{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.21,
  "eval_steps": 500,
  "global_step": 455000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0002,
      "grad_norm": 1.603668212890625,
      "learning_rate": 0.0001999604,
      "loss": 2.0926,
      "step": 100
    },
    {
      "epoch": 0.0004,
      "grad_norm": 1.1335899829864502,
      "learning_rate": 0.0001999204,
      "loss": 1.5915,
      "step": 200
    },
    {
      "epoch": 0.0006,
      "grad_norm": 0.9146980047225952,
      "learning_rate": 0.00019988040000000002,
      "loss": 1.413,
      "step": 300
    },
    {
      "epoch": 0.0008,
      "grad_norm": 1.1228628158569336,
      "learning_rate": 0.0001998404,
      "loss": 1.3208,
      "step": 400
    },
    {
      "epoch": 0.001,
      "grad_norm": 0.623985230922699,
      "learning_rate": 0.0001998004,
      "loss": 1.2557,
      "step": 500
    },
    {
      "epoch": 0.0012,
      "grad_norm": 0.7038842439651489,
      "learning_rate": 0.0001997604,
      "loss": 1.1387,
      "step": 600
    },
    {
      "epoch": 0.0014,
      "grad_norm": 0.820369303226471,
      "learning_rate": 0.00019972040000000002,
      "loss": 1.1653,
      "step": 700
    },
    {
      "epoch": 0.0016,
      "grad_norm": 0.6667912006378174,
      "learning_rate": 0.0001996804,
      "loss": 1.1342,
      "step": 800
    },
    {
      "epoch": 0.0018,
      "grad_norm": 0.616651713848114,
      "learning_rate": 0.0001996404,
      "loss": 1.1548,
      "step": 900
    },
    {
      "epoch": 0.002,
      "grad_norm": 0.5994811058044434,
      "learning_rate": 0.0001996004,
      "loss": 1.0464,
      "step": 1000
    },
    {
      "epoch": 0.0022,
      "grad_norm": 0.5694602131843567,
      "learning_rate": 0.0001995604,
      "loss": 1.1394,
      "step": 1100
    },
    {
      "epoch": 0.0024,
      "grad_norm": 0.5835832357406616,
      "learning_rate": 0.00019952040000000002,
      "loss": 1.0209,
      "step": 1200
    },
    {
      "epoch": 0.0026,
      "grad_norm": 1.4923994541168213,
      "learning_rate": 0.0001994804,
      "loss": 1.0708,
      "step": 1300
    },
    {
      "epoch": 0.0028,
      "grad_norm": 0.669029712677002,
      "learning_rate": 0.00019944040000000003,
      "loss": 0.9359,
      "step": 1400
    },
    {
      "epoch": 0.003,
      "grad_norm": 1.1236016750335693,
      "learning_rate": 0.00019940040000000002,
      "loss": 0.9963,
      "step": 1500
    },
    {
      "epoch": 0.0032,
      "grad_norm": 0.7666842341423035,
      "learning_rate": 0.0001993604,
      "loss": 1.0306,
      "step": 1600
    },
    {
      "epoch": 0.0034,
      "grad_norm": 1.5545086860656738,
      "learning_rate": 0.0001993204,
      "loss": 0.9544,
      "step": 1700
    },
    {
      "epoch": 0.0036,
      "grad_norm": 0.7571123242378235,
      "learning_rate": 0.0001992804,
      "loss": 0.9473,
      "step": 1800
    },
    {
      "epoch": 0.0038,
      "grad_norm": 0.5189706683158875,
      "learning_rate": 0.00019924040000000002,
      "loss": 0.9743,
      "step": 1900
    },
    {
      "epoch": 0.004,
      "grad_norm": 0.9225242137908936,
      "learning_rate": 0.00019920040000000002,
      "loss": 0.9029,
      "step": 2000
    },
    {
      "epoch": 0.0042,
      "grad_norm": 0.6427626609802246,
      "learning_rate": 0.0001991604,
      "loss": 0.9111,
      "step": 2100
    },
    {
      "epoch": 0.0044,
      "grad_norm": 0.5414828658103943,
      "learning_rate": 0.0001991204,
      "loss": 0.9591,
      "step": 2200
    },
    {
      "epoch": 0.0046,
      "grad_norm": 0.5731277465820312,
      "learning_rate": 0.0001990804,
      "loss": 0.9327,
      "step": 2300
    },
    {
      "epoch": 0.0048,
      "grad_norm": 0.6296040415763855,
      "learning_rate": 0.00019904040000000002,
      "loss": 0.8148,
      "step": 2400
    },
    {
      "epoch": 0.005,
      "grad_norm": 0.5396831631660461,
      "learning_rate": 0.0001990004,
      "loss": 0.8056,
      "step": 2500
    },
    {
      "epoch": 0.0052,
      "grad_norm": 0.5632336735725403,
      "learning_rate": 0.0001989604,
      "loss": 0.8224,
      "step": 2600
    },
    {
      "epoch": 0.0054,
      "grad_norm": 0.7668570280075073,
      "learning_rate": 0.0001989204,
      "loss": 0.8137,
      "step": 2700
    },
    {
      "epoch": 0.0056,
      "grad_norm": 0.5585981607437134,
      "learning_rate": 0.00019888040000000002,
      "loss": 0.8658,
      "step": 2800
    },
    {
      "epoch": 0.0058,
      "grad_norm": 0.5207762122154236,
      "learning_rate": 0.00019884040000000001,
      "loss": 0.796,
      "step": 2900
    },
    {
      "epoch": 0.006,
      "grad_norm": 0.6384051442146301,
      "learning_rate": 0.0001988004,
      "loss": 0.979,
      "step": 3000
    },
    {
      "epoch": 0.0062,
      "grad_norm": 0.4824763834476471,
      "learning_rate": 0.0001987604,
      "loss": 0.7749,
      "step": 3100
    },
    {
      "epoch": 0.0064,
      "grad_norm": 0.5262443423271179,
      "learning_rate": 0.0001987204,
      "loss": 0.7486,
      "step": 3200
    },
    {
      "epoch": 0.0066,
      "grad_norm": 3.371579170227051,
      "learning_rate": 0.00019868040000000002,
      "loss": 0.8155,
      "step": 3300
    },
    {
      "epoch": 0.0068,
      "grad_norm": 0.46106863021850586,
      "learning_rate": 0.0001986404,
      "loss": 0.7683,
      "step": 3400
    },
    {
      "epoch": 0.007,
      "grad_norm": 0.5194998979568481,
      "learning_rate": 0.00019860040000000003,
      "loss": 0.7919,
      "step": 3500
    },
    {
      "epoch": 0.0072,
      "grad_norm": 0.5358483195304871,
      "learning_rate": 0.0001985604,
      "loss": 0.7655,
      "step": 3600
    },
    {
      "epoch": 0.0074,
      "grad_norm": 0.4279918372631073,
      "learning_rate": 0.00019852040000000002,
      "loss": 0.7611,
      "step": 3700
    },
    {
      "epoch": 0.0076,
      "grad_norm": 0.3322402834892273,
      "learning_rate": 0.0001984804,
      "loss": 0.8131,
      "step": 3800
    },
    {
      "epoch": 0.0078,
      "grad_norm": 0.5196714401245117,
      "learning_rate": 0.0001984404,
      "loss": 0.7348,
      "step": 3900
    },
    {
      "epoch": 0.008,
      "grad_norm": 0.6269809603691101,
      "learning_rate": 0.00019840040000000003,
      "loss": 0.7274,
      "step": 4000
    },
    {
      "epoch": 0.0082,
      "grad_norm": 0.7291073203086853,
      "learning_rate": 0.0001983604,
      "loss": 0.8001,
      "step": 4100
    },
    {
      "epoch": 0.0084,
      "grad_norm": 1.0613459348678589,
      "learning_rate": 0.0001983204,
      "loss": 0.7169,
      "step": 4200
    },
    {
      "epoch": 0.0086,
      "grad_norm": 0.46023017168045044,
      "learning_rate": 0.0001982804,
      "loss": 0.7456,
      "step": 4300
    },
    {
      "epoch": 0.0088,
      "grad_norm": 0.3527376651763916,
      "learning_rate": 0.00019824040000000003,
      "loss": 0.699,
      "step": 4400
    },
    {
      "epoch": 0.009,
      "grad_norm": 0.5758853554725647,
