{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 13.2,
  "eval_steps": 500,
  "global_step": 33000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.04,
      "grad_norm": 1.1794846057891846,
      "learning_rate": 0.0002,
      "loss": 2.9634,
      "step": 100
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.9956796169281006,
      "learning_rate": 0.0002,
      "loss": 2.5897,
      "step": 200
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7088788747787476,
      "learning_rate": 0.0002,
      "loss": 2.3967,
      "step": 300
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.8429712057113647,
      "learning_rate": 0.0002,
      "loss": 2.403,
      "step": 400
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.7503288388252258,
      "learning_rate": 0.0002,
      "loss": 2.3169,
      "step": 500
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.6073698401451111,
      "learning_rate": 0.0002,
      "loss": 2.2718,
      "step": 600
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.6131550073623657,
      "learning_rate": 0.0002,
      "loss": 2.1968,
      "step": 700
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.0864523649215698,
      "learning_rate": 0.0002,
      "loss": 2.2322,
      "step": 800
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.7630054950714111,
      "learning_rate": 0.0002,
      "loss": 2.2196,
      "step": 900
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.8388053178787231,
      "learning_rate": 0.0002,
      "loss": 2.2967,
      "step": 1000
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.7474327087402344,
      "learning_rate": 0.0002,
      "loss": 2.222,
      "step": 1100
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.7345255613327026,
      "learning_rate": 0.0002,
      "loss": 2.0606,
      "step": 1200
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.5666030049324036,
      "learning_rate": 0.0002,
      "loss": 2.1423,
      "step": 1300
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.7595005631446838,
      "learning_rate": 0.0002,
      "loss": 2.1629,
      "step": 1400
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.5787087678909302,
      "learning_rate": 0.0002,
      "loss": 2.1169,
      "step": 1500
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.9300134778022766,
      "learning_rate": 0.0002,
      "loss": 2.0647,
      "step": 1600
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.6892948746681213,
      "learning_rate": 0.0002,
      "loss": 2.0773,
      "step": 1700
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.5084001421928406,
      "learning_rate": 0.0002,
      "loss": 2.1002,
      "step": 1800
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.470070481300354,
      "learning_rate": 0.0002,
      "loss": 2.0381,
      "step": 1900
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.6220455169677734,
      "learning_rate": 0.0002,
      "loss": 2.0517,
      "step": 2000
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.8656258583068848,
      "learning_rate": 0.0002,
      "loss": 1.9672,
      "step": 2100
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.7517567873001099,
      "learning_rate": 0.0002,
      "loss": 2.0848,
      "step": 2200
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.7987684011459351,
      "learning_rate": 0.0002,
      "loss": 2.0362,
      "step": 2300
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.6208759546279907,
      "learning_rate": 0.0002,
      "loss": 2.0221,
      "step": 2400
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.5592771768569946,
      "learning_rate": 0.0002,
      "loss": 1.9436,
      "step": 2500
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.8733409643173218,
      "learning_rate": 0.0002,
      "loss": 1.9028,
      "step": 2600
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.6601306200027466,
      "learning_rate": 0.0002,
      "loss": 1.9403,
      "step": 2700
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.7789013385772705,
      "learning_rate": 0.0002,
      "loss": 1.8257,
      "step": 2800
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.5855867862701416,
      "learning_rate": 0.0002,
      "loss": 1.9351,
      "step": 2900
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.5030935406684875,
      "learning_rate": 0.0002,
      "loss": 1.9253,
      "step": 3000
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.6684442758560181,
      "learning_rate": 0.0002,
      "loss": 1.9153,
      "step": 3100
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.6724442839622498,
      "learning_rate": 0.0002,
      "loss": 1.8726,
      "step": 3200
    },
    {
      "epoch": 1.32,
      "grad_norm": 0.48911118507385254,
      "learning_rate": 0.0002,
      "loss": 1.9423,
      "step": 3300
    },
    {
      "epoch": 1.3599999999999999,
      "grad_norm": 0.7769139409065247,
      "learning_rate": 0.0002,
      "loss": 1.7886,
      "step": 3400
    },
    {
      "epoch": 1.4,
      "grad_norm": 0.639460027217865,
      "learning_rate": 0.0002,
      "loss": 1.9118,
      "step": 3500
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.5745570659637451,
      "learning_rate": 0.0002,
      "loss": 1.8423,
      "step": 3600
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.6319829225540161,
      "learning_rate": 0.0002,
      "loss": 1.8893,
      "step": 3700
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.5839726328849792,
      "learning_rate": 0.0002,
      "loss": 1.8506,
      "step": 3800
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.7453562617301941,
      "learning_rate": 0.0002,
      "loss": 1.8698,
      "step": 3900
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.7091575264930725,
      "learning_rate": 0.0002,
      "loss": 1.8367,
      "step": 4000
    },
    {
      "epoch": 1.6400000000000001,
      "grad_norm": 0.722655177116394,
      "learning_rate": 0.0002,
      "loss": 1.8939,
      "step": 4100
    },
