
      "epoch": 4.09,
      "grad_norm": 0.3165563941001892,
      "learning_rate": 4.381231070169547e-06,
      "loss": 0.5854,
      "step": 16930
    },
    {
      "epoch": 4.09,
      "grad_norm": 0.3147006630897522,
      "learning_rate": 4.358700704746746e-06,
      "loss": 0.5852,
      "step": 16940
    },
    {
      "epoch": 4.1,
      "grad_norm": 0.3542732894420624,
      "learning_rate": 4.336222886000369e-06,
      "loss": 0.6078,
      "step": 16950
    },
    {
      "epoch": 4.1,
      "grad_norm": 0.4165996015071869,
      "learning_rate": 4.313797671152325e-06,
      "loss": 0.5607,
      "step": 16960
    },
    {
      "epoch": 4.1,
      "grad_norm": 0.28804096579551697,
      "learning_rate": 4.291425117290615e-06,
      "loss": 0.5853,
      "step": 16970
    },
    {
      "epoch": 4.1,
      "grad_norm": 0.2615216374397278,
      "learning_rate": 4.269105281369176e-06,
      "loss": 0.5687,
      "step": 16980
    },
    {
      "epoch": 4.11,
      "grad_norm": 0.2917400300502777,
      "learning_rate": 4.246838220207758e-06,
      "loss": 0.5795,
      "step": 16990
    },
    {
      "epoch": 4.11,
      "grad_norm": 0.22331537306308746,
      "learning_rate": 4.22462399049173e-06,
      "loss": 0.5618,
      "step": 17000
    },
    {
      "epoch": 4.11,
      "eval_loss": 0.6270723938941956,
      "eval_runtime": 67.167,
      "eval_samples_per_second": 29.777,
      "eval_steps_per_second": 0.938,
      "step": 17000
    },
    {
      "epoch": 4.11,
      "grad_norm": 0.30317845940589905,
      "learning_rate": 4.2024626487719986e-06,
      "loss": 0.5599,
      "step": 17010
    },
    {
      "epoch": 4.11,
      "grad_norm": 0.2718835771083832,
      "learning_rate": 4.180354251464816e-06,
      "loss": 0.5647,
      "step": 17020
    },
    {
      "epoch": 4.12,
      "grad_norm": 0.3022977113723755,
      "learning_rate": 4.158298854851661e-06,
      "loss": 0.5945,
      "step": 17030
    },
    {
      "epoch": 4.12,
      "grad_norm": 0.2802698314189911,
      "learning_rate": 4.136296515079088e-06,
      "loss": 0.599,
      "step": 17040
    },
    {
      "epoch": 4.12,
      "grad_norm": 0.2764960825443268,
      "learning_rate": 4.114347288158579e-06,
      "loss": 0.5843,
      "step": 17050
    },
    {
      "epoch": 4.12,
      "grad_norm": 0.25794321298599243,
      "learning_rate": 4.09245122996641e-06,
      "loss": 0.5922,
      "step": 17060
    },
    {
      "epoch": 4.13,
      "grad_norm": 0.3252635896205902,
      "learning_rate": 4.0706083962435025e-06,
      "loss": 0.6008,
      "step": 17070
    },
    {
      "epoch": 4.13,
      "grad_norm": 0.3116281032562256,
      "learning_rate": 4.048818842595284e-06,
      "loss": 0.5831,
      "step": 17080
    },
    {
      "epoch": 4.13,
      "grad_norm": 0.28487542271614075,
      "learning_rate": 4.02708262449156e-06,
      "loss": 0.5805,
      "step": 17090
    },
    {
      "epoch": 4.13,
      "grad_norm": 0.26027926802635193,
      "learning_rate": 4.00539979726634e-06,
      "loss": 0.5826,
      "step": 17100
    },
    {
      "epoch": 4.13,
      "grad_norm": 0.2977902591228485,
      "learning_rate": 3.983770416117727e-06,
      "loss": 0.6149,
      "step": 17110
    },
    {
      "epoch": 4.14,
      "grad_norm": 0.38094112277030945,
      "learning_rate": 3.962194536107761e-06,
      "loss": 0.5692,
      "step": 17120
    },
    {
      "epoch": 4.14,
      "grad_norm": 0.39290517568588257,
      "learning_rate": 3.940672212162285e-06,
      "loss": 0.5806,
      "step": 17130
    },
    {
      "epoch": 4.14,
      "grad_norm": 0.31928128004074097,
      "learning_rate": 3.919203499070812e-06,
      "loss": 0.5899,
      "step": 17140
    },
    {
      "epoch": 4.14,
      "grad_norm": 0.27730268239974976,
      "learning_rate": 3.897788451486359e-06,
      "loss": 0.5876,
      "step": 17150
    },
    {
      "epoch": 4.15,
      "grad_norm": 0.26537054777145386,
      "learning_rate": 3.876427123925347e-06,
      "loss": 0.6003,
      "step": 17160
    },
    {
      "epoch": 4.15,
      "grad_norm": 0.3302718698978424,
      "learning_rate": 3.855119570767429e-06,
      "loss": 0.6146,
      "step": 17170
    },
    {
      "epoch": 4.15,
      "grad_norm": 0.272029310464859,
      "learning_rate": 3.833865846255369e-06,
      "loss": 0.584,
      "step": 17180
    },
    {
      "epoch": 4.15,
      "grad_norm": 0.27373218536376953,
      "learning_rate": 3.812666004494897e-06,
      "loss": 0.5689,
      "step": 17190
    },
    {
      "epoch": 4.16,
      "grad_norm": 0.3247639536857605,
      "learning_rate": 3.791520099454576e-06,
      "loss": 0.5798,
      "step": 17200
    },
    {
      "epoch": 4.16,
      "grad_norm": 0.37551939487457275,
      "learning_rate": 3.7704281849656575e-06,
      "loss": 0.6041,
      "step": 17210
    },
    {
      "epoch": 4.16,
      "grad_norm": 0.2781738042831421,
      "learning_rate": 3.749390314721954e-06,
      "loss": 0.5951,
      "step": 17220
    },
    {
      "epoch": 4.16,
      "grad_norm": 0.31663358211517334,
      "learning_rate": 3.7284065422796945e-06,
      "loss": 0.5938,
      "step": 17230
    },
    {
      "epoch": 4.17,
      "grad_norm": 0.2771975100040436,
      "learning_rate": 3.70747692105739e-06,
      "loss": 0.547,
      "step": 17240
    },
    {
      "epoch": 4.17,
      "grad_norm": 0.42413491010665894,
      "learning_rate": 3.686601504335699e-06,
      "loss": 0.5971,
      "step": 17250
    },
    {
      "epoch": 4.17,
      "grad_norm": 0.2650463581085205,
      "learning_rate": 3.665780345257294e-06,
      "loss": 0.5877,
      "step": 17260
    },
    {
      "epoch": 4.17,
      "grad_norm": 0.28535643219947815,
      "learning_rate": 3.645013496826724e-06,
      "loss": 0.5601,
      "step": 17270
    },
    {
      "epoch": 4.18,
      "grad_norm": 0.30466383695602417,
      "learning_rate": 3.624301011910269e-06,
      "loss": 0.6031,
      "step": 17280
    },
    {
      "epoch": 4.18,
      "grad_norm": 0.2553720474243164,
      "learning_rate": 3.6036429432358327e-06,
      "loss": 0.5744,
      "step": 17290
    },
    {
      "epoch": 4.18,
      "grad_norm": 0.23872804641723633,
      "learning_rate": 3.5830393433927716e-06,
      "loss": 0.5894,
      "step": 17300
    },
    {
      "epoch": 4.18,
      "grad_norm": 0.2140451967716217,
      "learning_rate": 3.562490264831797e-06,
      "loss": 0.5768,
      "step": 17310
    },
    {
      "epoch": 4.19,
      "grad_norm": 0.2986193895339966,
      "learning_rate": 3.541995759864819e-06,
      "loss": 0.5831,
      "step": 17320
    },
    {
      "epoch": 4.19,
      "grad_norm": 0.3172728419303894,
      "learning_rate": 3.5215558806648145e-06,
      "loss": 0.5853,
      "step": 17330
    },
    {
      "epoch": 4.19,
      "grad_norm": 0.2729806900024414,
      "learning_rate": 3.5011706792657054e-06,
      "loss": 0.6025,
      "step": 17340
    },
    {
      "epoch": 4.19,
      "grad_norm": 0.27669742703437805,
      "learning_rate": 3.4808402075622194e-06,
      "loss": 0.5621,
      "step": 17350
    },
    {
      "epoch": 4.2,
      "grad_norm": 0.31161707639694214,
      "learning_rate": 3.4605645173097556e-06,
      "loss": 0.5997,
      "step": 17360
    },
    {
      "epoch": 4.2,
      "grad_norm": 0.30737054347991943,
      "learning_rate": 3.4403436601242567e-06,
      "loss": 0.604,
      "step": 17370