      "learning_rate": 0.00019820040000000002,
      "loss": 0.7226,
      "step": 4500
    },
    {
      "epoch": 0.0092,
      "grad_norm": 0.500385582447052,
      "learning_rate": 0.00019816040000000001,
      "loss": 0.753,
      "step": 4600
    },
    {
      "epoch": 0.0094,
      "grad_norm": 0.8873578906059265,
      "learning_rate": 0.0001981204,
      "loss": 0.7047,
      "step": 4700
    },
    {
      "epoch": 0.0096,
      "grad_norm": 0.6691761612892151,
      "learning_rate": 0.0001980804,
      "loss": 0.7139,
      "step": 4800
    },
    {
      "epoch": 0.0098,
      "grad_norm": 0.593289852142334,
      "learning_rate": 0.00019804040000000002,
      "loss": 0.7075,
      "step": 4900
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.7012218236923218,
      "learning_rate": 0.00019800040000000002,
      "loss": 0.6878,
      "step": 5000
    },
    {
      "epoch": 0.0102,
      "grad_norm": 0.5074142217636108,
      "learning_rate": 0.0001979604,
      "loss": 0.7574,
      "step": 5100
    },
    {
      "epoch": 0.0104,
      "grad_norm": 0.9699320197105408,
      "learning_rate": 0.0001979204,
      "loss": 0.6619,
      "step": 5200
    },
    {
      "epoch": 0.0106,
      "grad_norm": 0.44478604197502136,
      "learning_rate": 0.0001978804,
      "loss": 0.7029,
      "step": 5300
    },
    {
      "epoch": 0.0108,
      "grad_norm": 0.43934938311576843,
      "learning_rate": 0.00019784040000000002,
      "loss": 0.6991,
      "step": 5400
    },
    {
      "epoch": 0.011,
      "grad_norm": 0.4505891501903534,
      "learning_rate": 0.0001978004,
      "loss": 0.7019,
      "step": 5500
    },
    {
      "epoch": 0.0112,
      "grad_norm": 0.40732213854789734,
      "learning_rate": 0.0001977604,
      "loss": 0.7085,
      "step": 5600
    },
    {
      "epoch": 0.0114,
      "grad_norm": 0.5004173517227173,
      "learning_rate": 0.0001977204,
      "loss": 0.6757,
      "step": 5700
    },
    {
      "epoch": 0.0116,
      "grad_norm": 0.5011471509933472,
      "learning_rate": 0.00019768040000000002,
      "loss": 0.6752,
      "step": 5800
    },
    {
      "epoch": 0.0118,
      "grad_norm": 0.3992891311645508,
      "learning_rate": 0.0001976404,
      "loss": 0.6733,
      "step": 5900
    },
    {
      "epoch": 0.012,
      "grad_norm": 0.3974108099937439,
      "learning_rate": 0.0001976004,
      "loss": 0.6737,
      "step": 6000
    },
    {
      "epoch": 0.0122,
      "grad_norm": 0.4213123917579651,
      "learning_rate": 0.0001975604,
      "loss": 0.7769,
      "step": 6100
    },
    {
      "epoch": 0.0124,
      "grad_norm": 0.4016417860984802,
      "learning_rate": 0.0001975204,
      "loss": 0.6731,
      "step": 6200
    },
    {
      "epoch": 0.0126,
      "grad_norm": 0.45475897192955017,
      "learning_rate": 0.00019748040000000001,
      "loss": 0.6566,
      "step": 6300
    },
    {
      "epoch": 0.0128,
      "grad_norm": 0.298532098531723,
      "learning_rate": 0.0001974404,
      "loss": 0.6221,
      "step": 6400
    },
    {
      "epoch": 0.013,
      "grad_norm": 0.36682987213134766,
      "learning_rate": 0.00019740040000000003,
      "loss": 0.6402,
      "step": 6500
    },
    {
      "epoch": 0.0132,
      "grad_norm": 0.35242322087287903,
      "learning_rate": 0.0001973604,
      "loss": 0.6069,
      "step": 6600
    },
    {
      "epoch": 0.0134,
      "grad_norm": 0.49721652269363403,
      "learning_rate": 0.00019732040000000002,
      "loss": 0.6847,
      "step": 6700
    },
    {
      "epoch": 0.0136,
      "grad_norm": 0.371697336435318,
      "learning_rate": 0.0001972804,
      "loss": 0.617,
      "step": 6800
    },
    {
      "epoch": 0.0138,
      "grad_norm": 0.498140424489975,
      "learning_rate": 0.0001972404,
      "loss": 0.7371,
      "step": 6900
    },
    {
      "epoch": 0.014,
      "grad_norm": 0.4353884160518646,
      "learning_rate": 0.00019720040000000002,
      "loss": 0.6746,
      "step": 7000
    },
    {
      "epoch": 0.0142,
      "grad_norm": 0.47410452365875244,
      "learning_rate": 0.0001971604,
      "loss": 0.6462,
      "step": 7100
    },
    {
      "epoch": 0.0144,
      "grad_norm": 0.558781087398529,
      "learning_rate": 0.0001971204,
      "loss": 0.6459,
      "step": 7200
    },
    {
      "epoch": 0.0146,
      "grad_norm": 0.41540494561195374,
      "learning_rate": 0.0001970804,
      "loss": 0.6269,
      "step": 7300
    },
    {
      "epoch": 0.0148,
      "grad_norm": 0.5316945910453796,
      "learning_rate": 0.00019704040000000003,
      "loss": 0.6097,
      "step": 7400
    },
    {
      "epoch": 0.015,
      "grad_norm": 0.5045424103736877,
      "learning_rate": 0.00019700040000000002,
      "loss": 0.6496,
      "step": 7500
    },
    {
      "epoch": 0.0152,
      "grad_norm": 0.3779171407222748,
      "learning_rate": 0.0001969604,
      "loss": 0.632,
      "step": 7600
    },
    {
      "epoch": 0.0154,
      "grad_norm": 0.31508123874664307,
      "learning_rate": 0.0001969204,
      "loss": 0.595,
      "step": 7700
    },
    {
      "epoch": 0.0156,
      "grad_norm": 0.5666154623031616,
      "learning_rate": 0.0001968804,
      "loss": 0.6193,
      "step": 7800
    },
    {
      "epoch": 0.0158,
      "grad_norm": 0.36079517006874084,
      "learning_rate": 0.00019684040000000002,
      "loss": 0.6154,
      "step": 7900
    },
    {
      "epoch": 0.016,
      "grad_norm": 0.716064989566803,
      "learning_rate": 0.00019680040000000001,
      "loss": 0.6018,
      "step": 8000
    },
    {
      "epoch": 0.0162,
      "grad_norm": 0.3534316122531891,
      "learning_rate": 0.0001967604,
      "loss": 0.6298,
      "step": 8100
    },
    {
      "epoch": 0.0164,
      "grad_norm": 0.31048017740249634,
      "learning_rate": 0.0001967204,
      "loss": 0.6118,
      "step": 8200
    },
    {
      "epoch": 0.0166,
      "grad_norm": 0.7001683115959167,
      "learning_rate": 0.00019668040000000002,
      "loss": 0.6154,
      "step": 8300
    },
    {
      "epoch": 0.0168,
      "grad_norm": 0.577167809009552,
      "learning_rate": 0.00019664040000000002,
      "loss": 0.6081,
      "step": 8400
    },
    {
      "epoch": 0.017,
      "grad_norm": 0.48630014061927795,
      "learning_rate": 0.0001966004,
      "loss": 0.6225,
      "step": 8500
    },
    {
      "epoch": 0.0172,
      "grad_norm": 0.3509364426136017,
      "learning_rate": 0.0001965604,
      "loss": 0.6283,
      "step": 8600
    },
    {
      "epoch": 0.0174,
      "grad_norm": 0.5807685256004333,
      "learning_rate": 0.0001965204,
      "loss": 0.5904,
      "step": 8700
    },
    {
      "epoch": 0.0176,
      "grad_norm": 0.40310025215148926,
      "learning_rate": 0.00019648040000000002,
      "loss": 0.6043,
      "step": 8800
    },
    {
      "epoch": 0.0178,
      "grad_norm": 0.4283163845539093,
      "learning_rate": 0.0001964404,
      "loss": 0.5648,
      "step": 8900
    },
    {
      "epoch": 0.018,
      "grad_norm": 0.33734825253486633,
      "learning_rate": 0.0001964004,