    {
      "epoch": 1.6800000000000002,
      "grad_norm": 0.566392183303833,
      "learning_rate": 0.0002,
      "loss": 1.7557,
      "step": 4200
    },
    {
      "epoch": 1.72,
      "grad_norm": 0.564018189907074,
      "learning_rate": 0.0002,
      "loss": 1.72,
      "step": 4300
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.7120116949081421,
      "learning_rate": 0.0002,
      "loss": 1.7436,
      "step": 4400
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.7515490651130676,
      "learning_rate": 0.0002,
      "loss": 1.83,
      "step": 4500
    },
    {
      "epoch": 1.8399999999999999,
      "grad_norm": 0.6733573079109192,
      "learning_rate": 0.0002,
      "loss": 1.8012,
      "step": 4600
    },
    {
      "epoch": 1.88,
      "grad_norm": 1.0119801759719849,
      "learning_rate": 0.0002,
      "loss": 1.8833,
      "step": 4700
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.73843914270401,
      "learning_rate": 0.0002,
      "loss": 1.8012,
      "step": 4800
    },
    {
      "epoch": 1.96,
      "grad_norm": 0.5831142067909241,
      "learning_rate": 0.0002,
      "loss": 1.768,
      "step": 4900
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.5329481959342957,
      "learning_rate": 0.0002,
      "loss": 1.8615,
      "step": 5000
    },
    {
      "epoch": 2.04,
      "grad_norm": 0.5874722599983215,
      "learning_rate": 0.0002,
      "loss": 1.6998,
      "step": 5100
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.5984659194946289,
      "learning_rate": 0.0002,
      "loss": 1.7303,
      "step": 5200
    },
    {
      "epoch": 2.12,
      "grad_norm": 0.7614981532096863,
      "learning_rate": 0.0002,
      "loss": 1.6637,
      "step": 5300
    },
    {
      "epoch": 2.16,
      "grad_norm": 0.8689171671867371,
      "learning_rate": 0.0002,
      "loss": 1.7,
      "step": 5400
    },
    {
      "epoch": 2.2,
      "grad_norm": 0.8071489334106445,
      "learning_rate": 0.0002,
      "loss": 1.6588,
      "step": 5500
    },
    {
      "epoch": 2.24,
      "grad_norm": 0.7306439280509949,
      "learning_rate": 0.0002,
      "loss": 1.7062,
      "step": 5600
    },
    {
      "epoch": 2.2800000000000002,
      "grad_norm": 0.483328253030777,
      "learning_rate": 0.0002,
      "loss": 1.705,
      "step": 5700
    },
    {
      "epoch": 2.32,
      "grad_norm": 0.9662113189697266,
      "learning_rate": 0.0002,
      "loss": 1.6375,
      "step": 5800
    },
    {
      "epoch": 2.36,
      "grad_norm": 0.6570947766304016,
      "learning_rate": 0.0002,
      "loss": 1.6495,
      "step": 5900
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.6682040691375732,
      "learning_rate": 0.0002,
      "loss": 1.6971,
      "step": 6000
    },
    {
      "epoch": 2.44,
      "grad_norm": 0.6937342286109924,
      "learning_rate": 0.0002,
      "loss": 1.6179,
      "step": 6100
    },
    {
      "epoch": 2.48,
      "grad_norm": 0.6113543510437012,
      "learning_rate": 0.0002,
      "loss": 1.6769,
      "step": 6200
    },
    {
      "epoch": 2.52,
      "grad_norm": 0.6734089851379395,
      "learning_rate": 0.0002,
      "loss": 1.6866,
      "step": 6300
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.6203577518463135,
      "learning_rate": 0.0002,
      "loss": 1.6999,
      "step": 6400
    },
    {
      "epoch": 2.6,
      "grad_norm": 0.8067578077316284,
      "learning_rate": 0.0002,
      "loss": 1.6294,
      "step": 6500
    },
    {
      "epoch": 2.64,
      "grad_norm": 0.7386764883995056,
      "learning_rate": 0.0002,
      "loss": 1.6841,
      "step": 6600
    },
    {
      "epoch": 2.68,
      "grad_norm": 0.9529528617858887,
      "learning_rate": 0.0002,
      "loss": 1.665,
      "step": 6700
    },
    {
      "epoch": 2.7199999999999998,
      "grad_norm": 0.5639382600784302,
      "learning_rate": 0.0002,
      "loss": 1.6756,
      "step": 6800
    },
    {
      "epoch": 2.76,
      "grad_norm": 0.8711239695549011,
      "learning_rate": 0.0002,
      "loss": 1.5953,
      "step": 6900
    },
    {
      "epoch": 2.8,
      "grad_norm": 0.6677307486534119,
      "learning_rate": 0.0002,
      "loss": 1.6708,
      "step": 7000
    },
    {
      "epoch": 2.84,
      "grad_norm": 0.6820212006568909,
      "learning_rate": 0.0002,
      "loss": 1.6722,
      "step": 7100
    },
    {
      "epoch": 2.88,
      "grad_norm": 0.6516992449760437,
      "learning_rate": 0.0002,
      "loss": 1.6407,
      "step": 7200
    },
    {
      "epoch": 2.92,
      "grad_norm": 0.6071237325668335,
      "learning_rate": 0.0002,
      "loss": 1.6273,
      "step": 7300
    },
    {
      "epoch": 2.96,
      "grad_norm": 0.6759991645812988,
      "learning_rate": 0.0002,
      "loss": 1.6218,
      "step": 7400
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.6260673403739929,
      "learning_rate": 0.0002,
      "loss": 1.7184,
      "step": 7500
    },
    {
      "epoch": 3.04,
      "grad_norm": 0.6020120978355408,
      "learning_rate": 0.0002,
      "loss": 1.4686,
      "step": 7600
    },
    {
      "epoch": 3.08,
      "grad_norm": 0.9953874945640564,
      "learning_rate": 0.0002,
      "loss": 1.5124,
      "step": 7700
    },
    {
      "epoch": 3.12,
      "grad_norm": 0.7787545323371887,
      "learning_rate": 0.0002,
      "loss": 1.5184,
      "step": 7800
    },
    {
      "epoch": 3.16,
      "grad_norm": 0.9334218502044678,
      "learning_rate": 0.0002,
      "loss": 1.5031,
      "step": 7900
    },
    {
      "epoch": 3.2,
      "grad_norm": 0.9822279214859009,
      "learning_rate": 0.0002,
      "loss": 1.504,
      "step": 8000
    },
    {
      "epoch": 3.24,
      "grad_norm": 0.8602248430252075,
      "learning_rate": 0.0002,
      "loss": 1.5348,
      "step": 8100
    },
    {
      "epoch": 3.2800000000000002,
      "grad_norm": 0.88422691822052,
      "learning_rate": 0.0002,
      "loss": 1.4788,
      "step": 8200
    },
    {
      "epoch": 3.32,
      "grad_norm": 0.8250532746315002,