    },
    {
      "epoch": 4.2,
      "grad_norm": 0.31299006938934326,
      "learning_rate": 3.420177687482082e-06,
      "loss": 0.6099,
      "step": 17380
    },
    {
      "epoch": 4.2,
      "grad_norm": 0.23331117630004883,
      "learning_rate": 3.4000666507198605e-06,
      "loss": 0.5802,
      "step": 17390
    },
    {
      "epoch": 4.2,
      "grad_norm": 0.3147907853126526,
      "learning_rate": 3.3800106010343722e-06,
      "loss": 0.6073,
      "step": 17400
    },
    {
      "epoch": 4.21,
      "grad_norm": 0.24627135694026947,
      "learning_rate": 3.3600095894824322e-06,
      "loss": 0.565,
      "step": 17410
    },
    {
      "epoch": 4.21,
      "grad_norm": 0.24128040671348572,
      "learning_rate": 3.340063666980725e-06,
      "loss": 0.5737,
      "step": 17420
    },
    {
      "epoch": 4.21,
      "grad_norm": 0.3306833505630493,
      "learning_rate": 3.3201728843057055e-06,
      "loss": 0.5753,
      "step": 17430
    },
    {
      "epoch": 4.21,
      "grad_norm": 0.2970528304576874,
      "learning_rate": 3.3003372920934554e-06,
      "loss": 0.5977,
      "step": 17440
    },
    {
      "epoch": 4.22,
      "grad_norm": 0.4337664544582367,
      "learning_rate": 3.280556940839555e-06,
      "loss": 0.5905,
      "step": 17450
    },
    {
      "epoch": 4.22,
      "grad_norm": 0.2858019173145294,
      "learning_rate": 3.260831880898962e-06,
      "loss": 0.6008,
      "step": 17460
    },
    {
      "epoch": 4.22,
      "grad_norm": 0.2653151750564575,
      "learning_rate": 3.2411621624858777e-06,
      "loss": 0.5702,
      "step": 17470
    },
    {
      "epoch": 4.22,
      "grad_norm": 0.2580778896808624,
      "learning_rate": 3.2215478356736245e-06,
      "loss": 0.5906,
      "step": 17480
    },
    {
      "epoch": 4.23,
      "grad_norm": 0.3023454546928406,
      "learning_rate": 3.2019889503944937e-06,
      "loss": 0.6072,
      "step": 17490
    },
    {
      "epoch": 4.23,
      "grad_norm": 0.2779148817062378,
      "learning_rate": 3.1824855564396653e-06,
      "loss": 0.5749,
      "step": 17500
    },
    {
      "epoch": 4.23,
      "eval_loss": 0.627366840839386,
      "eval_runtime": 67.4291,
      "eval_samples_per_second": 29.661,
      "eval_steps_per_second": 0.934,
      "step": 17500
    },
    {
      "epoch": 4.23,
      "grad_norm": 0.2548415958881378,
      "learning_rate": 3.163037703459043e-06,
      "loss": 0.5809,
      "step": 17510
    },
    {
      "epoch": 4.23,
      "grad_norm": 0.2873731553554535,
      "learning_rate": 3.143645440961135e-06,
      "loss": 0.5751,
      "step": 17520
    },
    {
      "epoch": 4.24,
      "grad_norm": 0.2768021821975708,
      "learning_rate": 3.124308818312943e-06,
      "loss": 0.6002,
      "step": 17530
    },
    {
      "epoch": 4.24,
      "grad_norm": 0.31484872102737427,
      "learning_rate": 3.105027884739811e-06,
      "loss": 0.5677,
      "step": 17540
    },
    {
      "epoch": 4.24,
      "grad_norm": 0.28398895263671875,
      "learning_rate": 3.08580268932534e-06,
      "loss": 0.5653,
      "step": 17550
    },
    {
      "epoch": 4.24,
      "grad_norm": 0.2831834852695465,
      "learning_rate": 3.066633281011211e-06,
      "loss": 0.5676,
      "step": 17560
    },
    {
      "epoch": 4.25,
      "grad_norm": 0.3742017447948456,
      "learning_rate": 3.047519708597102e-06,
      "loss": 0.5776,
      "step": 17570
    },
    {
      "epoch": 4.25,
      "grad_norm": 0.26823127269744873,
      "learning_rate": 3.0284620207405467e-06,
      "loss": 0.5751,
      "step": 17580
    },
    {
      "epoch": 4.25,
      "grad_norm": 0.26558831334114075,
      "learning_rate": 3.009460265956812e-06,
      "loss": 0.5983,
      "step": 17590
    },
    {
      "epoch": 4.25,
      "grad_norm": 0.3036767244338989,
      "learning_rate": 2.9905144926187763e-06,
      "loss": 0.5681,
      "step": 17600
    },
    {
      "epoch": 4.26,
      "grad_norm": 0.28618937730789185,
      "learning_rate": 2.9716247489568055e-06,
      "loss": 0.5617,
      "step": 17610
    },
    {
      "epoch": 4.26,
      "grad_norm": 0.2805090844631195,
      "learning_rate": 2.9527910830586315e-06,
      "loss": 0.5787,
      "step": 17620
    },
    {
      "epoch": 4.26,
      "grad_norm": 0.30515792965888977,
      "learning_rate": 2.934013542869227e-06,
      "loss": 0.5848,
      "step": 17630
    },
    {
      "epoch": 4.26,
      "grad_norm": 0.35140377283096313,
      "learning_rate": 2.917161783691147e-06,
      "loss": 0.5993,
      "step": 17640
    },
    {
      "epoch": 4.27,
      "grad_norm": 0.47522976994514465,
      "learning_rate": 2.898491013924984e-06,
      "loss": 0.6019,
      "step": 17650
    },
    {
      "epoch": 4.27,
      "grad_norm": 0.27884793281555176,
      "learning_rate": 2.8798765080995963e-06,
      "loss": 0.5927,
      "step": 17660
    },
    {
      "epoch": 4.27,
      "grad_norm": 0.24438752233982086,
      "learning_rate": 2.861318313602038e-06,
      "loss": 0.5938,
      "step": 17670
    },
    {
      "epoch": 4.27,
      "grad_norm": 0.30406540632247925,
      "learning_rate": 2.8428164776760107e-06,
      "loss": 0.5548,
      "step": 17680
    },
    {
      "epoch": 4.28,
      "grad_norm": 0.3240637183189392,
      "learning_rate": 2.8243710474217535e-06,
      "loss": 0.5629,
      "step": 17690
    },
    {
      "epoch": 4.28,
      "grad_norm": 0.3322964310646057,
      "learning_rate": 2.805982069795901e-06,
      "loss": 0.6262,
      "step": 17700
    },
    {
      "epoch": 4.28,
      "grad_norm": 0.3318009078502655,
      "learning_rate": 2.787649591611377e-06,
      "loss": 0.5942,
      "step": 17710
    },
    {
      "epoch": 4.28,
      "grad_norm": 0.31209197640419006,
      "learning_rate": 2.7693736595372886e-06,
      "loss": 0.5721,
      "step": 17720
    },
    {
      "epoch": 4.28,
      "grad_norm": 0.6039661169052124,
      "learning_rate": 2.7511543200987784e-06,
      "loss": 0.5464,
      "step": 17730
    },
    {
      "epoch": 4.29,
      "grad_norm": 0.2902863621711731,
      "learning_rate": 2.732991619676925e-06,
      "loss": 0.5751,
      "step": 17740
    },
    {
      "epoch": 4.29,
      "grad_norm": 0.2638190686702728,
      "learning_rate": 2.714885604508621e-06,
      "loss": 0.588,
      "step": 17750
    },
    {
      "epoch": 4.29,
      "grad_norm": 0.3784666061401367,
      "learning_rate": 2.696836320686455e-06,
      "loss": 0.5983,
      "step": 17760
    },
    {
      "epoch": 4.29,
      "grad_norm": 0.26594826579093933,
      "learning_rate": 2.678843814158602e-06,
      "loss": 0.5533,
      "step": 17770
    },
    {
      "epoch": 4.3,
      "grad_norm": 0.25600722432136536,
      "learning_rate": 2.6609081307286786e-06,
      "loss": 0.564,
      "step": 17780
    },
    {
      "epoch": 4.3,
      "grad_norm": 0.3318096101284027,
      "learning_rate": 2.6430293160556644e-06,
      "loss": 0.5839,
      "step": 17790
    },
    {
      "epoch": 4.3,
      "grad_norm": 0.2860864996910095,
      "learning_rate": 2.625207415653763e-06,
      "loss": 0.6055,
      "step": 17800
    },
    {
      "epoch": 4.3,
      "grad_norm": 0.31288084387779236,
      "learning_rate": 2.607442474892291e-06,
      "loss": 0.551,
      "step": 17810
    },
    {
      "epoch": 4.31,
      "grad_norm": 0.2489456981420517,
      "learning_rate": 2.5897345389955586e-06,
      "loss": 0.5671,