      "loss": 0.5728,
      "step": 9000
    },
    {
      "epoch": 0.0182,
      "grad_norm": 0.2931211292743683,
      "learning_rate": 0.0001963604,
      "loss": 0.5937,
      "step": 9100
    },
    {
      "epoch": 0.0184,
      "grad_norm": 0.4612201154232025,
      "learning_rate": 0.0001963204,
      "loss": 0.5934,
      "step": 9200
    },
    {
      "epoch": 0.0186,
      "grad_norm": 0.3644779622554779,
      "learning_rate": 0.0001962804,
      "loss": 0.5766,
      "step": 9300
    },
    {
      "epoch": 0.0188,
      "grad_norm": 0.30560925602912903,
      "learning_rate": 0.0001962404,
      "loss": 0.5464,
      "step": 9400
    },
    {
      "epoch": 0.019,
      "grad_norm": 0.6216826438903809,
      "learning_rate": 0.00019620040000000003,
      "loss": 0.6206,
      "step": 9500
    },
    {
      "epoch": 0.0192,
      "grad_norm": 0.3877085745334625,
      "learning_rate": 0.0001961604,
      "loss": 0.5825,
      "step": 9600
    },
    {
      "epoch": 0.0194,
      "grad_norm": 0.4004894495010376,
      "learning_rate": 0.00019612040000000001,
      "loss": 0.5918,
      "step": 9700
    },
    {
      "epoch": 0.0196,
      "grad_norm": 0.49157753586769104,
      "learning_rate": 0.0001960804,
      "loss": 0.5891,
      "step": 9800
    },
    {
      "epoch": 0.0198,
      "grad_norm": 1.031232237815857,
      "learning_rate": 0.0001960404,
      "loss": 0.5748,
      "step": 9900
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.36149775981903076,
      "learning_rate": 0.00019600040000000002,
      "loss": 0.5844,
      "step": 10000
    },
    {
      "epoch": 0.0202,
      "grad_norm": 0.29936593770980835,
      "learning_rate": 0.0001959604,
      "loss": 0.5687,
      "step": 10100
    },
    {
      "epoch": 0.0204,
      "grad_norm": 4.51532506942749,
      "learning_rate": 0.0001959204,
      "loss": 0.6077,
      "step": 10200
    },
    {
      "epoch": 0.0206,
      "grad_norm": 0.44001829624176025,
      "learning_rate": 0.0001958804,
      "loss": 0.6138,
      "step": 10300
    },
    {
      "epoch": 0.0208,
      "grad_norm": 0.26877814531326294,
      "learning_rate": 0.00019584040000000002,
      "loss": 0.5787,
      "step": 10400
    },
    {
      "epoch": 0.021,
      "grad_norm": 0.441687673330307,
      "learning_rate": 0.00019580040000000002,
      "loss": 0.5353,
      "step": 10500
    },
    {
      "epoch": 0.0212,
      "grad_norm": 0.4635355472564697,
      "learning_rate": 0.0001957604,
      "loss": 0.5686,
      "step": 10600
    },
    {
      "epoch": 0.0214,
      "grad_norm": 0.3567267060279846,
      "learning_rate": 0.0001957204,
      "loss": 0.5363,
      "step": 10700
    },
    {
      "epoch": 0.0216,
      "grad_norm": 0.5282173156738281,
      "learning_rate": 0.0001956804,
      "loss": 0.582,
      "step": 10800
    },
    {
      "epoch": 0.0218,
      "grad_norm": 0.32945743203163147,
      "learning_rate": 0.00019564040000000002,
      "loss": 0.5695,
      "step": 10900
    },
    {
      "epoch": 0.022,
      "grad_norm": 1.3247750997543335,
      "learning_rate": 0.0001956004,
      "loss": 0.5665,
      "step": 11000
    },
    {
      "epoch": 0.0222,
      "grad_norm": 0.3688965439796448,
      "learning_rate": 0.0001955604,
      "loss": 0.5835,
      "step": 11100
    },
    {
      "epoch": 0.0224,
      "grad_norm": 0.4591335356235504,
      "learning_rate": 0.0001955204,
      "loss": 0.5638,
      "step": 11200
    },
    {
      "epoch": 0.0226,
      "grad_norm": 0.4654744267463684,
      "learning_rate": 0.00019548040000000002,
      "loss": 0.5821,
      "step": 11300
    },
    {
      "epoch": 0.0228,
      "grad_norm": 1.729272484779358,
      "learning_rate": 0.0001954404,
      "loss": 0.6013,
      "step": 11400
    },
    {
      "epoch": 0.023,
      "grad_norm": 0.3169005513191223,
      "learning_rate": 0.0001954004,
      "loss": 0.5678,
      "step": 11500
    },
    {
      "epoch": 0.0232,
      "grad_norm": 0.4935179650783539,
      "learning_rate": 0.0001953604,
      "loss": 0.5517,
      "step": 11600
    },
    {
      "epoch": 0.0234,
      "grad_norm": 0.3395957350730896,
      "learning_rate": 0.0001953204,
      "loss": 0.5387,
      "step": 11700
    },
    {
      "epoch": 0.0236,
      "grad_norm": 0.4479553997516632,
      "learning_rate": 0.00019528040000000001,
      "loss": 0.5748,
      "step": 11800
    },
    {
      "epoch": 0.0238,
      "grad_norm": 0.36801132559776306,
      "learning_rate": 0.0001952404,
      "loss": 0.6047,
      "step": 11900
    },
    {
      "epoch": 0.024,
      "grad_norm": 0.5720965266227722,
      "learning_rate": 0.00019520040000000003,
      "loss": 0.5458,
      "step": 12000
    },
    {
      "epoch": 0.0242,
      "grad_norm": 0.37826624512672424,
      "learning_rate": 0.0001951604,
      "loss": 0.5512,
      "step": 12100
    },
    {
      "epoch": 0.0244,
      "grad_norm": 0.39435476064682007,
      "learning_rate": 0.00019512040000000002,
      "loss": 0.5388,
      "step": 12200
    },
    {
      "epoch": 0.0246,
      "grad_norm": 0.4711388349533081,
      "learning_rate": 0.0001950804,
      "loss": 0.535,
      "step": 12300
    },
    {
      "epoch": 0.0248,
      "grad_norm": 0.38097721338272095,
      "learning_rate": 0.0001950404,
      "loss": 0.5675,
      "step": 12400
    },
    {
      "epoch": 0.025,
      "grad_norm": 0.3961537778377533,
      "learning_rate": 0.00019500040000000002,
      "loss": 0.5727,
      "step": 12500
    },
    {
      "epoch": 0.0252,
      "grad_norm": 0.5444319248199463,
      "learning_rate": 0.0001949604,
      "loss": 0.5879,
      "step": 12600
    },
    {
      "epoch": 0.0254,
      "grad_norm": 0.5849190354347229,
      "learning_rate": 0.0001949204,
      "loss": 0.5635,
      "step": 12700
    },
    {
      "epoch": 0.0256,
      "grad_norm": 0.6113494038581848,
      "learning_rate": 0.0001948804,
      "loss": 0.5467,
      "step": 12800
    },
    {
      "epoch": 0.0258,
      "grad_norm": 0.4276942312717438,
      "learning_rate": 0.0001948404,
      "loss": 0.5554,
      "step": 12900
    },
    {
      "epoch": 0.026,
      "grad_norm": 0.48870015144348145,
      "learning_rate": 0.00019480040000000002,
      "loss": 0.5986,
      "step": 13000
    },
    {
      "epoch": 0.0262,
      "grad_norm": 0.27192604541778564,
      "learning_rate": 0.00019476039999999999,
      "loss": 0.5551,
      "step": 13100
    },
    {
      "epoch": 0.0264,
      "grad_norm": 0.6044579148292542,
      "learning_rate": 0.0001947204,
      "loss": 0.5477,
      "step": 13200
    },
    {
      "epoch": 0.0266,
      "grad_norm": 0.4254133105278015,
      "learning_rate": 0.0001946804,
      "loss": 0.547,
      "step": 13300
    },
    {
      "epoch": 0.0268,
      "grad_norm": 1.369423747062683,
      "learning_rate": 0.00019464040000000002,
      "loss": 0.5492,
      "step": 13400
    },
    {
      "epoch": 0.027,
      "grad_norm": 0.38343751430511475,
      "learning_rate": 0.00019460040000000001,