      "learning_rate": 0.0002,
      "loss": 1.5258,
      "step": 8300
    },
    {
      "epoch": 3.36,
      "grad_norm": 1.3689357042312622,
      "learning_rate": 0.0002,
      "loss": 1.53,
      "step": 8400
    },
    {
      "epoch": 3.4,
      "grad_norm": 1.0472410917282104,
      "learning_rate": 0.0002,
      "loss": 1.5347,
      "step": 8500
    },
    {
      "epoch": 3.44,
      "grad_norm": 0.7986043095588684,
      "learning_rate": 0.0002,
      "loss": 1.4467,
      "step": 8600
    },
    {
      "epoch": 3.48,
      "grad_norm": 0.8166589736938477,
      "learning_rate": 0.0002,
      "loss": 1.6207,
      "step": 8700
    },
    {
      "epoch": 3.52,
      "grad_norm": 0.7566811442375183,
      "learning_rate": 0.0002,
      "loss": 1.5238,
      "step": 8800
    },
    {
      "epoch": 3.56,
      "grad_norm": 0.9338216185569763,
      "learning_rate": 0.0002,
      "loss": 1.5801,
      "step": 8900
    },
    {
      "epoch": 3.6,
      "grad_norm": 0.9602301716804504,
      "learning_rate": 0.0002,
      "loss": 1.4506,
      "step": 9000
    },
    {
      "epoch": 3.64,
      "grad_norm": 0.9274515509605408,
      "learning_rate": 0.0002,
      "loss": 1.5574,
      "step": 9100
    },
    {
      "epoch": 3.68,
      "grad_norm": 1.6966444253921509,
      "learning_rate": 0.0002,
      "loss": 1.4391,
      "step": 9200
    },
    {
      "epoch": 3.7199999999999998,
      "grad_norm": 0.9805380702018738,
      "learning_rate": 0.0002,
      "loss": 1.5071,
      "step": 9300
    },
    {
      "epoch": 3.76,
      "grad_norm": 1.1102324724197388,
      "learning_rate": 0.0002,
      "loss": 1.5041,
      "step": 9400
    },
    {
      "epoch": 3.8,
      "grad_norm": 1.1052135229110718,
      "learning_rate": 0.0002,
      "loss": 1.504,
      "step": 9500
    },
    {
      "epoch": 3.84,
      "grad_norm": 0.8187786936759949,
      "learning_rate": 0.0002,
      "loss": 1.5444,
      "step": 9600
    },
    {
      "epoch": 3.88,
      "grad_norm": 1.5404032468795776,
      "learning_rate": 0.0002,
      "loss": 1.5627,
      "step": 9700
    },
    {
      "epoch": 3.92,
      "grad_norm": 0.7880242466926575,
      "learning_rate": 0.0002,
      "loss": 1.508,
      "step": 9800
    },
    {
      "epoch": 3.96,
      "grad_norm": 1.2433679103851318,
      "learning_rate": 0.0002,
      "loss": 1.5252,
      "step": 9900
    },
    {
      "epoch": 4.0,
      "grad_norm": NaN,
      "learning_rate": 0.0002,
      "loss": 1.5194,
      "step": 10000
    },
    {
      "epoch": 4.04,
      "grad_norm": 0.9061052799224854,
      "learning_rate": 0.0002,
      "loss": 1.3939,
      "step": 10100
    },
    {
      "epoch": 4.08,
      "grad_norm": 0.9826890230178833,
      "learning_rate": 0.0002,
      "loss": 1.3602,
      "step": 10200
    },
    {
      "epoch": 4.12,
      "grad_norm": 1.1164418458938599,
      "learning_rate": 0.0002,
      "loss": 1.4036,
      "step": 10300
    },
    {
      "epoch": 4.16,
      "grad_norm": 1.2371020317077637,
      "learning_rate": 0.0002,
      "loss": 1.4325,
      "step": 10400
    },
    {
      "epoch": 4.2,
      "grad_norm": 0.9281136393547058,
      "learning_rate": 0.0002,
      "loss": 1.3633,
      "step": 10500
    },
    {
      "epoch": 4.24,
      "grad_norm": 1.5267653465270996,
      "learning_rate": 0.0002,
      "loss": 1.3531,
      "step": 10600
    },
    {
      "epoch": 4.28,
      "grad_norm": 0.9372194409370422,
      "learning_rate": 0.0002,
      "loss": 1.387,
      "step": 10700
    },
    {
      "epoch": 4.32,
      "grad_norm": 1.9180704355239868,
      "learning_rate": 0.0002,
      "loss": 1.4049,
      "step": 10800
    },
    {
      "epoch": 4.36,
      "grad_norm": 0.9102849960327148,
      "learning_rate": 0.0002,
      "loss": 1.3481,
      "step": 10900
    },
    {
      "epoch": 4.4,
      "grad_norm": 1.3661117553710938,
      "learning_rate": 0.0002,
      "loss": 1.3171,
      "step": 11000
    },
    {
      "epoch": 4.44,
      "grad_norm": 0.6796606183052063,
      "learning_rate": 0.0002,
      "loss": 1.3776,
      "step": 11100
    },
    {
      "epoch": 4.48,
      "grad_norm": 1.2240846157073975,
      "learning_rate": 0.0002,
      "loss": 1.3756,
      "step": 11200
    },
    {
      "epoch": 4.52,
      "grad_norm": 0.7911117672920227,
      "learning_rate": 0.0002,
      "loss": 1.3484,
      "step": 11300
    },
    {
      "epoch": 4.5600000000000005,
      "grad_norm": 0.6849353313446045,
      "learning_rate": 0.0002,
      "loss": 1.3985,
      "step": 11400
    },
    {
      "epoch": 4.6,
      "grad_norm": 1.292270541191101,
      "learning_rate": 0.0002,
      "loss": 1.3338,
      "step": 11500
    },
    {
      "epoch": 4.64,
      "grad_norm": 1.0751451253890991,
      "learning_rate": 0.0002,
      "loss": 1.4044,
      "step": 11600
    },
    {
      "epoch": 4.68,
      "grad_norm": 1.391108512878418,
      "learning_rate": 0.0002,
      "loss": 1.4003,
      "step": 11700
    },
    {
      "epoch": 4.72,
      "grad_norm": 0.8339885473251343,
      "learning_rate": 0.0002,
      "loss": 1.389,
      "step": 11800
    },
    {
      "epoch": 4.76,
      "grad_norm": 0.9836968779563904,
      "learning_rate": 0.0002,
      "loss": 1.3732,
      "step": 11900
    },
    {
      "epoch": 4.8,
      "grad_norm": 1.3942408561706543,
      "learning_rate": 0.0002,
      "loss": 1.3763,
      "step": 12000
    },
    {
      "epoch": 4.84,
      "grad_norm": 0.8473936915397644,
      "learning_rate": 0.0002,
      "loss": 1.3675,
      "step": 12100
    },
    {
      "epoch": 4.88,
      "grad_norm": 1.1180263757705688,
      "learning_rate": 0.0002,
      "loss": 1.3018,
      "step": 12200
    },
    {
      "epoch": 4.92,
      "grad_norm": 1.0825896263122559,
      "learning_rate": 0.0002,
      "loss": 1.3935,
      "step": 12300
    },
    {
      "epoch": 4.96,
      "grad_norm": 1.7957453727722168,
      "learning_rate": 0.0002,
      "loss": 1.4167,
      "step": 12400