      "step": 17820
    },
    {
      "epoch": 4.31,
      "grad_norm": 0.29000183939933777,
      "learning_rate": 2.5720836530427673e-06,
      "loss": 0.5718,
      "step": 17830
    },
    {
      "epoch": 4.31,
      "grad_norm": 0.3467647433280945,
      "learning_rate": 2.5544898619678794e-06,
      "loss": 0.6026,
      "step": 17840
    },
    {
      "epoch": 4.31,
      "grad_norm": 0.24173252284526825,
      "learning_rate": 2.5369532105595033e-06,
      "loss": 0.5755,
      "step": 17850
    },
    {
      "epoch": 4.32,
      "grad_norm": 0.27934640645980835,
      "learning_rate": 2.5194737434608066e-06,
      "loss": 0.611,
      "step": 17860
    },
    {
      "epoch": 4.32,
      "grad_norm": 0.22576917707920074,
      "learning_rate": 2.50205150516937e-06,
      "loss": 0.5526,
      "step": 17870
    },
    {
      "epoch": 4.32,
      "grad_norm": 0.27904582023620605,
      "learning_rate": 2.484686540037082e-06,
      "loss": 0.5997,
      "step": 17880
    },
    {
      "epoch": 4.32,
      "grad_norm": 0.34738051891326904,
      "learning_rate": 2.467378892270039e-06,
      "loss": 0.5606,
      "step": 17890
    },
    {
      "epoch": 4.33,
      "grad_norm": 0.25818055868148804,
      "learning_rate": 2.45012860592842e-06,
      "loss": 0.5699,
      "step": 17900
    },
    {
      "epoch": 4.33,
      "grad_norm": 0.2724223732948303,
      "learning_rate": 2.43293572492638e-06,
      "loss": 0.557,
      "step": 17910
    },
    {
      "epoch": 4.33,
      "grad_norm": 0.24570220708847046,
      "learning_rate": 2.415800293031939e-06,
      "loss": 0.5724,
      "step": 17920
    },
    {
      "epoch": 4.33,
      "grad_norm": 0.2438490241765976,
      "learning_rate": 2.398722353866864e-06,
      "loss": 0.5703,
      "step": 17930
    },
    {
      "epoch": 4.34,
      "grad_norm": 0.28835931420326233,
      "learning_rate": 2.3817019509065646e-06,
      "loss": 0.593,
      "step": 17940
    },
    {
      "epoch": 4.34,
      "grad_norm": 0.26829794049263,
      "learning_rate": 2.3647391274799786e-06,
      "loss": 0.5691,
      "step": 17950
    },
    {
      "epoch": 4.34,
      "grad_norm": 0.3007825016975403,
      "learning_rate": 2.3478339267694687e-06,
      "loss": 0.5585,
      "step": 17960
    },
    {
      "epoch": 4.34,
      "grad_norm": 0.26816418766975403,
      "learning_rate": 2.330986391810702e-06,
      "loss": 0.5726,
      "step": 17970
    },
    {
      "epoch": 4.35,
      "grad_norm": 0.27209803462028503,
      "learning_rate": 2.314196565492546e-06,
      "loss": 0.5896,
      "step": 17980
    },
    {
      "epoch": 4.35,
      "grad_norm": 0.27638113498687744,
      "learning_rate": 2.2974644905569608e-06,
      "loss": 0.5715,
      "step": 17990
    },
    {
      "epoch": 4.35,
      "grad_norm": 0.25953102111816406,
      "learning_rate": 2.2807902095988847e-06,
      "loss": 0.577,
      "step": 18000
    },
    {
      "epoch": 4.35,
      "eval_loss": 0.6267811059951782,
      "eval_runtime": 67.1155,
      "eval_samples_per_second": 29.799,
      "eval_steps_per_second": 0.939,
      "step": 18000
    },
    {
      "epoch": 4.35,
      "grad_norm": 0.26673948764801025,
      "learning_rate": 2.2641737650661383e-06,
      "loss": 0.5802,
      "step": 18010
    },
    {
      "epoch": 4.35,
      "grad_norm": 0.2779202163219452,
      "learning_rate": 2.2476151992592955e-06,
      "loss": 0.5804,
      "step": 18020
    },
    {
      "epoch": 4.36,
      "grad_norm": 0.27395954728126526,
      "learning_rate": 2.2311145543315954e-06,
      "loss": 0.5602,
      "step": 18030
    },
    {
      "epoch": 4.36,
      "grad_norm": 0.30662351846694946,
      "learning_rate": 2.2146718722888316e-06,
      "loss": 0.5573,
      "step": 18040
    },
    {
      "epoch": 4.36,
      "grad_norm": 0.32358744740486145,
      "learning_rate": 2.198287194989229e-06,
      "loss": 0.5615,
      "step": 18050
    },
    {
      "epoch": 4.36,
      "grad_norm": 0.2596151530742645,
      "learning_rate": 2.181960564143359e-06,
      "loss": 0.5814,
      "step": 18060
    },
    {
      "epoch": 4.37,
      "grad_norm": 0.2692429721355438,
      "learning_rate": 2.1656920213140227e-06,
      "loss": 0.579,
      "step": 18070
    },
    {
      "epoch": 4.37,
      "grad_norm": 0.26572030782699585,
      "learning_rate": 2.149481607916143e-06,
      "loss": 0.5661,
      "step": 18080
    },
    {
      "epoch": 4.37,
      "grad_norm": 0.265151709318161,
      "learning_rate": 2.133329365216666e-06,
      "loss": 0.5656,
      "step": 18090
    },
    {
      "epoch": 4.37,
      "grad_norm": 0.26654693484306335,
      "learning_rate": 2.1172353343344496e-06,
      "loss": 0.5778,
      "step": 18100
    },
    {
      "epoch": 4.38,
      "grad_norm": 0.25736740231513977,
      "learning_rate": 2.101199556240163e-06,
      "loss": 0.5723,
      "step": 18110
    },
    {
      "epoch": 4.38,
      "grad_norm": 0.2923225164413452,
      "learning_rate": 2.0852220717561794e-06,
      "loss": 0.5718,
      "step": 18120
    },
    {
      "epoch": 4.38,
      "grad_norm": 0.29171547293663025,
      "learning_rate": 2.0693029215564737e-06,
      "loss": 0.5682,
      "step": 18130
    },
    {
      "epoch": 4.38,
      "grad_norm": 0.23074455559253693,
      "learning_rate": 2.0534421461665202e-06,
      "loss": 0.5807,
      "step": 18140
    },
    {
      "epoch": 4.39,
      "grad_norm": 0.23554468154907227,
      "learning_rate": 2.0376397859631887e-06,
      "loss": 0.5888,
      "step": 18150
    },
    {
      "epoch": 4.39,
      "grad_norm": 0.3034656345844269,
      "learning_rate": 2.0218958811746403e-06,
      "loss": 0.5741,
      "step": 18160
    },
    {
      "epoch": 4.39,
      "grad_norm": 0.28760990500450134,
      "learning_rate": 2.006210471880221e-06,
      "loss": 0.5948,
      "step": 18170
    },
    {
      "epoch": 4.39,
      "grad_norm": 0.2663377821445465,
      "learning_rate": 1.9905835980103758e-06,
      "loss": 0.5779,
      "step": 18180
    },
    {
      "epoch": 4.4,
      "grad_norm": 0.284000426530838,
      "learning_rate": 1.975015299346528e-06,
      "loss": 0.5692,
      "step": 18190
    },
    {
      "epoch": 4.4,
      "grad_norm": 0.28214696049690247,
      "learning_rate": 1.9595056155209847e-06,
      "loss": 0.5836,
      "step": 18200
    },
    {
      "epoch": 4.4,
      "grad_norm": 0.3069036602973938,
      "learning_rate": 1.9440545860168356e-06,
      "loss": 0.5614,
      "step": 18210
    },
    {
      "epoch": 4.4,
      "grad_norm": 0.30323630571365356,
      "learning_rate": 1.928662250167862e-06,
      "loss": 0.5747,
      "step": 18220
    },
    {
      "epoch": 4.41,
      "grad_norm": 0.3179965615272522,
      "learning_rate": 1.9133286471584216e-06,
      "loss": 0.5971,
      "step": 18230
    },
    {
      "epoch": 4.41,
      "grad_norm": 0.2338808923959732,
      "learning_rate": 1.8980538160233568e-06,
      "loss": 0.582,
      "step": 18240
    },
    {
      "epoch": 4.41,
      "grad_norm": 0.26593130826950073,
      "learning_rate": 1.882837795647896e-06,
      "loss": 0.5819,
      "step": 18250
    },
    {
      "epoch": 4.41,
      "grad_norm": 0.35234713554382324,
      "learning_rate": 1.867680624767551e-06,
      "loss": 0.6131,
      "step": 18260
    },
    {
      "epoch": 4.42,
      "grad_norm": 0.27029749751091003,
      "learning_rate": 1.8525823419680204e-06,