      "loss": 0.5275,
      "step": 13500
    },
    {
      "epoch": 0.0272,
      "grad_norm": 0.45244088768959045,
      "learning_rate": 0.0001945604,
      "loss": 0.4827,
      "step": 13600
    },
    {
      "epoch": 0.0274,
      "grad_norm": 0.3227088451385498,
      "learning_rate": 0.0001945204,
      "loss": 0.4863,
      "step": 13700
    },
    {
      "epoch": 0.0276,
      "grad_norm": 0.45713531970977783,
      "learning_rate": 0.0001944804,
      "loss": 0.4933,
      "step": 13800
    },
    {
      "epoch": 0.0278,
      "grad_norm": 0.42994385957717896,
      "learning_rate": 0.00019444040000000002,
      "loss": 0.484,
      "step": 13900
    },
    {
      "epoch": 0.028,
      "grad_norm": 0.3284667134284973,
      "learning_rate": 0.0001944004,
      "loss": 0.4956,
      "step": 14000
    },
    {
      "epoch": 0.0282,
      "grad_norm": 0.44729650020599365,
      "learning_rate": 0.00019436040000000003,
      "loss": 0.4595,
      "step": 14100
    },
    {
      "epoch": 0.0284,
      "grad_norm": 0.31246569752693176,
      "learning_rate": 0.0001943204,
      "loss": 0.4796,
      "step": 14200
    },
    {
      "epoch": 0.0286,
      "grad_norm": 0.46601220965385437,
      "learning_rate": 0.00019428040000000002,
      "loss": 0.4956,
      "step": 14300
    },
    {
      "epoch": 0.0288,
      "grad_norm": 0.33735036849975586,
      "learning_rate": 0.0001942404,
      "loss": 0.4689,
      "step": 14400
    },
    {
      "epoch": 0.029,
      "grad_norm": 0.44884735345840454,
      "learning_rate": 0.0001942004,
      "loss": 0.4913,
      "step": 14500
    },
    {
      "epoch": 0.0292,
      "grad_norm": 1.72175133228302,
      "learning_rate": 0.00019416040000000003,
      "loss": 0.4932,
      "step": 14600
    },
    {
      "epoch": 0.0294,
      "grad_norm": 0.8482571244239807,
      "learning_rate": 0.0001941204,
      "loss": 0.4834,
      "step": 14700
    },
    {
      "epoch": 0.0296,
      "grad_norm": 0.35360008478164673,
      "learning_rate": 0.0001940804,
      "loss": 0.498,
      "step": 14800
    },
    {
      "epoch": 0.0298,
      "grad_norm": 0.32927367091178894,
      "learning_rate": 0.0001940404,
      "loss": 0.4824,
      "step": 14900
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.4251364469528198,
      "learning_rate": 0.00019400040000000003,
      "loss": 0.4941,
      "step": 15000
    },
    {
      "epoch": 0.0302,
      "grad_norm": 0.3267940878868103,
      "learning_rate": 0.00019396040000000002,
      "loss": 0.4943,
      "step": 15100
    },
    {
      "epoch": 0.0304,
      "grad_norm": 0.29757174849510193,
      "learning_rate": 0.00019392040000000001,
      "loss": 0.4856,
      "step": 15200
    },
    {
      "epoch": 0.0306,
      "grad_norm": 0.4625142216682434,
      "learning_rate": 0.0001938804,
      "loss": 0.4759,
      "step": 15300
    },
    {
      "epoch": 0.0308,
      "grad_norm": 0.3954981565475464,
      "learning_rate": 0.0001938404,
      "loss": 0.4731,
      "step": 15400
    },
    {
      "epoch": 0.031,
      "grad_norm": 0.42979204654693604,
      "learning_rate": 0.00019380040000000002,
      "loss": 0.518,
      "step": 15500
    },
    {
      "epoch": 0.0312,
      "grad_norm": 0.5260553956031799,
      "learning_rate": 0.00019376040000000002,
      "loss": 0.4934,
      "step": 15600
    },
    {
      "epoch": 0.0314,
      "grad_norm": 0.6792057156562805,
      "learning_rate": 0.0001937204,
      "loss": 0.478,
      "step": 15700
    },
    {
      "epoch": 0.0316,
      "grad_norm": 0.3626260459423065,
      "learning_rate": 0.0001936804,
      "loss": 0.4859,
      "step": 15800
    },
    {
      "epoch": 0.0318,
      "grad_norm": 0.6549792885780334,
      "learning_rate": 0.00019364040000000002,
      "loss": 0.4843,
      "step": 15900
    },
    {
      "epoch": 0.032,
      "grad_norm": 0.40819457173347473,
      "learning_rate": 0.00019360040000000002,
      "loss": 0.4876,
      "step": 16000
    },
    {
      "epoch": 0.0322,
      "grad_norm": 0.46695244312286377,
      "learning_rate": 0.0001935604,
      "loss": 0.493,
      "step": 16100
    },
    {
      "epoch": 0.0324,
      "grad_norm": 0.29340672492980957,
      "learning_rate": 0.0001935204,
      "loss": 0.4822,
      "step": 16200
    },
    {
      "epoch": 0.0326,
      "grad_norm": 0.4895103871822357,
      "learning_rate": 0.0001934804,
      "loss": 0.4775,
      "step": 16300
    },
    {
      "epoch": 0.0328,
      "grad_norm": 0.38141900300979614,
      "learning_rate": 0.00019344040000000002,
      "loss": 0.5175,
      "step": 16400
    },
    {
      "epoch": 0.033,
      "grad_norm": 0.2875644862651825,
      "learning_rate": 0.0001934004,
      "loss": 0.5004,
      "step": 16500
    },
    {
      "epoch": 0.0332,
      "grad_norm": 1.0335603952407837,
      "learning_rate": 0.0001933604,
      "loss": 0.4917,
      "step": 16600
    },
    {
      "epoch": 0.0334,
      "grad_norm": 0.3046559691429138,
      "learning_rate": 0.0001933204,
      "loss": 0.5966,
      "step": 16700
    },
    {
      "epoch": 0.0336,
      "grad_norm": 0.3726640045642853,
      "learning_rate": 0.0001932804,
      "loss": 0.4882,
      "step": 16800
    },
    {
      "epoch": 0.0338,
      "grad_norm": 0.3167787194252014,
      "learning_rate": 0.00019324040000000001,
      "loss": 0.4979,
      "step": 16900
    },
    {
      "epoch": 0.034,
      "grad_norm": 0.23184147477149963,
      "learning_rate": 0.0001932004,
      "loss": 0.4873,
      "step": 17000
    },
    {
      "epoch": 0.0342,
      "grad_norm": 0.27494415640830994,
      "learning_rate": 0.00019316040000000003,
      "loss": 0.4735,
      "step": 17100
    },
    {
      "epoch": 0.0344,
      "grad_norm": 2.1850268840789795,
      "learning_rate": 0.0001931204,
      "loss": 0.4706,
      "step": 17200
    },
    {
      "epoch": 0.0346,
      "grad_norm": 0.6089339852333069,
      "learning_rate": 0.00019308040000000002,
      "loss": 0.4736,
      "step": 17300
    },
    {
      "epoch": 0.0348,
      "grad_norm": 0.33707159757614136,
      "learning_rate": 0.0001930404,
      "loss": 0.4639,
      "step": 17400
    },
    {
      "epoch": 0.035,
      "grad_norm": 0.3079136610031128,
      "learning_rate": 0.0001930004,
      "loss": 0.4576,
      "step": 17500
    },
    {
      "epoch": 0.0352,
      "grad_norm": 0.9146600961685181,
      "learning_rate": 0.00019296040000000002,
      "loss": 0.4713,
      "step": 17600
    },
    {
      "epoch": 0.0354,
      "grad_norm": 0.3352242708206177,
      "learning_rate": 0.0001929204,
      "loss": 0.4628,
      "step": 17700
    },
    {
      "epoch": 0.0356,
      "grad_norm": 0.3531840443611145,
      "learning_rate": 0.0001928804,
      "loss": 0.4814,
      "step": 17800
    },
    {
      "epoch": 0.0358,
      "grad_norm": 0.3614370822906494,
      "learning_rate": 0.0001928404,
      "loss": 0.4786,
      "step": 17900
    },
    {
      "epoch": 0.036,
      "grad_norm": 0.3174838721752167,
      "learning_rate": 0.00019280040000000003,