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.4984807968139648,
      "learning_rate": 0.0002,
      "loss": 1.3562,
      "step": 12500
    },
    {
      "epoch": 5.04,
      "grad_norm": 1.2871723175048828,
      "learning_rate": 0.0002,
      "loss": 1.2327,
      "step": 12600
    },
    {
      "epoch": 5.08,
      "grad_norm": 0.5957996249198914,
      "learning_rate": 0.0002,
      "loss": 1.1613,
      "step": 12700
    },
    {
      "epoch": 5.12,
      "grad_norm": 1.0731712579727173,
      "learning_rate": 0.0002,
      "loss": 1.2683,
      "step": 12800
    },
    {
      "epoch": 5.16,
      "grad_norm": 1.2785401344299316,
      "learning_rate": 0.0002,
      "loss": 1.2215,
      "step": 12900
    },
    {
      "epoch": 5.2,
      "grad_norm": 0.6945757269859314,
      "learning_rate": 0.0002,
      "loss": 1.2665,
      "step": 13000
    },
    {
      "epoch": 5.24,
      "grad_norm": 0.8779969215393066,
      "learning_rate": 0.0002,
      "loss": 1.2028,
      "step": 13100
    },
    {
      "epoch": 5.28,
      "grad_norm": 0.9905422329902649,
      "learning_rate": 0.0002,
      "loss": 1.2691,
      "step": 13200
    },
    {
      "epoch": 5.32,
      "grad_norm": 0.7267619967460632,
      "learning_rate": 0.0002,
      "loss": 1.2794,
      "step": 13300
    },
    {
      "epoch": 5.36,
      "grad_norm": 1.1280577182769775,
      "learning_rate": 0.0002,
      "loss": 1.2608,
      "step": 13400
    },
    {
      "epoch": 5.4,
      "grad_norm": 1.3053045272827148,
      "learning_rate": 0.0002,
      "loss": 1.253,
      "step": 13500
    },
    {
      "epoch": 5.44,
      "grad_norm": 1.1373580694198608,
      "learning_rate": 0.0002,
      "loss": 1.2236,
      "step": 13600
    },
    {
      "epoch": 5.48,
      "grad_norm": 0.9823132753372192,
      "learning_rate": 0.0002,
      "loss": 1.2841,
      "step": 13700
    },
    {
      "epoch": 5.52,
      "grad_norm": 1.0225436687469482,
      "learning_rate": 0.0002,
      "loss": 1.2052,
      "step": 13800
    },
    {
      "epoch": 5.5600000000000005,
      "grad_norm": 1.6617635488510132,
      "learning_rate": 0.0002,
      "loss": 1.2314,
      "step": 13900
    },
    {
      "epoch": 5.6,
      "grad_norm": 1.0195096731185913,
      "learning_rate": 0.0002,
      "loss": 1.2959,
      "step": 14000
    },
    {
      "epoch": 5.64,
      "grad_norm": 1.3563017845153809,
      "learning_rate": 0.0002,
      "loss": 1.2683,
      "step": 14100
    },
    {
      "epoch": 5.68,
      "grad_norm": 2.1966164112091064,
      "learning_rate": 0.0002,
      "loss": 1.2247,
      "step": 14200
    },
    {
      "epoch": 5.72,
      "grad_norm": 1.057099461555481,
      "learning_rate": 0.0002,
      "loss": 1.2577,
      "step": 14300
    },
    {
      "epoch": 5.76,
      "grad_norm": 1.2523263692855835,
      "learning_rate": 0.0002,
      "loss": 1.2865,
      "step": 14400
    },
    {
      "epoch": 5.8,
      "grad_norm": 1.7607208490371704,
      "learning_rate": 0.0002,
      "loss": 1.2205,
      "step": 14500
    },
    {
      "epoch": 5.84,
      "grad_norm": 1.9455257654190063,
      "learning_rate": 0.0002,
      "loss": 1.2829,
      "step": 14600
    },
    {
      "epoch": 5.88,
      "grad_norm": 1.0618771314620972,
      "learning_rate": 0.0002,
      "loss": 1.3037,
      "step": 14700
    },
    {
      "epoch": 5.92,
      "grad_norm": 1.0242942571640015,
      "learning_rate": 0.0002,
      "loss": 1.2165,
      "step": 14800
    },
    {
      "epoch": 5.96,
      "grad_norm": 1.2692897319793701,
      "learning_rate": 0.0002,
      "loss": 1.2873,
      "step": 14900
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.0533056259155273,
      "learning_rate": 0.0002,
      "loss": 1.2528,
      "step": 15000
    },
    {
      "epoch": 6.04,
      "grad_norm": 0.8311458230018616,
      "learning_rate": 0.0002,
      "loss": 1.0705,
      "step": 15100
    },
    {
      "epoch": 6.08,
      "grad_norm": 1.0614266395568848,
      "learning_rate": 0.0002,
      "loss": 1.0803,
      "step": 15200
    },
    {
      "epoch": 6.12,
      "grad_norm": 1.4157719612121582,
      "learning_rate": 0.0002,
      "loss": 1.0864,
      "step": 15300
    },
    {
      "epoch": 6.16,
      "grad_norm": 1.0785095691680908,
      "learning_rate": 0.0002,
      "loss": 1.1076,
      "step": 15400
    },
    {
      "epoch": 6.2,
      "grad_norm": 1.3798463344573975,
      "learning_rate": 0.0002,
      "loss": 1.1038,
      "step": 15500
    },
    {
      "epoch": 6.24,
      "grad_norm": 0.9821926355361938,
      "learning_rate": 0.0002,
      "loss": 1.1482,
      "step": 15600
    },
    {
      "epoch": 6.28,
      "grad_norm": 2.189770221710205,
      "learning_rate": 0.0002,
      "loss": 1.1154,
      "step": 15700
    },
    {
      "epoch": 6.32,
      "grad_norm": 5.5393500328063965,
      "learning_rate": 0.0002,
      "loss": 1.1186,
      "step": 15800
    },
    {
      "epoch": 6.36,
      "grad_norm": 1.2127723693847656,
      "learning_rate": 0.0002,
      "loss": 1.1888,
      "step": 15900
    },
    {
      "epoch": 6.4,
      "grad_norm": 2.876635789871216,
      "learning_rate": 0.0002,
      "loss": 1.1029,
      "step": 16000
    },
    {
      "epoch": 6.44,
      "grad_norm": 1.3155653476715088,
      "learning_rate": 0.0002,
      "loss": 1.1092,
      "step": 16100
    },
    {
      "epoch": 6.48,
      "grad_norm": 1.1871975660324097,
      "learning_rate": 0.0002,
      "loss": 1.1488,
      "step": 16200
    },
    {
      "epoch": 6.52,
      "grad_norm": 1.6706851720809937,
      "learning_rate": 0.0002,
      "loss": 1.153,
      "step": 16300
    },
    {
      "epoch": 6.5600000000000005,
      "grad_norm": 2.3989503383636475,
      "learning_rate": 0.0002,
      "loss": 1.1049,
      "step": 16400
    },
    {
      "epoch": 6.6,
      "grad_norm": 1.0962737798690796,
      "learning_rate": 0.0002,
      "loss": 1.182,
      "step": 16500
    },
    {
      "epoch": 6.64,
      "grad_norm": 1.2321207523345947,
      "learning_rate": 0.0002,