      "loss": 0.5761,
      "step": 18270
    },
    {
      "epoch": 4.42,
      "grad_norm": 0.2861502170562744,
      "learning_rate": 1.8375429856850895e-06,
      "loss": 0.5739,
      "step": 18280
    },
    {
      "epoch": 4.42,
      "grad_norm": 0.3253205418586731,
      "learning_rate": 1.8225625942045356e-06,
      "loss": 0.5903,
      "step": 18290
    },
    {
      "epoch": 4.42,
      "grad_norm": 0.2866848111152649,
      "learning_rate": 1.8076412056620307e-06,
      "loss": 0.5738,
      "step": 18300
    },
    {
      "epoch": 4.42,
      "grad_norm": 0.28978246450424194,
      "learning_rate": 1.7927788580430343e-06,
      "loss": 0.5904,
      "step": 18310
    },
    {
      "epoch": 4.43,
      "grad_norm": 0.3061251640319824,
      "learning_rate": 1.7779755891827232e-06,
      "loss": 0.5844,
      "step": 18320
    },
    {
      "epoch": 4.43,
      "grad_norm": 0.2831498086452484,
      "learning_rate": 1.7632314367658582e-06,
      "loss": 0.5815,
      "step": 18330
    },
    {
      "epoch": 4.43,
      "grad_norm": 0.3293684422969818,
      "learning_rate": 1.7485464383267175e-06,
      "loss": 0.542,
      "step": 18340
    },
    {
      "epoch": 4.43,
      "grad_norm": 0.27200835943222046,
      "learning_rate": 1.7339206312489864e-06,
      "loss": 0.5825,
      "step": 18350
    },
    {
      "epoch": 4.44,
      "grad_norm": 0.24985529482364655,
      "learning_rate": 1.7193540527656648e-06,
      "loss": 0.5983,
      "step": 18360
    },
    {
      "epoch": 4.44,
      "grad_norm": 0.24906408786773682,
      "learning_rate": 1.704846739958982e-06,
      "loss": 0.5791,
      "step": 18370
    },
    {
      "epoch": 4.44,
      "grad_norm": 1.195361852645874,
      "learning_rate": 1.6903987297602842e-06,
      "loss": 0.5862,
      "step": 18380
    },
    {
      "epoch": 4.44,
      "grad_norm": 0.32085666060447693,
      "learning_rate": 1.6760100589499616e-06,
      "loss": 0.6076,
      "step": 18390
    },
    {
      "epoch": 4.45,
      "grad_norm": 0.29969578981399536,
      "learning_rate": 1.6616807641573307e-06,
      "loss": 0.5862,
      "step": 18400
    },
    {
      "epoch": 4.45,
      "grad_norm": 0.28302955627441406,
      "learning_rate": 1.6474108818605644e-06,
      "loss": 0.5829,
      "step": 18410
    },
    {
      "epoch": 4.45,
      "grad_norm": 0.2879648506641388,
      "learning_rate": 1.6332004483865876e-06,
      "loss": 0.5908,
      "step": 18420
    },
    {
      "epoch": 4.45,
      "grad_norm": 0.2986491918563843,
      "learning_rate": 1.6190494999109795e-06,
      "loss": 0.5807,
      "step": 18430
    },
    {
      "epoch": 4.46,
      "grad_norm": 0.3248368203639984,
      "learning_rate": 1.604958072457896e-06,
      "loss": 0.576,
      "step": 18440
    },
    {
      "epoch": 4.46,
      "grad_norm": 0.2517332434654236,
      "learning_rate": 1.5909262018999644e-06,
      "loss": 0.579,
      "step": 18450
    },
    {
      "epoch": 4.46,
      "grad_norm": 0.30697351694107056,
      "learning_rate": 1.5769539239582021e-06,
      "loss": 0.5802,
      "step": 18460
    },
    {
      "epoch": 4.46,
      "grad_norm": 0.31047147512435913,
      "learning_rate": 1.5630412742019152e-06,
      "loss": 0.5758,
      "step": 18470
    },
    {
      "epoch": 4.47,
      "grad_norm": 0.25356003642082214,
      "learning_rate": 1.5491882880486197e-06,
      "loss": 0.5651,
      "step": 18480
    },
    {
      "epoch": 4.47,
      "grad_norm": 0.42373108863830566,
      "learning_rate": 1.5353950007639422e-06,
      "loss": 0.5903,
      "step": 18490
    },
    {
      "epoch": 4.47,
      "grad_norm": 0.2725335657596588,
      "learning_rate": 1.5216614474615366e-06,
      "loss": 0.5947,
      "step": 18500
    },
    {
      "epoch": 4.47,
      "eval_loss": 0.6266624927520752,
      "eval_runtime": 67.2798,
      "eval_samples_per_second": 29.727,
      "eval_steps_per_second": 0.936,
      "step": 18500
    },
    {
      "epoch": 4.47,
      "grad_norm": 0.24809718132019043,
      "learning_rate": 1.5079876631029944e-06,
      "loss": 0.5976,
      "step": 18510
    },
    {
      "epoch": 4.48,
      "grad_norm": 0.31159552931785583,
      "learning_rate": 1.4943736824977406e-06,
      "loss": 0.556,
      "step": 18520
    },
    {
      "epoch": 4.48,
      "grad_norm": 0.262481689453125,
      "learning_rate": 1.480819540302969e-06,
      "loss": 0.5769,
      "step": 18530
    },
    {
      "epoch": 4.48,
      "grad_norm": 0.2672281861305237,
      "learning_rate": 1.467325271023537e-06,
      "loss": 0.585,
      "step": 18540
    },
    {
      "epoch": 4.48,
      "grad_norm": 0.26489323377609253,
      "learning_rate": 1.4538909090118846e-06,
      "loss": 0.5761,
      "step": 18550
    },
    {
      "epoch": 4.49,
      "grad_norm": 0.261283814907074,
      "learning_rate": 1.4405164884679496e-06,
      "loss": 0.5758,
      "step": 18560
    },
    {
      "epoch": 4.49,
      "grad_norm": 0.2721174359321594,
      "learning_rate": 1.4272020434390658e-06,
      "loss": 0.5958,
      "step": 18570
    },
    {
      "epoch": 4.49,
      "grad_norm": 0.3530591130256653,
      "learning_rate": 1.4139476078199016e-06,
      "loss": 0.5998,
      "step": 18580
    },
    {
      "epoch": 4.49,
      "grad_norm": 0.30697134137153625,
      "learning_rate": 1.4007532153523462e-06,
      "loss": 0.5946,
      "step": 18590
    },
    {
      "epoch": 4.49,
      "grad_norm": 0.33676719665527344,
      "learning_rate": 1.3876188996254452e-06,
      "loss": 0.5844,
      "step": 18600
    },
    {
      "epoch": 4.5,
      "grad_norm": 0.27862054109573364,
      "learning_rate": 1.3745446940752998e-06,
      "loss": 0.5796,
      "step": 18610
    },
    {
      "epoch": 4.5,
      "grad_norm": 0.2915698289871216,
      "learning_rate": 1.361530631984989e-06,
      "loss": 0.5708,
      "step": 18620
    },
    {
      "epoch": 4.5,
      "grad_norm": 0.3161049485206604,
      "learning_rate": 1.348576746484495e-06,
      "loss": 0.583,
      "step": 18630
    },
    {
      "epoch": 4.5,
      "grad_norm": 0.308633029460907,
      "learning_rate": 1.3356830705505974e-06,
      "loss": 0.5805,
      "step": 18640
    },
    {
      "epoch": 4.51,
      "grad_norm": 0.33652573823928833,
      "learning_rate": 1.3228496370067983e-06,
      "loss": 0.5868,
      "step": 18650
    },
    {
      "epoch": 4.51,
      "grad_norm": 0.2793577015399933,
      "learning_rate": 1.3100764785232478e-06,
      "loss": 0.5708,
      "step": 18660
    },
    {
      "epoch": 4.51,
      "grad_norm": 0.3058006167411804,
      "learning_rate": 1.297363627616649e-06,
      "loss": 0.567,
      "step": 18670
    },
    {
      "epoch": 4.51,
      "grad_norm": 0.25099068880081177,
      "learning_rate": 1.2847111166501836e-06,
      "loss": 0.5727,
      "step": 18680
    },
    {
      "epoch": 4.52,
      "grad_norm": 0.3448689877986908,
      "learning_rate": 1.2721189778334169e-06,
      "loss": 0.5903,
      "step": 18690
    },
    {
      "epoch": 4.52,
      "grad_norm": 0.2721264064311981,
      "learning_rate": 1.259587243222235e-06,
      "loss": 0.5777,
      "step": 18700
    },
    {
      "epoch": 4.52,
      "grad_norm": 0.3201538324356079,
      "learning_rate": 1.247115944718749e-06,
      "loss": 0.5636,
      "step": 18710
    },
    {
      "epoch": 4.52,
      "grad_norm": 0.24066606163978577,