      "loss": 0.4618,
      "step": 18000
    },
    {
      "epoch": 0.0362,
      "grad_norm": 0.5685095191001892,
      "learning_rate": 0.00019276040000000002,
      "loss": 0.4678,
      "step": 18100
    },
    {
      "epoch": 0.0364,
      "grad_norm": 0.35674697160720825,
      "learning_rate": 0.0001927204,
      "loss": 0.4709,
      "step": 18200
    },
    {
      "epoch": 0.0366,
      "grad_norm": 0.3839879631996155,
      "learning_rate": 0.0001926804,
      "loss": 0.4709,
      "step": 18300
    },
    {
      "epoch": 0.0368,
      "grad_norm": 0.401214599609375,
      "learning_rate": 0.0001926404,
      "loss": 0.4716,
      "step": 18400
    },
    {
      "epoch": 0.037,
      "grad_norm": 0.4122023582458496,
      "learning_rate": 0.00019260040000000002,
      "loss": 0.481,
      "step": 18500
    },
    {
      "epoch": 0.0372,
      "grad_norm": 0.42760711908340454,
      "learning_rate": 0.00019256040000000001,
      "loss": 0.4534,
      "step": 18600
    },
    {
      "epoch": 0.0374,
      "grad_norm": 4.317193508148193,
      "learning_rate": 0.0001925204,
      "loss": 0.4838,
      "step": 18700
    },
    {
      "epoch": 0.0376,
      "grad_norm": 0.3325408697128296,
      "learning_rate": 0.0001924804,
      "loss": 0.4756,
      "step": 18800
    },
    {
      "epoch": 0.0378,
      "grad_norm": 0.4391893744468689,
      "learning_rate": 0.00019244040000000002,
      "loss": 0.4625,
      "step": 18900
    },
    {
      "epoch": 0.038,
      "grad_norm": 0.3271416127681732,
      "learning_rate": 0.00019240040000000002,
      "loss": 0.452,
      "step": 19000
    },
    {
      "epoch": 0.0382,
      "grad_norm": 0.2682945132255554,
      "learning_rate": 0.0001923604,
      "loss": 0.4677,
      "step": 19100
    },
    {
      "epoch": 0.0384,
      "grad_norm": 0.2958075702190399,
      "learning_rate": 0.0001923204,
      "loss": 0.4715,
      "step": 19200
    },
    {
      "epoch": 0.0386,
      "grad_norm": 0.6007779240608215,
      "learning_rate": 0.0001922804,
      "loss": 0.4581,
      "step": 19300
    },
    {
      "epoch": 0.0388,
      "grad_norm": 0.40477871894836426,
      "learning_rate": 0.00019224040000000002,
      "loss": 0.4675,
      "step": 19400
    },
    {
      "epoch": 0.039,
      "grad_norm": 0.44704627990722656,
      "learning_rate": 0.0001922004,
      "loss": 0.4846,
      "step": 19500
    },
    {
      "epoch": 0.0392,
      "grad_norm": 0.4317639172077179,
      "learning_rate": 0.00019216040000000003,
      "loss": 0.4681,
      "step": 19600
    },
    {
      "epoch": 0.0394,
      "grad_norm": 0.274308443069458,
      "learning_rate": 0.0001921204,
      "loss": 0.4577,
      "step": 19700
    },
    {
      "epoch": 0.0396,
      "grad_norm": 0.3873860239982605,
      "learning_rate": 0.0001920804,
      "loss": 0.4505,
      "step": 19800
    },
    {
      "epoch": 0.0398,
      "grad_norm": 0.4344724714756012,
      "learning_rate": 0.0001920404,
      "loss": 0.4531,
      "step": 19900
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.3240549564361572,
      "learning_rate": 0.0001920004,
      "loss": 0.4616,
      "step": 20000
    },
    {
      "epoch": 0.0402,
      "grad_norm": 0.3235743045806885,
      "learning_rate": 0.00019196040000000003,
      "loss": 0.4435,
      "step": 20100
    },
    {
      "epoch": 0.0404,
      "grad_norm": 0.6763646602630615,
      "learning_rate": 0.0001919204,
      "loss": 0.4806,
      "step": 20200
    },
    {
      "epoch": 0.0406,
      "grad_norm": 0.3265356123447418,
      "learning_rate": 0.00019188040000000001,
      "loss": 0.4674,
      "step": 20300
    },
    {
      "epoch": 0.0408,
      "grad_norm": 0.41400811076164246,
      "learning_rate": 0.0001918404,
      "loss": 0.4687,
      "step": 20400
    },
    {
      "epoch": 0.041,
      "grad_norm": 1.1848185062408447,
      "learning_rate": 0.0001918004,
      "loss": 0.4602,
      "step": 20500
    },
    {
      "epoch": 0.0412,
      "grad_norm": 0.4341808259487152,
      "learning_rate": 0.00019176040000000002,
      "loss": 0.4559,
      "step": 20600
    },
    {
      "epoch": 0.0414,
      "grad_norm": 0.40686967968940735,
      "learning_rate": 0.0001917204,
      "loss": 0.4408,
      "step": 20700
    },
    {
      "epoch": 0.0416,
      "grad_norm": 0.28525295853614807,
      "learning_rate": 0.0001916804,
      "loss": 0.4841,
      "step": 20800
    },
    {
      "epoch": 0.0418,
      "grad_norm": 0.29934221506118774,
      "learning_rate": 0.0001916404,
      "loss": 0.4562,
      "step": 20900
    },
    {
      "epoch": 0.042,
      "grad_norm": 0.6304417848587036,
      "learning_rate": 0.00019160040000000002,
      "loss": 0.4609,
      "step": 21000
    },
    {
      "epoch": 0.0422,
      "grad_norm": 0.2794235348701477,
      "learning_rate": 0.00019156040000000002,
      "loss": 0.4578,
      "step": 21100
    },
    {
      "epoch": 0.0424,
      "grad_norm": 0.9360523223876953,
      "learning_rate": 0.0001915204,
      "loss": 0.4563,
      "step": 21200
    },
    {
      "epoch": 0.0426,
      "grad_norm": 0.3164953291416168,
      "learning_rate": 0.0001914804,
      "loss": 0.4563,
      "step": 21300
    },
    {
      "epoch": 0.0428,
      "grad_norm": 0.2829827070236206,
      "learning_rate": 0.0001914404,
      "loss": 0.4509,
      "step": 21400
    },
    {
      "epoch": 0.043,
      "grad_norm": 0.3804665803909302,
      "learning_rate": 0.00019140040000000002,
      "loss": 0.4523,
      "step": 21500
    },
    {
      "epoch": 0.0432,
      "grad_norm": 0.2858908176422119,
      "learning_rate": 0.0001913604,
      "loss": 0.4669,
      "step": 21600
    },
    {
      "epoch": 0.0434,
      "grad_norm": 0.29328569769859314,
      "learning_rate": 0.0001913204,
      "loss": 0.4571,
      "step": 21700
    },
    {
      "epoch": 0.0436,
      "grad_norm": 0.3529110252857208,
      "learning_rate": 0.0001912804,
      "loss": 0.4546,
      "step": 21800
    },
    {
      "epoch": 0.0438,
      "grad_norm": 0.2541522681713104,
      "learning_rate": 0.00019124040000000002,
      "loss": 0.4618,
      "step": 21900
    },
    {
      "epoch": 0.044,
      "grad_norm": 0.3136732280254364,
      "learning_rate": 0.0001912004,
      "loss": 0.4579,
      "step": 22000
    },
    {
      "epoch": 0.0442,
      "grad_norm": 0.35282132029533386,
      "learning_rate": 0.0001911604,
      "loss": 0.4801,
      "step": 22100
    },
    {
      "epoch": 0.0444,
      "grad_norm": 0.32009491324424744,
      "learning_rate": 0.0001911204,
      "loss": 0.4422,
      "step": 22200
    },
    {
      "epoch": 0.0446,
      "grad_norm": 0.25222790241241455,
      "learning_rate": 0.0001910804,
      "loss": 0.5024,
      "step": 22300
    },
    {
      "epoch": 0.0448,
      "grad_norm": 0.24487198889255524,
      "learning_rate": 0.00019104040000000001,
      "loss": 0.44,
      "step": 22400
    },
    {
      "epoch": 0.045,
      "grad_norm": 0.25609472393989563,