      "loss": 1.1196,
      "step": 16600
    },
    {
      "epoch": 6.68,
      "grad_norm": 1.6745890378952026,
      "learning_rate": 0.0002,
      "loss": 1.1776,
      "step": 16700
    },
    {
      "epoch": 6.72,
      "grad_norm": 3.3587148189544678,
      "learning_rate": 0.0002,
      "loss": 1.1335,
      "step": 16800
    },
    {
      "epoch": 6.76,
      "grad_norm": 1.2802035808563232,
      "learning_rate": 0.0002,
      "loss": 1.2144,
      "step": 16900
    },
    {
      "epoch": 6.8,
      "grad_norm": 1.5999842882156372,
      "learning_rate": 0.0002,
      "loss": 1.1143,
      "step": 17000
    },
    {
      "epoch": 6.84,
      "grad_norm": 1.5529465675354004,
      "learning_rate": 0.0002,
      "loss": 1.1503,
      "step": 17100
    },
    {
      "epoch": 6.88,
      "grad_norm": 0.9041572213172913,
      "learning_rate": 0.0002,
      "loss": 1.1681,
      "step": 17200
    },
    {
      "epoch": 6.92,
      "grad_norm": 1.2176377773284912,
      "learning_rate": 0.0002,
      "loss": 1.1284,
      "step": 17300
    },
    {
      "epoch": 6.96,
      "grad_norm": 1.1047685146331787,
      "learning_rate": 0.0002,
      "loss": 1.1622,
      "step": 17400
    },
    {
      "epoch": 7.0,
      "grad_norm": 2.3316566944122314,
      "learning_rate": 0.0002,
      "loss": 1.149,
      "step": 17500
    },
    {
      "epoch": 7.04,
      "grad_norm": 0.9057099223136902,
      "learning_rate": 0.0002,
      "loss": 0.9952,
      "step": 17600
    },
    {
      "epoch": 7.08,
      "grad_norm": 2.2556183338165283,
      "learning_rate": 0.0002,
      "loss": 0.9667,
      "step": 17700
    },
    {
      "epoch": 7.12,
      "grad_norm": 1.2181649208068848,
      "learning_rate": 0.0002,
      "loss": 0.9606,
      "step": 17800
    },
    {
      "epoch": 7.16,
      "grad_norm": 1.1756635904312134,
      "learning_rate": 0.0002,
      "loss": 1.0176,
      "step": 17900
    },
    {
      "epoch": 7.2,
      "grad_norm": 1.3955613374710083,
      "learning_rate": 0.0002,
      "loss": 1.0423,
      "step": 18000
    },
    {
      "epoch": 7.24,
      "grad_norm": 2.0451719760894775,
      "learning_rate": 0.0002,
      "loss": 1.0152,
      "step": 18100
    },
    {
      "epoch": 7.28,
      "grad_norm": 1.8395252227783203,
      "learning_rate": 0.0002,
      "loss": 0.9896,
      "step": 18200
    },
    {
      "epoch": 7.32,
      "grad_norm": 0.8587050437927246,
      "learning_rate": 0.0002,
      "loss": 0.9976,
      "step": 18300
    },
    {
      "epoch": 7.36,
      "grad_norm": 1.848895788192749,
      "learning_rate": 0.0002,
      "loss": 0.9711,
      "step": 18400
    },
    {
      "epoch": 7.4,
      "grad_norm": 0.8069963455200195,
      "learning_rate": 0.0002,
      "loss": 1.02,
      "step": 18500
    },
    {
      "epoch": 7.44,
      "grad_norm": 1.0655219554901123,
      "learning_rate": 0.0002,
      "loss": 1.0069,
      "step": 18600
    },
    {
      "epoch": 7.48,
      "grad_norm": 1.474328637123108,
      "learning_rate": 0.0002,
      "loss": 1.0311,
      "step": 18700
    },
    {
      "epoch": 7.52,
      "grad_norm": 1.3253296613693237,
      "learning_rate": 0.0002,
      "loss": 1.0158,
      "step": 18800
    },
    {
      "epoch": 7.5600000000000005,
      "grad_norm": 1.1345421075820923,
      "learning_rate": 0.0002,
      "loss": 1.0391,
      "step": 18900
    },
    {
      "epoch": 7.6,
      "grad_norm": 1.0736902952194214,
      "learning_rate": 0.0002,
      "loss": 1.0498,
      "step": 19000
    },
    {
      "epoch": 7.64,
      "grad_norm": 1.6537193059921265,
      "learning_rate": 0.0002,
      "loss": 1.0235,
      "step": 19100
    },
    {
      "epoch": 7.68,
      "grad_norm": 1.3010786771774292,
      "learning_rate": 0.0002,
      "loss": 1.0307,
      "step": 19200
    },
    {
      "epoch": 7.72,
      "grad_norm": 0.9209179282188416,
      "learning_rate": 0.0002,
      "loss": 1.0111,
      "step": 19300
    },
    {
      "epoch": 7.76,
      "grad_norm": 5.0836310386657715,
      "learning_rate": 0.0002,
      "loss": 1.059,
      "step": 19400
    },
    {
      "epoch": 7.8,
      "grad_norm": 1.7362360954284668,
      "learning_rate": 0.0002,
      "loss": 1.0704,
      "step": 19500
    },
    {
      "epoch": 7.84,
      "grad_norm": 1.692413330078125,
      "learning_rate": 0.0002,
      "loss": 1.0618,
      "step": 19600
    },
    {
      "epoch": 7.88,
      "grad_norm": 3.538470506668091,
      "learning_rate": 0.0002,
      "loss": 1.0778,
      "step": 19700
    },
    {
      "epoch": 7.92,
      "grad_norm": 2.578237533569336,
      "learning_rate": 0.0002,
      "loss": 1.1153,
      "step": 19800
    },
    {
      "epoch": 7.96,
      "grad_norm": 1.356609582901001,
      "learning_rate": 0.0002,
      "loss": 1.0597,
      "step": 19900
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.89506995677948,
      "learning_rate": 0.0002,
      "loss": 1.0892,
      "step": 20000
    },
    {
      "epoch": 8.04,
      "grad_norm": 1.5506243705749512,
      "learning_rate": 0.0002,
      "loss": 0.8659,
      "step": 20100
    },
    {
      "epoch": 8.08,
      "grad_norm": 1.69996178150177,
      "learning_rate": 0.0002,
      "loss": 0.9142,
      "step": 20200
    },
    {
      "epoch": 8.12,
      "grad_norm": 2.0094783306121826,
      "learning_rate": 0.0002,
      "loss": 0.8807,
      "step": 20300
    },
    {
      "epoch": 8.16,
      "grad_norm": 1.6503652334213257,
      "learning_rate": 0.0002,
      "loss": 0.857,
      "step": 20400
    },
    {
      "epoch": 8.2,
      "grad_norm": 0.933527410030365,
      "learning_rate": 0.0002,
      "loss": 0.8855,
      "step": 20500
    },
    {
      "epoch": 8.24,
      "grad_norm": 1.6827526092529297,
      "learning_rate": 0.0002,
      "loss": 0.9295,
      "step": 20600
    },
    {
      "epoch": 8.28,
      "grad_norm": 2.89079213142395,
      "learning_rate": 0.0002,
      "loss": 0.9146,
      "step": 20700
    },
    {
      "epoch": 8.32,