      "learning_rate": 1.2347051140712134e-06,
      "loss": 0.6149,
      "step": 18720
    },
    {
      "epoch": 4.53,
      "grad_norm": 0.3968261480331421,
      "learning_rate": 1.2223547828739523e-06,
      "loss": 0.5845,
      "step": 18730
    },
    {
      "epoch": 4.53,
      "grad_norm": 0.2825827896595001,
      "learning_rate": 1.2100649825672799e-06,
      "loss": 0.6046,
      "step": 18740
    },
    {
      "epoch": 4.53,
      "grad_norm": 0.5150123834609985,
      "learning_rate": 1.197835744437409e-06,
      "loss": 0.5717,
      "step": 18750
    },
    {
      "epoch": 4.53,
      "grad_norm": 0.2709062993526459,
      "learning_rate": 1.1856670996163832e-06,
      "loss": 0.5678,
      "step": 18760
    },
    {
      "epoch": 4.54,
      "grad_norm": 0.27491089701652527,
      "learning_rate": 1.1735590790819955e-06,
      "loss": 0.5617,
      "step": 18770
    },
    {
      "epoch": 4.54,
      "grad_norm": 0.3479272723197937,
      "learning_rate": 1.1615117136577008e-06,
      "loss": 0.5914,
      "step": 18780
    },
    {
      "epoch": 4.54,
      "grad_norm": 0.2816771864891052,
      "learning_rate": 1.1495250340125496e-06,
      "loss": 0.5996,
      "step": 18790
    },
    {
      "epoch": 4.54,
      "grad_norm": 0.29737162590026855,
      "learning_rate": 1.1375990706611024e-06,
      "loss": 0.579,
      "step": 18800
    },
    {
      "epoch": 4.55,
      "grad_norm": 0.257581889629364,
      "learning_rate": 1.1257338539633566e-06,
      "loss": 0.5777,
      "step": 18810
    },
    {
      "epoch": 4.55,
      "grad_norm": 0.28490132093429565,
      "learning_rate": 1.1139294141246593e-06,
      "loss": 0.5807,
      "step": 18820
    },
    {
      "epoch": 4.55,
      "grad_norm": 0.23307104408740997,
      "learning_rate": 1.1021857811956443e-06,
      "loss": 0.5869,
      "step": 18830
    },
    {
      "epoch": 4.55,
      "grad_norm": 0.3434465229511261,
      "learning_rate": 1.0905029850721449e-06,
      "loss": 0.5874,
      "step": 18840
    },
    {
      "epoch": 4.56,
      "grad_norm": 0.2812347710132599,
      "learning_rate": 1.0788810554951263e-06,
      "loss": 0.5555,
      "step": 18850
    },
    {
      "epoch": 4.56,
      "grad_norm": 0.3239988386631012,
      "learning_rate": 1.0673200220505998e-06,
      "loss": 0.5869,
      "step": 18860
    },
    {
      "epoch": 4.56,
      "grad_norm": 0.275582879781723,
      "learning_rate": 1.055819914169559e-06,
      "loss": 0.5695,
      "step": 18870
    },
    {
      "epoch": 4.56,
      "grad_norm": 0.34836065769195557,
      "learning_rate": 1.0443807611278945e-06,
      "loss": 0.5664,
      "step": 18880
    },
    {
      "epoch": 4.57,
      "grad_norm": 0.5555301904678345,
      "learning_rate": 1.0330025920463227e-06,
      "loss": 0.5678,
      "step": 18890
    },
    {
      "epoch": 4.57,
      "grad_norm": 0.2862575352191925,
      "learning_rate": 1.021685435890321e-06,
      "loss": 0.5684,
      "step": 18900
    },
    {
      "epoch": 4.57,
      "grad_norm": 0.32544222474098206,
      "learning_rate": 1.0104293214700328e-06,
      "loss": 0.5718,
      "step": 18910
    },
    {
      "epoch": 4.57,
      "grad_norm": 0.2881731688976288,
      "learning_rate": 9.992342774402198e-07,
      "loss": 0.5823,
      "step": 18920
    },
    {
      "epoch": 4.57,
      "grad_norm": 0.28005796670913696,
      "learning_rate": 9.88100332300171e-07,
      "loss": 0.5952,
      "step": 18930
    },
    {
      "epoch": 4.58,
      "grad_norm": 0.3245788514614105,
      "learning_rate": 9.770275143936274e-07,
      "loss": 0.6031,
      "step": 18940
    },
    {
      "epoch": 4.58,
      "grad_norm": 0.3854841887950897,
      "learning_rate": 9.660158519087386e-07,
      "loss": 0.5618,
      "step": 18950
    },
    {
      "epoch": 4.58,
      "grad_norm": 0.2594067454338074,
      "learning_rate": 9.55065372877953e-07,
      "loss": 0.5848,
      "step": 18960
    },
    {
      "epoch": 4.58,
      "grad_norm": 0.2725635766983032,
      "learning_rate": 9.441761051779685e-07,
      "loss": 0.5943,
      "step": 18970
    },
    {
      "epoch": 4.59,
      "grad_norm": 0.29124271869659424,
      "learning_rate": 9.333480765296582e-07,
      "loss": 0.5875,
      "step": 18980
    },
    {
      "epoch": 4.59,
      "grad_norm": 0.30971428751945496,
      "learning_rate": 9.225813144979972e-07,
      "loss": 0.5846,
      "step": 18990
    },
    {
      "epoch": 4.59,
      "grad_norm": 0.33653488755226135,
      "learning_rate": 9.118758464919968e-07,
      "loss": 0.5902,
      "step": 19000
    },
    {
      "epoch": 4.59,
      "eval_loss": 0.6267839074134827,
      "eval_runtime": 67.1305,
      "eval_samples_per_second": 29.793,
      "eval_steps_per_second": 0.938,
      "step": 19000
    },
    {
      "epoch": 4.59,
      "grad_norm": 0.29609283804893494,
      "learning_rate": 9.012316997646292e-07,
      "loss": 0.5613,
      "step": 19010
    },
    {
      "epoch": 4.6,
      "grad_norm": 0.2698591649532318,
      "learning_rate": 8.906489014127551e-07,
      "loss": 0.5893,
      "step": 19020
    },
    {
      "epoch": 4.6,
      "grad_norm": 0.2906891107559204,
      "learning_rate": 8.801274783770741e-07,
      "loss": 0.5603,
      "step": 19030
    },
    {
      "epoch": 4.6,
      "grad_norm": 0.8006889820098877,
      "learning_rate": 8.696674574420305e-07,
      "loss": 0.5906,
      "step": 19040
    },
    {
      "epoch": 4.6,
      "grad_norm": 0.29602134227752686,
      "learning_rate": 8.5926886523576e-07,
      "loss": 0.5847,
      "step": 19050
    },
    {
      "epoch": 4.61,
      "grad_norm": 0.3691641092300415,
      "learning_rate": 8.489317282300207e-07,
      "loss": 0.5696,
      "step": 19060
    },
    {
      "epoch": 4.61,
      "grad_norm": 0.2935909628868103,
      "learning_rate": 8.386560727401266e-07,
      "loss": 0.5794,
      "step": 19070
    },
    {
      "epoch": 4.61,
      "grad_norm": 0.3069007694721222,
      "learning_rate": 8.284419249248638e-07,
      "loss": 0.595,
      "step": 19080
    },
    {
      "epoch": 4.61,
      "grad_norm": 0.2680169939994812,
      "learning_rate": 8.182893107864603e-07,
      "loss": 0.5914,
      "step": 19090
    },
    {
      "epoch": 4.62,
      "grad_norm": 0.2621663212776184,
      "learning_rate": 8.081982561704809e-07,
      "loss": 0.5783,
      "step": 19100
    },
    {
      "epoch": 4.62,
      "grad_norm": 0.5907593965530396,
      "learning_rate": 7.981687867657794e-07,
      "loss": 0.5625,
      "step": 19110
    },
    {
      "epoch": 4.62,
      "grad_norm": 0.31429407000541687,
      "learning_rate": 7.882009281044378e-07,
      "loss": 0.5982,
      "step": 19120
    },
    {
      "epoch": 4.62,
      "grad_norm": 0.26952117681503296,
      "learning_rate": 7.782947055616912e-07,
      "loss": 0.5849,
      "step": 19130
    },
    {
      "epoch": 4.63,
      "grad_norm": 0.2835392355918884,
      "learning_rate": 7.684501443558673e-07,
      "loss": 0.5706,
      "step": 19140
    },
    {
      "epoch": 4.63,
      "grad_norm": 0.2576424777507782,
      "learning_rate": 7.586672695483271e-07,
      "loss": 0.577,
      "step": 19150
    },
    {
      "epoch": 4.63,
      "grad_norm": 0.2863716185092926,
      "learning_rate": 7.489461060433856e-07,
      "loss": 0.5751,
      "step": 19160
    },
    {
      "epoch": 4.63,
      "grad_norm": 0.2309756726026535,