      "learning_rate": 0.0001910004,
      "loss": 0.4403,
      "step": 22500
    },
    {
      "epoch": 0.0452,
      "grad_norm": 0.24061603844165802,
      "learning_rate": 0.00019096040000000003,
      "loss": 0.434,
      "step": 22600
    },
    {
      "epoch": 0.0454,
      "grad_norm": 0.24821767210960388,
      "learning_rate": 0.0001909204,
      "loss": 0.4383,
      "step": 22700
    },
    {
      "epoch": 0.0456,
      "grad_norm": 0.3346046507358551,
      "learning_rate": 0.00019088040000000002,
      "loss": 0.4337,
      "step": 22800
    },
    {
      "epoch": 0.0458,
      "grad_norm": 0.4589807093143463,
      "learning_rate": 0.0001908404,
      "loss": 0.436,
      "step": 22900
    },
    {
      "epoch": 0.046,
      "grad_norm": 0.24371834099292755,
      "learning_rate": 0.0001908004,
      "loss": 0.4374,
      "step": 23000
    },
    {
      "epoch": 0.0462,
      "grad_norm": 0.30695632100105286,
      "learning_rate": 0.00019076040000000002,
      "loss": 0.4507,
      "step": 23100
    },
    {
      "epoch": 0.0464,
      "grad_norm": 0.3627791702747345,
      "learning_rate": 0.0001907204,
      "loss": 0.4853,
      "step": 23200
    },
    {
      "epoch": 0.0466,
      "grad_norm": 0.4218233525753021,
      "learning_rate": 0.0001906804,
      "loss": 0.4422,
      "step": 23300
    },
    {
      "epoch": 0.0468,
      "grad_norm": 0.3387954831123352,
      "learning_rate": 0.0001906404,
      "loss": 0.4443,
      "step": 23400
    },
    {
      "epoch": 0.047,
      "grad_norm": 0.3261500597000122,
      "learning_rate": 0.0001906004,
      "loss": 0.4438,
      "step": 23500
    },
    {
      "epoch": 0.0472,
      "grad_norm": 0.27388980984687805,
      "learning_rate": 0.00019056040000000002,
      "loss": 0.4383,
      "step": 23600
    },
    {
      "epoch": 0.0474,
      "grad_norm": 0.2767557203769684,
      "learning_rate": 0.00019052039999999999,
      "loss": 0.444,
      "step": 23700
    },
    {
      "epoch": 0.0476,
      "grad_norm": 0.37184062600135803,
      "learning_rate": 0.0001904804,
      "loss": 0.443,
      "step": 23800
    },
    {
      "epoch": 0.0478,
      "grad_norm": 0.33180898427963257,
      "learning_rate": 0.0001904404,
      "loss": 0.4601,
      "step": 23900
    },
    {
      "epoch": 0.048,
      "grad_norm": 0.3857232630252838,
      "learning_rate": 0.00019040040000000002,
      "loss": 0.4645,
      "step": 24000
    },
    {
      "epoch": 0.0482,
      "grad_norm": 0.4319915175437927,
      "learning_rate": 0.00019036040000000001,
      "loss": 0.4464,
      "step": 24100
    },
    {
      "epoch": 0.0484,
      "grad_norm": 1.3287200927734375,
      "learning_rate": 0.0001903204,
      "loss": 0.4419,
      "step": 24200
    },
    {
      "epoch": 0.0486,
      "grad_norm": 0.2736344337463379,
      "learning_rate": 0.0001902804,
      "loss": 0.4541,
      "step": 24300
    },
    {
      "epoch": 0.0488,
      "grad_norm": 0.2820674777030945,
      "learning_rate": 0.0001902404,
      "loss": 0.4284,
      "step": 24400
    },
    {
      "epoch": 0.049,
      "grad_norm": 0.4350157678127289,
      "learning_rate": 0.00019020040000000002,
      "loss": 0.4487,
      "step": 24500
    },
    {
      "epoch": 0.0492,
      "grad_norm": 0.29447850584983826,
      "learning_rate": 0.0001901604,
      "loss": 0.4321,
      "step": 24600
    },
    {
      "epoch": 0.0494,
      "grad_norm": 0.5137728452682495,
      "learning_rate": 0.0001901204,
      "loss": 0.4482,
      "step": 24700
    },
    {
      "epoch": 0.0496,
      "grad_norm": 0.33691874146461487,
      "learning_rate": 0.0001900804,
      "loss": 0.4471,
      "step": 24800
    },
    {
      "epoch": 0.0498,
      "grad_norm": 0.4733399748802185,
      "learning_rate": 0.00019004040000000002,
      "loss": 0.4448,
      "step": 24900
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.2643273174762726,
      "learning_rate": 0.0001900004,
      "loss": 0.4413,
      "step": 25000
    },
    {
      "epoch": 0.0502,
      "grad_norm": 0.33353281021118164,
      "learning_rate": 0.0001899604,
      "loss": 0.4519,
      "step": 25100
    },
    {
      "epoch": 0.0504,
      "grad_norm": 0.7628878355026245,
      "learning_rate": 0.0001899204,
      "loss": 0.4843,
      "step": 25200
    },
    {
      "epoch": 0.0506,
      "grad_norm": 0.33647382259368896,
      "learning_rate": 0.0001898804,
      "loss": 0.4402,
      "step": 25300
    },
    {
      "epoch": 0.0508,
      "grad_norm": 0.2926424741744995,
      "learning_rate": 0.0001898404,
      "loss": 0.4441,
      "step": 25400
    },
    {
      "epoch": 0.051,
      "grad_norm": 0.24321198463439941,
      "learning_rate": 0.0001898004,
      "loss": 0.4485,
      "step": 25500
    },
    {
      "epoch": 0.0512,
      "grad_norm": 0.2866199314594269,
      "learning_rate": 0.00018976040000000003,
      "loss": 0.4311,
      "step": 25600
    },
    {
      "epoch": 0.0514,
      "grad_norm": 0.27528929710388184,
      "learning_rate": 0.0001897204,
      "loss": 0.4478,
      "step": 25700
    },
    {
      "epoch": 0.0516,
      "grad_norm": 0.27291688323020935,
      "learning_rate": 0.00018968040000000001,
      "loss": 0.4504,
      "step": 25800
    },
    {
      "epoch": 0.0518,
      "grad_norm": 0.36198413372039795,
      "learning_rate": 0.0001896404,
      "loss": 0.4304,
      "step": 25900
    },
    {
      "epoch": 0.052,
      "grad_norm": 1.4908273220062256,
      "learning_rate": 0.0001896004,
      "loss": 0.4516,
      "step": 26000
    },
    {
      "epoch": 0.0522,
      "grad_norm": 0.36965224146842957,
      "learning_rate": 0.00018956040000000002,
      "loss": 0.4227,
      "step": 26100
    },
    {
      "epoch": 0.0524,
      "grad_norm": 0.3467487394809723,
      "learning_rate": 0.00018952040000000002,
      "loss": 0.4352,
      "step": 26200
    },
    {
      "epoch": 0.0526,
      "grad_norm": 0.5308062434196472,
      "learning_rate": 0.0001894804,
      "loss": 0.4636,
      "step": 26300
    },
    {
      "epoch": 0.0528,
      "grad_norm": 0.30632877349853516,
      "learning_rate": 0.0001894404,
      "loss": 0.4416,
      "step": 26400
    },
    {
      "epoch": 0.053,
      "grad_norm": 0.9236398339271545,
      "learning_rate": 0.00018940040000000002,
      "loss": 0.4302,
      "step": 26500
    },
    {
      "epoch": 0.0532,
      "grad_norm": 0.37798625230789185,
      "learning_rate": 0.00018936040000000002,
      "loss": 0.4479,
      "step": 26600
    },
    {
      "epoch": 0.0534,
      "grad_norm": 0.20882748067378998,
      "learning_rate": 0.0001893204,
      "loss": 0.4256,
      "step": 26700
    },
    {
      "epoch": 0.0536,
      "grad_norm": 0.3316884934902191,
      "learning_rate": 0.0001892804,
      "loss": 0.4281,
      "step": 26800
    },
    {
      "epoch": 0.0538,
      "grad_norm": 0.2426268756389618,
      "learning_rate": 0.0001892404,
      "loss": 0.4328,
      "step": 26900
    },
    {
      "epoch": 0.054,
      "grad_norm": 0.263194739818573,