      "grad_norm": 1.0603892803192139,
      "learning_rate": 0.0002,
      "loss": 0.9279,
      "step": 20800
    },
    {
      "epoch": 8.36,
      "grad_norm": 3.3014211654663086,
      "learning_rate": 0.0002,
      "loss": 0.9406,
      "step": 20900
    },
    {
      "epoch": 8.4,
      "grad_norm": 0.8049854040145874,
      "learning_rate": 0.0002,
      "loss": 0.907,
      "step": 21000
    },
    {
      "epoch": 8.44,
      "grad_norm": 2.991314649581909,
      "learning_rate": 0.0002,
      "loss": 0.9721,
      "step": 21100
    },
    {
      "epoch": 8.48,
      "grad_norm": 1.043578028678894,
      "learning_rate": 0.0002,
      "loss": 0.9557,
      "step": 21200
    },
    {
      "epoch": 8.52,
      "grad_norm": 1.6888822317123413,
      "learning_rate": 0.0002,
      "loss": 0.9477,
      "step": 21300
    },
    {
      "epoch": 8.56,
      "grad_norm": 1.4089540243148804,
      "learning_rate": 0.0002,
      "loss": 0.9424,
      "step": 21400
    },
    {
      "epoch": 8.6,
      "grad_norm": 7.123160362243652,
      "learning_rate": 0.0002,
      "loss": 0.9232,
      "step": 21500
    },
    {
      "epoch": 8.64,
      "grad_norm": 1.149699091911316,
      "learning_rate": 0.0002,
      "loss": 0.9189,
      "step": 21600
    },
    {
      "epoch": 8.68,
      "grad_norm": 1.3870540857315063,
      "learning_rate": 0.0002,
      "loss": 0.9318,
      "step": 21700
    },
    {
      "epoch": 8.72,
      "grad_norm": 1.4124248027801514,
      "learning_rate": 0.0002,
      "loss": 0.9657,
      "step": 21800
    },
    {
      "epoch": 8.76,
      "grad_norm": 0.6599737405776978,
      "learning_rate": 0.0002,
      "loss": 0.9493,
      "step": 21900
    },
    {
      "epoch": 8.8,
      "grad_norm": 1.0088489055633545,
      "learning_rate": 0.0002,
      "loss": 0.9739,
      "step": 22000
    },
    {
      "epoch": 8.84,
      "grad_norm": 0.7540724873542786,
      "learning_rate": 0.0002,
      "loss": 0.9492,
      "step": 22100
    },
    {
      "epoch": 8.88,
      "grad_norm": 1.0838185548782349,
      "learning_rate": 0.0002,
      "loss": 1.0105,
      "step": 22200
    },
    {
      "epoch": 8.92,
      "grad_norm": 0.8989962935447693,
      "learning_rate": 0.0002,
      "loss": 0.9256,
      "step": 22300
    },
    {
      "epoch": 8.96,
      "grad_norm": 0.9646226167678833,
      "learning_rate": 0.0002,
      "loss": 0.9766,
      "step": 22400
    },
    {
      "epoch": 9.0,
      "grad_norm": 1.0363638401031494,
      "learning_rate": 0.0002,
      "loss": 0.9676,
      "step": 22500
    },
    {
      "epoch": 9.04,
      "grad_norm": 1.2631175518035889,
      "learning_rate": 0.0002,
      "loss": 0.7845,
      "step": 22600
    },
    {
      "epoch": 9.08,
      "grad_norm": 1.2183212041854858,
      "learning_rate": 0.0002,
      "loss": 0.7785,
      "step": 22700
    },
    {
      "epoch": 9.12,
      "grad_norm": 1.4566229581832886,
      "learning_rate": 0.0002,
      "loss": 0.7972,
      "step": 22800
    },
    {
      "epoch": 9.16,
      "grad_norm": 1.2739365100860596,
      "learning_rate": 0.0002,
      "loss": 0.7729,
      "step": 22900
    },
    {
      "epoch": 9.2,
      "grad_norm": 1.4455575942993164,
      "learning_rate": 0.0002,
      "loss": 0.8359,
      "step": 23000
    },
    {
      "epoch": 9.24,
      "grad_norm": 1.1576048135757446,
      "learning_rate": 0.0002,
      "loss": 0.8324,
      "step": 23100
    },
    {
      "epoch": 9.28,
      "grad_norm": 1.3177443742752075,
      "learning_rate": 0.0002,
      "loss": 0.8223,
      "step": 23200
    },
    {
      "epoch": 9.32,
      "grad_norm": 4.67700719833374,
      "learning_rate": 0.0002,
      "loss": 0.8154,
      "step": 23300
    },
    {
      "epoch": 9.36,
      "grad_norm": 1.2279400825500488,
      "learning_rate": 0.0002,
      "loss": 0.8047,
      "step": 23400
    },
    {
      "epoch": 9.4,
      "grad_norm": 0.6378070712089539,
      "learning_rate": 0.0002,
      "loss": 0.8539,
      "step": 23500
    },
    {
      "epoch": 9.44,
      "grad_norm": 1.7342982292175293,
      "learning_rate": 0.0002,
      "loss": 0.8405,
      "step": 23600
    },
    {
      "epoch": 9.48,
      "grad_norm": 1.534493088722229,
      "learning_rate": 0.0002,
      "loss": 0.8493,
      "step": 23700
    },
    {
      "epoch": 9.52,
      "grad_norm": 1.0669933557510376,
      "learning_rate": 0.0002,
      "loss": 0.8931,
      "step": 23800
    },
    {
      "epoch": 9.56,
      "grad_norm": 1.2402708530426025,
      "learning_rate": 0.0002,
      "loss": 0.8689,
      "step": 23900
    },
    {
      "epoch": 9.6,
      "grad_norm": 1.2599835395812988,
      "learning_rate": 0.0002,
      "loss": 0.8733,
      "step": 24000
    },
    {
      "epoch": 9.64,
      "grad_norm": 2.5779175758361816,
      "learning_rate": 0.0002,
      "loss": 0.8934,
      "step": 24100
    },
    {
      "epoch": 9.68,
      "grad_norm": 1.8797070980072021,
      "learning_rate": 0.0002,
      "loss": 0.8536,
      "step": 24200
    },
    {
      "epoch": 9.72,
      "grad_norm": 1.3404995203018188,
      "learning_rate": 0.0002,
      "loss": 0.9191,
      "step": 24300
    },
    {
      "epoch": 9.76,
      "grad_norm": 1.4789234399795532,
      "learning_rate": 0.0002,
      "loss": 0.8198,
      "step": 24400
    },
    {
      "epoch": 9.8,
      "grad_norm": 0.6044953465461731,
      "learning_rate": 0.0002,
      "loss": 0.8304,
      "step": 24500
    },
    {
      "epoch": 9.84,
      "grad_norm": 1.2792354822158813,
      "learning_rate": 0.0002,
      "loss": 0.9196,
      "step": 24600
    },
    {
      "epoch": 9.88,
      "grad_norm": 2.047617197036743,
      "learning_rate": 0.0002,
      "loss": 0.9136,
      "step": 24700
    },
    {
      "epoch": 9.92,
      "grad_norm": 5.776161193847656,
      "learning_rate": 0.0002,
      "loss": 0.8715,
      "step": 24800
    },
    {
      "epoch": 9.96,
      "grad_norm": 1.9059360027313232,
      "learning_rate": 0.0002,
      "loss": 0.8712,