      "learning_rate": 7.392866785882718e-07,
      "loss": 0.5671,
      "step": 19170
    },
    {
      "epoch": 4.64,
      "grad_norm": 0.27092790603637695,
      "learning_rate": 7.296890117730437e-07,
      "loss": 0.5875,
      "step": 19180
    },
    {
      "epoch": 4.64,
      "grad_norm": 0.4372580647468567,
      "learning_rate": 7.201531300305403e-07,
      "loss": 0.5697,
      "step": 19190
    },
    {
      "epoch": 4.64,
      "grad_norm": 0.2907099425792694,
      "learning_rate": 7.10679057636307e-07,
      "loss": 0.581,
      "step": 19200
    },
    {
      "epoch": 4.64,
      "grad_norm": 0.2552661895751953,
      "learning_rate": 7.012668187085514e-07,
      "loss": 0.5902,
      "step": 19210
    },
    {
      "epoch": 4.64,
      "grad_norm": 0.31329628825187683,
      "learning_rate": 6.919164372080595e-07,
      "loss": 0.5653,
      "step": 19220
    },
    {
      "epoch": 4.65,
      "grad_norm": 0.3134840130805969,
      "learning_rate": 6.826279369381544e-07,
      "loss": 0.5437,
      "step": 19230
    },
    {
      "epoch": 4.65,
      "grad_norm": 0.4079614579677582,
      "learning_rate": 6.734013415446272e-07,
      "loss": 0.6099,
      "step": 19240
    },
    {
      "epoch": 4.65,
      "grad_norm": 0.25411513447761536,
      "learning_rate": 6.642366745156753e-07,
      "loss": 0.5687,
      "step": 19250
    },
    {
      "epoch": 4.65,
      "grad_norm": 0.2598119378089905,
      "learning_rate": 6.551339591818417e-07,
      "loss": 0.5638,
      "step": 19260
    },
    {
      "epoch": 4.66,
      "grad_norm": 0.2853028178215027,
      "learning_rate": 6.46093218715968e-07,
      "loss": 0.5875,
      "step": 19270
    },
    {
      "epoch": 4.66,
      "grad_norm": 0.2891433835029602,
      "learning_rate": 6.371144761331132e-07,
      "loss": 0.6056,
      "step": 19280
    },
    {
      "epoch": 4.66,
      "grad_norm": 0.3109332025051117,
      "learning_rate": 6.281977542905215e-07,
      "loss": 0.5784,
      "step": 19290
    },
    {
      "epoch": 4.66,
      "grad_norm": 0.3131010830402374,
      "learning_rate": 6.193430758875379e-07,
      "loss": 0.5869,
      "step": 19300
    },
    {
      "epoch": 4.67,
      "grad_norm": 0.2744022011756897,
      "learning_rate": 6.105504634655757e-07,
      "loss": 0.5728,
      "step": 19310
    },
    {
      "epoch": 4.67,
      "grad_norm": 0.31314247846603394,
      "learning_rate": 6.018199394080326e-07,
      "loss": 0.5774,
      "step": 19320
    },
    {
      "epoch": 4.67,
      "grad_norm": 0.30135002732276917,
      "learning_rate": 5.931515259402609e-07,
      "loss": 0.5713,
      "step": 19330
    },
    {
      "epoch": 4.67,
      "grad_norm": 0.271109402179718,
      "learning_rate": 5.845452451294892e-07,
      "loss": 0.5834,
      "step": 19340
    },
    {
      "epoch": 4.68,
      "grad_norm": 0.2904196083545685,
      "learning_rate": 5.760011188847753e-07,
      "loss": 0.5668,
      "step": 19350
    },
    {
      "epoch": 4.68,
      "grad_norm": 0.3036612272262573,
      "learning_rate": 5.675191689569509e-07,
      "loss": 0.5885,
      "step": 19360
    },
    {
      "epoch": 4.68,
      "grad_norm": 0.36015647649765015,
      "learning_rate": 5.590994169385688e-07,
      "loss": 0.6312,
      "step": 19370
    },
    {
      "epoch": 4.68,
      "grad_norm": 0.2702130973339081,
      "learning_rate": 5.507418842638417e-07,
      "loss": 0.5705,
      "step": 19380
    },
    {
      "epoch": 4.69,
      "grad_norm": 0.2625230550765991,
      "learning_rate": 5.424465922085842e-07,
      "loss": 0.5731,
      "step": 19390
    },
    {
      "epoch": 4.69,
      "grad_norm": 0.35318925976753235,
      "learning_rate": 5.342135618901734e-07,
      "loss": 0.5678,
      "step": 19400
    },
    {
      "epoch": 4.69,
      "grad_norm": 0.31287771463394165,
      "learning_rate": 5.260428142674828e-07,
      "loss": 0.5678,
      "step": 19410
    },
    {
      "epoch": 4.69,
      "grad_norm": 0.30367550253868103,
      "learning_rate": 5.179343701408324e-07,
      "loss": 0.5715,
      "step": 19420
    },
    {
      "epoch": 4.7,
      "grad_norm": 0.28252947330474854,
      "learning_rate": 5.098882501519353e-07,
      "loss": 0.596,
      "step": 19430
    },
    {
      "epoch": 4.7,
      "grad_norm": 0.2828138470649719,
      "learning_rate": 5.019044747838458e-07,
      "loss": 0.5683,
      "step": 19440
    },
    {
      "epoch": 4.7,
      "grad_norm": 0.3079724609851837,
      "learning_rate": 4.939830643609117e-07,
      "loss": 0.5995,
      "step": 19450
    },
    {
      "epoch": 4.7,
      "grad_norm": 0.3176003098487854,
      "learning_rate": 4.861240390487076e-07,
      "loss": 0.5769,
      "step": 19460
    },
    {
      "epoch": 4.71,
      "grad_norm": 0.30581873655319214,
      "learning_rate": 4.783274188540021e-07,
      "loss": 0.592,
      "step": 19470
    },
    {
      "epoch": 4.71,
      "grad_norm": 0.23425906896591187,
      "learning_rate": 4.7059322362469905e-07,
      "loss": 0.5775,
      "step": 19480
    },
    {
      "epoch": 4.71,
      "grad_norm": 0.27035319805145264,
      "learning_rate": 4.6292147304978215e-07,
      "loss": 0.5857,
      "step": 19490
    },
    {
      "epoch": 4.71,
      "grad_norm": 0.33754491806030273,
      "learning_rate": 4.5531218665927354e-07,
      "loss": 0.5869,
      "step": 19500
    },
    {
      "epoch": 4.71,
      "eval_loss": 0.6268126368522644,
      "eval_runtime": 67.11,
      "eval_samples_per_second": 29.802,
      "eval_steps_per_second": 0.939,
      "step": 19500
    },
    {
      "epoch": 4.71,
      "grad_norm": 0.3014700710773468,
      "learning_rate": 4.47765383824178e-07,
      "loss": 0.5663,
      "step": 19510
    },
    {
      "epoch": 4.72,
      "grad_norm": 0.27844566106796265,
      "learning_rate": 4.402810837564358e-07,
      "loss": 0.5837,
      "step": 19520
    },
    {
      "epoch": 4.72,
      "grad_norm": 0.33787259459495544,
      "learning_rate": 4.3285930550887034e-07,
      "loss": 0.5866,
      "step": 19530
    },
    {
      "epoch": 4.72,
      "grad_norm": 0.3213830888271332,
      "learning_rate": 4.2550006797514586e-07,
      "loss": 0.6091,
      "step": 19540
    },
    {
      "epoch": 4.72,
      "grad_norm": 0.2916968762874603,
      "learning_rate": 4.1820338988972085e-07,
      "loss": 0.5768,
      "step": 19550
    },
    {
      "epoch": 4.73,
      "grad_norm": 0.2971319854259491,
      "learning_rate": 4.1096928982779227e-07,
      "loss": 0.5848,
      "step": 19560
    },
    {
      "epoch": 4.73,
      "grad_norm": 0.30203232169151306,
      "learning_rate": 4.037977862052456e-07,
      "loss": 0.5973,
      "step": 19570
    },
    {
      "epoch": 4.73,
      "grad_norm": 0.34591034054756165,
      "learning_rate": 3.9668889727862144e-07,
      "loss": 0.5945,
      "step": 19580
    },
    {
      "epoch": 4.73,
      "grad_norm": 0.33844736218452454,
      "learning_rate": 3.896426411450632e-07,
      "loss": 0.5768,
      "step": 19590
    },
    {
      "epoch": 4.74,
      "grad_norm": 0.30640116333961487,
      "learning_rate": 3.82659035742261e-07,
      "loss": 0.5987,
      "step": 19600
    },
    {
      "epoch": 4.74,
      "grad_norm": 0.2866669297218323,
      "learning_rate": 3.757380988484244e-07,
      "loss": 0.572,
      "step": 19610
    },
    {
      "epoch": 4.74,
      "grad_norm": 0.317213237285614,