      "learning_rate": 0.00018920040000000002,
      "loss": 0.4321,
      "step": 27000
    },
    {
      "epoch": 0.0542,
      "grad_norm": 0.2998006343841553,
      "learning_rate": 0.0001891604,
      "loss": 0.4322,
      "step": 27100
    },
    {
      "epoch": 0.0544,
      "grad_norm": 0.32721447944641113,
      "learning_rate": 0.00018912040000000003,
      "loss": 0.4483,
      "step": 27200
    },
    {
      "epoch": 0.0546,
      "grad_norm": 0.33025509119033813,
      "learning_rate": 0.0001890804,
      "loss": 0.4429,
      "step": 27300
    },
    {
      "epoch": 0.0548,
      "grad_norm": 0.2779247462749481,
      "learning_rate": 0.0001890404,
      "loss": 0.4243,
      "step": 27400
    },
    {
      "epoch": 0.055,
      "grad_norm": 0.4281350374221802,
      "learning_rate": 0.00018900040000000001,
      "loss": 0.43,
      "step": 27500
    },
    {
      "epoch": 0.0552,
      "grad_norm": 0.3283655345439911,
      "learning_rate": 0.0001889604,
      "loss": 0.4358,
      "step": 27600
    },
    {
      "epoch": 0.0554,
      "grad_norm": 0.4739823341369629,
      "learning_rate": 0.00018892040000000003,
      "loss": 0.4729,
      "step": 27700
    },
    {
      "epoch": 0.0556,
      "grad_norm": 0.5889720916748047,
      "learning_rate": 0.0001888804,
      "loss": 0.4436,
      "step": 27800
    },
    {
      "epoch": 0.0558,
      "grad_norm": 0.6123796105384827,
      "learning_rate": 0.00018884040000000002,
      "loss": 0.4222,
      "step": 27900
    },
    {
      "epoch": 0.056,
      "grad_norm": 0.24797767400741577,
      "learning_rate": 0.0001888004,
      "loss": 0.4527,
      "step": 28000
    },
    {
      "epoch": 0.0562,
      "grad_norm": 0.2619722783565521,
      "learning_rate": 0.0001887604,
      "loss": 0.4307,
      "step": 28100
    },
    {
      "epoch": 0.0564,
      "grad_norm": 0.309741348028183,
      "learning_rate": 0.00018872040000000002,
      "loss": 0.4343,
      "step": 28200
    },
    {
      "epoch": 0.0566,
      "grad_norm": 0.6912897825241089,
      "learning_rate": 0.0001886804,
      "loss": 0.4137,
      "step": 28300
    },
    {
      "epoch": 0.0568,
      "grad_norm": 0.40055686235427856,
      "learning_rate": 0.0001886404,
      "loss": 0.4395,
      "step": 28400
    },
    {
      "epoch": 0.057,
      "grad_norm": 0.19984565675258636,
      "learning_rate": 0.0001886004,
      "loss": 0.4418,
      "step": 28500
    },
    {
      "epoch": 0.0572,
      "grad_norm": 0.4430752694606781,
      "learning_rate": 0.00018856040000000003,
      "loss": 0.4269,
      "step": 28600
    },
    {
      "epoch": 0.0574,
      "grad_norm": 0.37985754013061523,
      "learning_rate": 0.00018852040000000002,
      "loss": 0.4349,
      "step": 28700
    },
    {
      "epoch": 0.0576,
      "grad_norm": 0.29130980372428894,
      "learning_rate": 0.0001884804,
      "loss": 0.4394,
      "step": 28800
    },
    {
      "epoch": 0.0578,
      "grad_norm": 0.5696423053741455,
      "learning_rate": 0.0001884404,
      "loss": 0.442,
      "step": 28900
    },
    {
      "epoch": 0.058,
      "grad_norm": 0.25018632411956787,
      "learning_rate": 0.0001884004,
      "loss": 0.4433,
      "step": 29000
    },
    {
      "epoch": 0.0582,
      "grad_norm": 0.2951875627040863,
      "learning_rate": 0.00018836040000000002,
      "loss": 0.4188,
      "step": 29100
    },
    {
      "epoch": 0.0584,
      "grad_norm": 0.23352210223674774,
      "learning_rate": 0.00018832040000000001,
      "loss": 0.4222,
      "step": 29200
    },
    {
      "epoch": 0.0586,
      "grad_norm": 0.3203904330730438,
      "learning_rate": 0.0001882804,
      "loss": 0.4261,
      "step": 29300
    },
    {
      "epoch": 0.0588,
      "grad_norm": 0.34108293056488037,
      "learning_rate": 0.0001882404,
      "loss": 0.4341,
      "step": 29400
    },
    {
      "epoch": 0.059,
      "grad_norm": 0.45176732540130615,
      "learning_rate": 0.00018820040000000002,
      "loss": 0.4667,
      "step": 29500
    },
    {
      "epoch": 0.0592,
      "grad_norm": 0.4928261637687683,
      "learning_rate": 0.00018816040000000002,
      "loss": 0.4546,
      "step": 29600
    },
    {
      "epoch": 0.0594,
      "grad_norm": 0.31960615515708923,
      "learning_rate": 0.0001881204,
      "loss": 0.4346,
      "step": 29700
    },
    {
      "epoch": 0.0596,
      "grad_norm": 0.2550900876522064,
      "learning_rate": 0.0001880804,
      "loss": 0.4181,
      "step": 29800
    },
    {
      "epoch": 0.0598,
      "grad_norm": 0.21990257501602173,
      "learning_rate": 0.0001880404,
      "loss": 0.4103,
      "step": 29900
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.2592441439628601,
      "learning_rate": 0.00018800040000000002,
      "loss": 0.4092,
      "step": 30000
    },
    {
      "epoch": 0.0602,
      "grad_norm": 0.2601667046546936,
      "learning_rate": 0.0001879604,
      "loss": 0.3927,
      "step": 30100
    },
    {
      "epoch": 0.0604,
      "grad_norm": 0.24636943638324738,
      "learning_rate": 0.00018792040000000003,
      "loss": 0.3929,
      "step": 30200
    },
    {
      "epoch": 0.0606,
      "grad_norm": 0.3238808810710907,
      "learning_rate": 0.0001878804,
      "loss": 0.3894,
      "step": 30300
    },
    {
      "epoch": 0.0608,
      "grad_norm": 0.2241670787334442,
      "learning_rate": 0.00018784040000000002,
      "loss": 0.3862,
      "step": 30400
    },
    {
      "epoch": 0.061,
      "grad_norm": 0.26162776350975037,
      "learning_rate": 0.0001878004,
      "loss": 0.3836,
      "step": 30500
    },
    {
      "epoch": 0.0612,
      "grad_norm": 0.21369214355945587,
      "learning_rate": 0.0001877604,
      "loss": 0.3829,
      "step": 30600
    },
    {
      "epoch": 0.0614,
      "grad_norm": 0.311394602060318,
      "learning_rate": 0.00018772040000000003,
      "loss": 0.384,
      "step": 30700
    },
    {
      "epoch": 0.0616,
      "grad_norm": 0.2316521257162094,
      "learning_rate": 0.0001876804,
      "loss": 0.3934,
      "step": 30800
    },
    {
      "epoch": 0.0618,
      "grad_norm": 0.4280303120613098,
      "learning_rate": 0.00018764040000000001,
      "loss": 0.3983,
      "step": 30900
    },
    {
      "epoch": 0.062,
      "grad_norm": 0.26460182666778564,
      "learning_rate": 0.0001876004,
      "loss": 0.3815,
      "step": 31000
    },
    {
      "epoch": 0.0622,
      "grad_norm": 0.2175382673740387,
      "learning_rate": 0.0001875604,
      "loss": 0.393,
      "step": 31100
    },
    {
      "epoch": 0.0624,
      "grad_norm": 0.2641454041004181,
      "learning_rate": 0.00018752040000000002,
      "loss": 0.3982,
      "step": 31200
    },
    {
      "epoch": 0.0626,
      "grad_norm": 0.28961893916130066,
      "learning_rate": 0.0001874804,
      "loss": 0.393,
      "step": 31300
    },
    {
      "epoch": 0.0628,
      "grad_norm": 0.2512940764427185,
      "learning_rate": 0.0001874404,
      "loss": 0.3896,
      "step": 31400
    },
    {
      "epoch": 0.063,