      "step": 24900
    },
    {
      "epoch": 10.0,
      "grad_norm": 1.1497496366500854,
      "learning_rate": 0.0002,
      "loss": 0.8607,
      "step": 25000
    },
    {
      "epoch": 10.04,
      "grad_norm": 0.6882659792900085,
      "learning_rate": 0.0002,
      "loss": 0.7556,
      "step": 25100
    },
    {
      "epoch": 10.08,
      "grad_norm": 0.9709841012954712,
      "learning_rate": 0.0002,
      "loss": 0.787,
      "step": 25200
    },
    {
      "epoch": 10.12,
      "grad_norm": 0.9609636664390564,
      "learning_rate": 0.0002,
      "loss": 0.701,
      "step": 25300
    },
    {
      "epoch": 10.16,
      "grad_norm": 1.397544026374817,
      "learning_rate": 0.0002,
      "loss": 0.7548,
      "step": 25400
    },
    {
      "epoch": 10.2,
      "grad_norm": 2.043348550796509,
      "learning_rate": 0.0002,
      "loss": 0.7357,
      "step": 25500
    },
    {
      "epoch": 10.24,
      "grad_norm": 0.7744215726852417,
      "learning_rate": 0.0002,
      "loss": 0.7481,
      "step": 25600
    },
    {
      "epoch": 10.28,
      "grad_norm": 0.7471018433570862,
      "learning_rate": 0.0002,
      "loss": 0.7206,
      "step": 25700
    },
    {
      "epoch": 10.32,
      "grad_norm": 1.5226207971572876,
      "learning_rate": 0.0002,
      "loss": 0.7199,
      "step": 25800
    },
    {
      "epoch": 10.36,
      "grad_norm": 0.9229708313941956,
      "learning_rate": 0.0002,
      "loss": 0.7449,
      "step": 25900
    },
    {
      "epoch": 10.4,
      "grad_norm": 0.8783457279205322,
      "learning_rate": 0.0002,
      "loss": 0.8026,
      "step": 26000
    },
    {
      "epoch": 10.44,
      "grad_norm": 1.858168125152588,
      "learning_rate": 0.0002,
      "loss": 0.758,
      "step": 26100
    },
    {
      "epoch": 10.48,
      "grad_norm": 9.689515113830566,
      "learning_rate": 0.0002,
      "loss": 0.7394,
      "step": 26200
    },
    {
      "epoch": 10.52,
      "grad_norm": 1.7214679718017578,
      "learning_rate": 0.0002,
      "loss": 0.7704,
      "step": 26300
    },
    {
      "epoch": 10.56,
      "grad_norm": 4.615940570831299,
      "learning_rate": 0.0002,
      "loss": 0.7957,
      "step": 26400
    },
    {
      "epoch": 10.6,
      "grad_norm": 1.4114075899124146,
      "learning_rate": 0.0002,
      "loss": 0.7613,
      "step": 26500
    },
    {
      "epoch": 10.64,
      "grad_norm": 1.1253297328948975,
      "learning_rate": 0.0002,
      "loss": 0.8119,
      "step": 26600
    },
    {
      "epoch": 10.68,
      "grad_norm": 1.9775581359863281,
      "learning_rate": 0.0002,
      "loss": 0.8083,
      "step": 26700
    },
    {
      "epoch": 10.72,
      "grad_norm": 0.6344081163406372,
      "learning_rate": 0.0002,
      "loss": 0.8078,
      "step": 26800
    },
    {
      "epoch": 10.76,
      "grad_norm": 1.287255048751831,
      "learning_rate": 0.0002,
      "loss": 0.7732,
      "step": 26900
    },
    {
      "epoch": 10.8,
      "grad_norm": 0.826394259929657,
      "learning_rate": 0.0002,
      "loss": 0.7874,
      "step": 27000
    },
    {
      "epoch": 10.84,
      "grad_norm": 1.9858311414718628,
      "learning_rate": 0.0002,
      "loss": 0.7672,
      "step": 27100
    },
    {
      "epoch": 10.88,
      "grad_norm": 2.411587715148926,
      "learning_rate": 0.0002,
      "loss": 0.7926,
      "step": 27200
    },
    {
      "epoch": 10.92,
      "grad_norm": 1.1117106676101685,
      "learning_rate": 0.0002,
      "loss": 0.8031,
      "step": 27300
    },
    {
      "epoch": 10.96,
      "grad_norm": 0.9955012798309326,
      "learning_rate": 0.0002,
      "loss": 0.8018,
      "step": 27400
    },
    {
      "epoch": 11.0,
      "grad_norm": 6.443946838378906,
      "learning_rate": 0.0002,
      "loss": 0.7974,
      "step": 27500
    },
    {
      "epoch": 11.04,
      "grad_norm": 0.9066771864891052,
      "learning_rate": 0.0002,
      "loss": 0.6354,
      "step": 27600
    },
    {
      "epoch": 11.08,
      "grad_norm": 1.2832303047180176,
      "learning_rate": 0.0002,
      "loss": 0.6669,
      "step": 27700
    },
    {
      "epoch": 11.12,
      "grad_norm": 1.3047817945480347,
      "learning_rate": 0.0002,
      "loss": 0.6408,
      "step": 27800
    },
    {
      "epoch": 11.16,
      "grad_norm": 1.2970690727233887,
      "learning_rate": 0.0002,
      "loss": 0.6777,
      "step": 27900
    },
    {
      "epoch": 11.2,
      "grad_norm": 1.2775633335113525,
      "learning_rate": 0.0002,
      "loss": 0.6366,
      "step": 28000
    },
    {
      "epoch": 11.24,
      "grad_norm": 1.1797577142715454,
      "learning_rate": 0.0002,
      "loss": 0.6618,
      "step": 28100
    },
    {
      "epoch": 11.28,
      "grad_norm": 0.4641984701156616,
      "learning_rate": 0.0002,
      "loss": 0.714,
      "step": 28200
    },
    {
      "epoch": 11.32,
      "grad_norm": 5.440238952636719,
      "learning_rate": 0.0002,
      "loss": 0.707,
      "step": 28300
    },
    {
      "epoch": 11.36,
      "grad_norm": 2.7342915534973145,
      "learning_rate": 0.0002,
      "loss": 0.7155,
      "step": 28400
    },
    {
      "epoch": 11.4,
      "grad_norm": 1.025922179222107,
      "learning_rate": 0.0002,
      "loss": 0.6761,
      "step": 28500
    },
    {
      "epoch": 11.44,
      "grad_norm": 1.4162850379943848,
      "learning_rate": 0.0002,
      "loss": 0.7079,
      "step": 28600
    },
    {
      "epoch": 11.48,
      "grad_norm": 1.673281192779541,
      "learning_rate": 0.0002,
      "loss": 0.6957,
      "step": 28700
    },
    {
      "epoch": 11.52,
      "grad_norm": 1.6640900373458862,
      "learning_rate": 0.0002,
      "loss": 0.7616,
      "step": 28800
    },
    {
      "epoch": 11.56,
      "grad_norm": 0.5658434629440308,
      "learning_rate": 0.0002,
      "loss": 0.6939,
      "step": 28900
    },
    {
      "epoch": 11.6,
      "grad_norm": 0.6061828136444092,
      "learning_rate": 0.0002,
      "loss": 0.7162,
      "step": 29000
    },
    {
      "epoch": 11.64,