      "learning_rate": 3.688798480822209e-07,
      "loss": 0.5979,
      "step": 19620
    },
    {
      "epoch": 4.74,
      "grad_norm": 0.257652223110199,
      "learning_rate": 3.6208430090274024e-07,
      "loss": 0.5507,
      "step": 19630
    },
    {
      "epoch": 4.75,
      "grad_norm": 0.29283446073532104,
      "learning_rate": 3.553514746094444e-07,
      "loss": 0.6035,
      "step": 19640
    },
    {
      "epoch": 4.75,
      "grad_norm": 0.2867329716682434,
      "learning_rate": 3.48681386342134e-07,
      "loss": 0.5545,
      "step": 19650
    },
    {
      "epoch": 4.75,
      "grad_norm": 0.271506667137146,
      "learning_rate": 3.420740530808847e-07,
      "loss": 0.5904,
      "step": 19660
    },
    {
      "epoch": 4.75,
      "grad_norm": 0.24992626905441284,
      "learning_rate": 3.3552949164603054e-07,
      "loss": 0.5829,
      "step": 19670
    },
    {
      "epoch": 4.76,
      "grad_norm": 0.2408900409936905,
      "learning_rate": 3.2904771869809446e-07,
      "loss": 0.5622,
      "step": 19680
    },
    {
      "epoch": 4.76,
      "grad_norm": 0.2378418892621994,
      "learning_rate": 3.226287507377662e-07,
      "loss": 0.5935,
      "step": 19690
    },
    {
      "epoch": 4.76,
      "grad_norm": 0.321590393781662,
      "learning_rate": 3.1627260410584945e-07,
      "loss": 0.5648,
      "step": 19700
    },
    {
      "epoch": 4.76,
      "grad_norm": 0.2992785573005676,
      "learning_rate": 3.0997929498322585e-07,
      "loss": 0.5624,
      "step": 19710
    },
    {
      "epoch": 4.77,
      "grad_norm": 0.25593581795692444,
      "learning_rate": 3.0374883939080224e-07,
      "loss": 0.5938,
      "step": 19720
    },
    {
      "epoch": 4.77,
      "grad_norm": 0.2549329400062561,
      "learning_rate": 2.975812531894967e-07,
      "loss": 0.5674,
      "step": 19730
    },
    {
      "epoch": 4.77,
      "grad_norm": 0.269255131483078,
      "learning_rate": 2.914765520801638e-07,
      "loss": 0.5774,
      "step": 19740
    },
    {
      "epoch": 4.77,
      "grad_norm": 0.45250871777534485,
      "learning_rate": 2.8543475160357494e-07,
      "loss": 0.5907,
      "step": 19750
    },
    {
      "epoch": 4.78,
      "grad_norm": 0.34732845425605774,
      "learning_rate": 2.7945586714037965e-07,
      "loss": 0.5573,
      "step": 19760
    },
    {
      "epoch": 4.78,
      "grad_norm": 0.2744540870189667,
      "learning_rate": 2.7353991391105837e-07,
      "loss": 0.5697,
      "step": 19770
    },
    {
      "epoch": 4.78,
      "grad_norm": 0.3116280138492584,
      "learning_rate": 2.6768690697588914e-07,
      "loss": 0.5852,
      "step": 19780
    },
    {
      "epoch": 4.78,
      "grad_norm": 0.4008103013038635,
      "learning_rate": 2.61896861234906e-07,
      "loss": 0.5903,
      "step": 19790
    },
    {
      "epoch": 4.78,
      "grad_norm": 0.22020401060581207,
      "learning_rate": 2.5616979142785713e-07,
      "loss": 0.5773,
      "step": 19800
    },
    {
      "epoch": 4.79,
      "grad_norm": 0.30938223004341125,
      "learning_rate": 2.5050571213418584e-07,
      "loss": 0.6048,
      "step": 19810
    },
    {
      "epoch": 4.79,
      "grad_norm": 0.2577236592769623,
      "learning_rate": 2.4490463777296634e-07,
      "loss": 0.5716,
      "step": 19820
    },
    {
      "epoch": 4.79,
      "grad_norm": 0.3309877812862396,
      "learning_rate": 2.3936658260288734e-07,
      "loss": 0.603,
      "step": 19830
    },
    {
      "epoch": 4.79,
      "grad_norm": 0.3032202124595642,
      "learning_rate": 2.3389156072220753e-07,
      "loss": 0.5796,
      "step": 19840
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.260646253824234,
      "learning_rate": 2.28479586068725e-07,
      "loss": 0.5855,
      "step": 19850
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.3001609146595001,
      "learning_rate": 2.2313067241973295e-07,
      "loss": 0.5842,
      "step": 19860
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.24422255158424377,
      "learning_rate": 2.178448333919919e-07,
      "loss": 0.5709,
      "step": 19870
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.3161880671977997,
      "learning_rate": 2.1262208244169636e-07,
      "loss": 0.5848,
      "step": 19880
    },
    {
      "epoch": 4.81,
      "grad_norm": 0.29053041338920593,
      "learning_rate": 2.0746243286443036e-07,
      "loss": 0.5767,
      "step": 19890
    },
    {
      "epoch": 4.81,
      "grad_norm": 0.2951963543891907,
      "learning_rate": 2.0236589779514815e-07,
      "loss": 0.5938,
      "step": 19900
    },
    {
      "epoch": 4.81,
      "grad_norm": 0.3630070984363556,
      "learning_rate": 1.9733249020812416e-07,
      "loss": 0.6206,
      "step": 19910
    },
    {
      "epoch": 4.81,
      "grad_norm": 0.3287997245788574,
      "learning_rate": 1.923622229169364e-07,
      "loss": 0.6034,
      "step": 19920
    },
    {
      "epoch": 4.82,
      "grad_norm": 0.3185696303844452,
      "learning_rate": 1.8745510857442193e-07,
      "loss": 0.5904,
      "step": 19930
    },
    {
      "epoch": 4.82,
      "grad_norm": 0.3890605568885803,
      "learning_rate": 1.826111596726493e-07,
      "loss": 0.5825,
      "step": 19940
    },
    {
      "epoch": 4.82,
      "grad_norm": 0.3463825583457947,
      "learning_rate": 1.7783038854289336e-07,
      "loss": 0.5921,
      "step": 19950
    },
    {
      "epoch": 4.82,
      "grad_norm": 0.3115726709365845,
      "learning_rate": 1.7311280735558545e-07,
      "loss": 0.591,
      "step": 19960
    },
    {
      "epoch": 4.83,
      "grad_norm": 0.27066686749458313,
      "learning_rate": 1.684584281203022e-07,
      "loss": 0.607,
      "step": 19970
    },
    {
      "epoch": 4.83,
      "grad_norm": 0.3084637224674225,
      "learning_rate": 1.6386726268572128e-07,
      "loss": 0.5772,
      "step": 19980
    },
    {
      "epoch": 4.83,
      "grad_norm": 0.28908705711364746,
      "learning_rate": 1.593393227396045e-07,
      "loss": 0.5815,
      "step": 19990
    },
    {
      "epoch": 4.83,
      "grad_norm": 0.3324044644832611,
      "learning_rate": 1.5487461980874807e-07,
      "loss": 0.5829,
      "step": 20000
    },
    {
      "epoch": 4.83,
      "eval_loss": 0.6268470287322998,
      "eval_runtime": 67.2244,
      "eval_samples_per_second": 29.751,
      "eval_steps_per_second": 0.937,
      "step": 20000
    },
    {
      "epoch": 4.84,
      "grad_norm": 0.2697450816631317,
      "learning_rate": 1.5047316525897693e-07,
      "loss": 0.5788,
      "step": 20010
    },
    {
      "epoch": 4.84,
      "grad_norm": 0.3514525294303894,
      "learning_rate": 1.4613497029510038e-07,
      "loss": 0.5692,
      "step": 20020
    },
    {
      "epoch": 4.84,
      "grad_norm": 0.3335043787956238,
      "learning_rate": 1.418600459608871e-07,
      "loss": 0.5949,
      "step": 20030
    },
    {
      "epoch": 4.84,
      "grad_norm": 0.2783183455467224,
      "learning_rate": 1.3764840313903182e-07,
      "loss": 0.5643,
      "step": 20040
    },
    {
      "epoch": 4.85,
      "grad_norm": 0.3691387176513672,
      "learning_rate": 1.3350005255114428e-07,
      "loss": 0.586,
      "step": 20050
    },
    {
      "epoch": 4.85,
      "grad_norm": 0.29611578583717346,
      "learning_rate": 1.2941500475770195e-07,
      "loss": 0.5577,
      "step": 20060
    },
    {
      "epoch": 4.85,