      "grad_norm": 1.6504275798797607,
      "learning_rate": 0.0002,
      "loss": 0.7266,
      "step": 29100
    },
    {
      "epoch": 11.68,
      "grad_norm": 2.936732530593872,
      "learning_rate": 0.0002,
      "loss": 0.7053,
      "step": 29200
    },
    {
      "epoch": 11.72,
      "grad_norm": 1.2476632595062256,
      "learning_rate": 0.0002,
      "loss": 0.6966,
      "step": 29300
    },
    {
      "epoch": 11.76,
      "grad_norm": 0.8274354934692383,
      "learning_rate": 0.0002,
      "loss": 0.6928,
      "step": 29400
    },
    {
      "epoch": 11.8,
      "grad_norm": 1.6214333772659302,
      "learning_rate": 0.0002,
      "loss": 0.7288,
      "step": 29500
    },
    {
      "epoch": 11.84,
      "grad_norm": 1.2420477867126465,
      "learning_rate": 0.0002,
      "loss": 0.7501,
      "step": 29600
    },
    {
      "epoch": 11.88,
      "grad_norm": 1.3407083749771118,
      "learning_rate": 0.0002,
      "loss": 0.7472,
      "step": 29700
    },
    {
      "epoch": 11.92,
      "grad_norm": 2.469486951828003,
      "learning_rate": 0.0002,
      "loss": 0.7077,
      "step": 29800
    },
    {
      "epoch": 11.96,
      "grad_norm": 2.422231912612915,
      "learning_rate": 0.0002,
      "loss": 0.7587,
      "step": 29900
    },
    {
      "epoch": 12.0,
      "grad_norm": 2.655271053314209,
      "learning_rate": 0.0002,
      "loss": 0.734,
      "step": 30000
    },
    {
      "epoch": 12.04,
      "grad_norm": 3.5966265201568604,
      "learning_rate": 0.0002,
      "loss": 0.5812,
      "step": 30100
    },
    {
      "epoch": 12.08,
      "grad_norm": 0.46480792760849,
      "learning_rate": 0.0002,
      "loss": 0.589,
      "step": 30200
    },
    {
      "epoch": 12.12,
      "grad_norm": 0.8438324928283691,
      "learning_rate": 0.0002,
      "loss": 0.6005,
      "step": 30300
    },
    {
      "epoch": 12.16,
      "grad_norm": 1.9844694137573242,
      "learning_rate": 0.0002,
      "loss": 0.6312,
      "step": 30400
    },
    {
      "epoch": 12.2,
      "grad_norm": 1.1500381231307983,
      "learning_rate": 0.0002,
      "loss": 0.621,
      "step": 30500
    },
    {
      "epoch": 12.24,
      "grad_norm": 1.3909460306167603,
      "learning_rate": 0.0002,
      "loss": 0.6353,
      "step": 30600
    },
    {
      "epoch": 12.28,
      "grad_norm": 3.504803419113159,
      "learning_rate": 0.0002,
      "loss": 0.6191,
      "step": 30700
    },
    {
      "epoch": 12.32,
      "grad_norm": 1.116080641746521,
      "learning_rate": 0.0002,
      "loss": 0.628,
      "step": 30800
    },
    {
      "epoch": 12.36,
      "grad_norm": 0.9176204800605774,
      "learning_rate": 0.0002,
      "loss": 0.6314,
      "step": 30900
    },
    {
      "epoch": 12.4,
      "grad_norm": 1.0510369539260864,
      "learning_rate": 0.0002,
      "loss": 0.6532,
      "step": 31000
    },
    {
      "epoch": 12.44,
      "grad_norm": 0.5242018103599548,
      "learning_rate": 0.0002,
      "loss": 0.6422,
      "step": 31100
    },
    {
      "epoch": 12.48,
      "grad_norm": 5.052489280700684,
      "learning_rate": 0.0002,
      "loss": 0.64,
      "step": 31200
    },
    {
      "epoch": 12.52,
      "grad_norm": 0.5426860451698303,
      "learning_rate": 0.0002,
      "loss": 0.6488,
      "step": 31300
    },
    {
      "epoch": 12.56,
      "grad_norm": 0.7503352165222168,
      "learning_rate": 0.0002,
      "loss": 0.6764,
      "step": 31400
    },
    {
      "epoch": 12.6,
      "grad_norm": 2.2711548805236816,
      "learning_rate": 0.0002,
      "loss": 0.6177,
      "step": 31500
    },
    {
      "epoch": 12.64,
      "grad_norm": 1.3384021520614624,
      "learning_rate": 0.0002,
      "loss": 0.6425,
      "step": 31600
    },
    {
      "epoch": 12.68,
      "grad_norm": 1.098314642906189,
      "learning_rate": 0.0002,
      "loss": 0.6453,
      "step": 31700
    },
    {
      "epoch": 12.72,
      "grad_norm": 0.8752951622009277,
      "learning_rate": 0.0002,
      "loss": 0.6724,
      "step": 31800
    },
    {
      "epoch": 12.76,
      "grad_norm": 1.700907826423645,
      "learning_rate": 0.0002,
      "loss": 0.6707,
      "step": 31900
    },
    {
      "epoch": 12.8,
      "grad_norm": 2.7010860443115234,
      "learning_rate": 0.0002,
      "loss": 0.6551,
      "step": 32000
    },
    {
      "epoch": 12.84,
      "grad_norm": 0.9509829878807068,
      "learning_rate": 0.0002,
      "loss": 0.6766,
      "step": 32100
    },
    {
      "epoch": 12.88,
      "grad_norm": 1.8936227560043335,
      "learning_rate": 0.0002,
      "loss": 0.6889,
      "step": 32200
    },
    {
      "epoch": 12.92,
      "grad_norm": 3.870959997177124,
      "learning_rate": 0.0002,
      "loss": 0.6526,
      "step": 32300
    },
    {
      "epoch": 12.96,
      "grad_norm": 0.8048538565635681,
      "learning_rate": 0.0002,
      "loss": 0.6658,
      "step": 32400
    },
    {
      "epoch": 13.0,
      "grad_norm": 3.0256259441375732,
      "learning_rate": 0.0002,
      "loss": 0.6521,
      "step": 32500
    },
    {
      "epoch": 13.04,
      "grad_norm": 0.5219417810440063,
      "learning_rate": 0.0002,
      "loss": 0.5415,
      "step": 32600
    },
    {
      "epoch": 13.08,
      "grad_norm": 0.6729668378829956,
      "learning_rate": 0.0002,
      "loss": 0.5442,
      "step": 32700
    },
    {
      "epoch": 13.12,
      "grad_norm": 5.041486740112305,
      "learning_rate": 0.0002,
      "loss": 0.5794,
      "step": 32800
    },
    {
      "epoch": 13.16,
      "grad_norm": 0.7740932106971741,
      "learning_rate": 0.0002,
      "loss": 0.5739,
      "step": 32900
    },
    {
      "epoch": 13.2,
      "grad_norm": 2.4038619995117188,
      "learning_rate": 0.0002,
      "loss": 0.5527,
      "step": 33000
    }
  ],
  "logging_steps": 100,
  "max_steps": 125000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 50,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 8.330326307619963e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}