      "grad_norm": 0.3332057595252991,
      "learning_rate": 1.253932701580335e-07,
      "loss": 0.5776,
      "step": 20070
    },
    {
      "epoch": 4.85,
      "grad_norm": 0.3926698863506317,
      "learning_rate": 1.2143485899029638e-07,
      "loss": 0.6003,
      "step": 20080
    },
    {
      "epoch": 4.86,
      "grad_norm": 0.6085434556007385,
      "learning_rate": 1.1753978133143828e-07,
      "loss": 0.5708,
      "step": 20090
    },
    {
      "epoch": 4.86,
      "grad_norm": 0.3020778298377991,
      "learning_rate": 1.137080470971802e-07,
      "loss": 0.5656,
      "step": 20100
    },
    {
      "epoch": 4.86,
      "grad_norm": 0.2559867799282074,
      "learning_rate": 1.0993966604199158e-07,
      "loss": 0.5727,
      "step": 20110
    },
    {
      "epoch": 4.86,
      "grad_norm": 0.26182061433792114,
      "learning_rate": 1.0623464775906533e-07,
      "loss": 0.5509,
      "step": 20120
    },
    {
      "epoch": 4.86,
      "grad_norm": 0.8149334192276001,
      "learning_rate": 1.0259300168028174e-07,
      "loss": 0.6203,
      "step": 20130
    },
    {
      "epoch": 4.87,
      "grad_norm": 0.38119813799858093,
      "learning_rate": 9.901473707620568e-08,
      "loss": 0.5954,
      "step": 20140
    },
    {
      "epoch": 4.87,
      "grad_norm": 0.3422250747680664,
      "learning_rate": 9.549986305604497e-08,
      "loss": 0.5863,
      "step": 20150
    },
    {
      "epoch": 4.87,
      "grad_norm": 0.3847030997276306,
      "learning_rate": 9.204838856763375e-08,
      "loss": 0.572,
      "step": 20160
    },
    {
      "epoch": 4.87,
      "grad_norm": 0.29762041568756104,
      "learning_rate": 8.866032239741307e-08,
      "loss": 0.578,
      "step": 20170
    },
    {
      "epoch": 4.88,
      "grad_norm": 0.3200927972793579,
      "learning_rate": 8.533567317040026e-08,
      "loss": 0.5988,
      "step": 20180
    },
    {
      "epoch": 4.88,
      "grad_norm": 0.47101739048957825,
      "learning_rate": 8.207444935017516e-08,
      "loss": 0.5921,
      "step": 20190
    },
    {
      "epoch": 4.88,
      "grad_norm": 0.29855531454086304,
      "learning_rate": 7.887665923885512e-08,
      "loss": 0.5852,
      "step": 20200
    },
    {
      "epoch": 4.88,
      "grad_norm": 0.3227376639842987,
      "learning_rate": 7.574231097707274e-08,
      "loss": 0.5963,
      "step": 20210
    },
    {
      "epoch": 4.89,
      "grad_norm": 0.25560078024864197,
      "learning_rate": 7.267141254395649e-08,
      "loss": 0.582,
      "step": 20220
    },
    {
      "epoch": 4.89,
      "grad_norm": 0.2820698916912079,
      "learning_rate": 6.966397175711126e-08,
      "loss": 0.6048,
      "step": 20230
    },
    {
      "epoch": 4.89,
      "grad_norm": 0.2958727180957794,
      "learning_rate": 6.671999627259617e-08,
      "loss": 0.5882,
      "step": 20240
    },
    {
      "epoch": 4.89,
      "grad_norm": 0.2670610547065735,
      "learning_rate": 6.383949358490793e-08,
      "loss": 0.5546,
      "step": 20250
    },
    {
      "epoch": 4.9,
      "grad_norm": 0.27611351013183594,
      "learning_rate": 6.10224710269558e-08,
      "loss": 0.5974,
      "step": 20260
    },
    {
      "epoch": 4.9,
      "grad_norm": 0.27489909529685974,
      "learning_rate": 5.826893577005332e-08,
      "loss": 0.5643,
      "step": 20270
    },
    {
      "epoch": 4.9,
      "grad_norm": 0.3481590449810028,
      "learning_rate": 5.557889482389056e-08,
      "loss": 0.5998,
      "step": 20280
    },
    {
      "epoch": 4.9,
      "grad_norm": 0.26827698945999146,
      "learning_rate": 5.2952355036520183e-08,
      "loss": 0.5994,
      "step": 20290
    },
    {
      "epoch": 4.91,
      "grad_norm": 0.3229430615901947,
      "learning_rate": 5.038932309434086e-08,
      "loss": 0.5851,
      "step": 20300
    },
    {
      "epoch": 4.91,
      "grad_norm": 0.23925012350082397,
      "learning_rate": 4.788980552207778e-08,
      "loss": 0.5895,
      "step": 20310
    },
    {
      "epoch": 4.91,
      "grad_norm": 0.26190707087516785,
      "learning_rate": 4.545380868276605e-08,
      "loss": 0.586,
      "step": 20320
    },
    {
      "epoch": 4.91,
      "grad_norm": 0.2643202245235443,
      "learning_rate": 4.3081338777739566e-08,
      "loss": 0.5556,
      "step": 20330
    },
    {
      "epoch": 4.92,
      "grad_norm": 0.26436647772789,
      "learning_rate": 4.07724018466088e-08,
      "loss": 0.5884,
      "step": 20340
    },
    {
      "epoch": 4.92,
      "grad_norm": 0.29357632994651794,
      "learning_rate": 3.852700376724416e-08,
      "loss": 0.574,
      "step": 20350
    },
    {
      "epoch": 4.92,
      "grad_norm": 0.27444806694984436,
      "learning_rate": 3.634515025577323e-08,
      "loss": 0.5846,
      "step": 20360
    },
    {
      "epoch": 4.92,
      "grad_norm": 0.284024715423584,
      "learning_rate": 3.42268468665502e-08,
      "loss": 0.5841,
      "step": 20370
    },
    {
      "epoch": 4.93,
      "grad_norm": 0.2982480823993683,
      "learning_rate": 3.217209899215312e-08,
      "loss": 0.5605,
      "step": 20380
    },
    {
      "epoch": 4.93,
      "grad_norm": 0.23872561752796173,
      "learning_rate": 3.018091186336447e-08,
      "loss": 0.581,
      "step": 20390
    },
    {
      "epoch": 4.93,
      "grad_norm": 0.28587600588798523,
      "learning_rate": 2.825329054916559e-08,
      "loss": 0.5609,
      "step": 20400
    },
    {
      "epoch": 4.93,
      "grad_norm": 0.29186132550239563,
      "learning_rate": 2.6389239956708943e-08,
      "loss": 0.5824,
      "step": 20410
    },
    {
      "epoch": 4.93,
      "grad_norm": 0.3008376657962799,
      "learning_rate": 2.4588764831320887e-08,
      "loss": 0.5782,
      "step": 20420
    },
    {
      "epoch": 4.94,
      "grad_norm": 0.24783658981323242,
      "learning_rate": 2.285186975647946e-08,
      "loss": 0.6024,
      "step": 20430
    },
    {
      "epoch": 4.94,
      "grad_norm": 0.2890031039714813,
      "learning_rate": 2.1178559153811617e-08,
      "loss": 0.5637,
      "step": 20440
    },
    {
      "epoch": 4.94,
      "grad_norm": 0.4320888817310333,
      "learning_rate": 1.956883728307102e-08,
      "loss": 0.5915,
      "step": 20450
    },
    {
      "epoch": 4.94,
      "grad_norm": 0.3041839599609375,
      "learning_rate": 1.802270824213803e-08,
      "loss": 0.5931,
      "step": 20460
    },
    {
      "epoch": 4.95,
      "grad_norm": 0.2837877869606018,
      "learning_rate": 1.6540175967003078e-08,
      "loss": 0.5924,
      "step": 20470
    },
    {
      "epoch": 4.95,
      "grad_norm": 0.26986050605773926,
      "learning_rate": 1.512124423175554e-08,
      "loss": 0.6065,
      "step": 20480
    },
    {
      "epoch": 4.95,
      "grad_norm": 0.3388214111328125,
      "learning_rate": 1.376591664857818e-08,
      "loss": 0.5638,
      "step": 20490
    },
    {
      "epoch": 4.95,
      "grad_norm": 0.29277268052101135,
      "learning_rate": 1.2474196667738857e-08,
      "loss": 0.5587,
      "step": 20500
    },
    {
      "epoch": 4.95,
      "eval_loss": 0.6266828775405884,
      "eval_runtime": 67.3012,
      "eval_samples_per_second": 29.717,
      "eval_steps_per_second": 0.936,
      "step": 20500
    }
  ],
  "logging_steps": 10,
  "max_steps": 20690,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 4.432085131947147e+19,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}