      "grad_norm": 0.26094546914100647,
      "learning_rate": 0.0001874004,
      "loss": 0.3861,
      "step": 31500
    },
    {
      "epoch": 0.0632,
      "grad_norm": 1.4088988304138184,
      "learning_rate": 0.00018736040000000002,
      "loss": 0.3894,
      "step": 31600
    },
    {
      "epoch": 0.0634,
      "grad_norm": 0.21915282309055328,
      "learning_rate": 0.00018732040000000002,
      "loss": 0.3895,
      "step": 31700
    },
    {
      "epoch": 0.0636,
      "grad_norm": 0.282810240983963,
      "learning_rate": 0.0001872804,
      "loss": 0.3814,
      "step": 31800
    },
    {
      "epoch": 0.0638,
      "grad_norm": 0.24024060368537903,
      "learning_rate": 0.0001872404,
      "loss": 0.3802,
      "step": 31900
    },
    {
      "epoch": 0.064,
      "grad_norm": 0.254407674074173,
      "learning_rate": 0.0001872004,
      "loss": 0.3916,
      "step": 32000
    },
    {
      "epoch": 0.0642,
      "grad_norm": 0.22265967726707458,
      "learning_rate": 0.00018716040000000002,
      "loss": 0.3941,
      "step": 32100
    },
    {
      "epoch": 0.0644,
      "grad_norm": 0.2248506098985672,
      "learning_rate": 0.0001871204,
      "loss": 0.3878,
      "step": 32200
    },
    {
      "epoch": 0.0646,
      "grad_norm": 0.30514708161354065,
      "learning_rate": 0.0001870804,
      "loss": 0.3861,
      "step": 32300
    },
    {
      "epoch": 0.0648,
      "grad_norm": 0.24802802503108978,
      "learning_rate": 0.0001870404,
      "loss": 0.3802,
      "step": 32400
    },
    {
      "epoch": 0.065,
      "grad_norm": 0.25993114709854126,
      "learning_rate": 0.00018700040000000002,
      "loss": 0.3832,
      "step": 32500
    },
    {
      "epoch": 0.0652,
      "grad_norm": 0.2758074104785919,
      "learning_rate": 0.0001869604,
      "loss": 0.3898,
      "step": 32600
    },
    {
      "epoch": 0.0654,
      "grad_norm": 0.2821817696094513,
      "learning_rate": 0.0001869204,
      "loss": 0.3901,
      "step": 32700
    },
    {
      "epoch": 0.0656,
      "grad_norm": 0.3593108057975769,
      "learning_rate": 0.0001868804,
      "loss": 0.3835,
      "step": 32800
    },
    {
      "epoch": 0.0658,
      "grad_norm": 0.21605613827705383,
      "learning_rate": 0.0001868404,
      "loss": 0.3862,
      "step": 32900
    },
    {
      "epoch": 0.066,
      "grad_norm": 0.34763646125793457,
      "learning_rate": 0.00018680040000000001,
      "loss": 0.3861,
      "step": 33000
    },
    {
      "epoch": 0.0662,
      "grad_norm": 0.2341729700565338,
      "learning_rate": 0.0001867604,
      "loss": 0.3876,
      "step": 33100
    },
    {
      "epoch": 0.0664,
      "grad_norm": 0.6066027879714966,
      "learning_rate": 0.00018672040000000003,
      "loss": 0.3904,
      "step": 33200
    },
    {
      "epoch": 0.0666,
      "grad_norm": 0.2845059037208557,
      "learning_rate": 0.0001866804,
      "loss": 0.387,
      "step": 33300
    },
    {
      "epoch": 0.0668,
      "grad_norm": 0.23616880178451538,
      "learning_rate": 0.00018664040000000002,
      "loss": 0.3862,
      "step": 33400
    },
    {
      "epoch": 0.067,
      "grad_norm": 0.21980926394462585,
      "learning_rate": 0.0001866004,
      "loss": 0.3888,
      "step": 33500
    },
    {
      "epoch": 0.0672,
      "grad_norm": 0.2976222336292267,
      "learning_rate": 0.0001865604,
      "loss": 0.3873,
      "step": 33600
    },
    {
      "epoch": 0.0674,
      "grad_norm": 0.23099921643733978,
      "learning_rate": 0.00018652040000000002,
      "loss": 0.3816,
      "step": 33700
    },
    {
      "epoch": 0.0676,
      "grad_norm": 0.2811383306980133,
      "learning_rate": 0.0001864804,
      "loss": 0.3849,
      "step": 33800
    },
    {
      "epoch": 0.0678,
      "grad_norm": 0.2094060480594635,
      "learning_rate": 0.0001864404,
      "loss": 0.3761,
      "step": 33900
    },
    {
      "epoch": 0.068,
      "grad_norm": 0.2863057553768158,
      "learning_rate": 0.0001864004,
      "loss": 0.384,
      "step": 34000
    },
    {
      "epoch": 0.0682,
      "grad_norm": 0.23131683468818665,
      "learning_rate": 0.00018636040000000003,
      "loss": 0.3898,
      "step": 34100
    },
    {
      "epoch": 0.0684,
      "grad_norm": 0.2947816848754883,
      "learning_rate": 0.00018632040000000002,
      "loss": 0.3853,
      "step": 34200
    },
    {
      "epoch": 0.0686,
      "grad_norm": 0.2566851079463959,
      "learning_rate": 0.00018628039999999999,
      "loss": 0.3743,
      "step": 34300
    },
    {
      "epoch": 0.0688,
      "grad_norm": 0.23540231585502625,
      "learning_rate": 0.0001862404,
      "loss": 0.3777,
      "step": 34400
    },
    {
      "epoch": 0.069,
      "grad_norm": 0.3022463917732239,
      "learning_rate": 0.0001862004,
      "loss": 0.3865,
      "step": 34500
    },
    {
      "epoch": 0.0692,
      "grad_norm": 0.2179107815027237,
      "learning_rate": 0.00018616040000000002,
      "loss": 0.3802,
      "step": 34600
    },
    {
      "epoch": 0.0694,
      "grad_norm": 0.3824068009853363,
      "learning_rate": 0.00018612040000000001,
      "loss": 0.385,
      "step": 34700
    },
    {
      "epoch": 0.0696,
      "grad_norm": 0.2288985550403595,
      "learning_rate": 0.0001860804,
      "loss": 0.3767,
      "step": 34800
    },
    {
      "epoch": 0.0698,
      "grad_norm": 0.228486567735672,
      "learning_rate": 0.0001860404,
      "loss": 0.3827,
      "step": 34900
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.2173689603805542,
      "learning_rate": 0.0001860004,
      "loss": 0.3865,
      "step": 35000
    },
    {
      "epoch": 0.0702,
      "grad_norm": 0.27074554562568665,
      "learning_rate": 0.00018596040000000002,
      "loss": 0.3756,
      "step": 35100
    },
    {
      "epoch": 0.0704,
      "grad_norm": 0.3166576623916626,
      "learning_rate": 0.0001859204,
      "loss": 0.3889,
      "step": 35200
    },
    {
      "epoch": 0.0706,
      "grad_norm": 0.3322249948978424,
      "learning_rate": 0.0001858804,
      "loss": 0.3793,
      "step": 35300
    },
    {
      "epoch": 0.0708,
      "grad_norm": 0.2725512981414795,
      "learning_rate": 0.0001858404,
      "loss": 0.3828,
      "step": 35400
    },
    {
      "epoch": 0.071,
      "grad_norm": 0.20953910052776337,
      "learning_rate": 0.00018580040000000002,
      "loss": 0.3771,
      "step": 35500
    },
    {
      "epoch": 0.0712,
      "grad_norm": 0.2445133924484253,
      "learning_rate": 0.0001857604,
      "loss": 0.3737,
      "step": 35600
    },
    {
      "epoch": 0.0714,
      "grad_norm": 0.5710493922233582,
      "learning_rate": 0.0001857204,
      "loss": 0.3712,
      "step": 35700
    },
    {
      "epoch": 0.0716,
      "grad_norm": 0.3139127790927887,
      "learning_rate": 0.0001856804,
      "loss": 0.3787,
      "step": 35800
    },
    {
      "epoch": 0.0718,
      "grad_norm": 0.23770715296268463,
      "learning_rate": 0.0001856404,
      "loss": 0.3814,
      "step": 35900
    },
    {
      "epoch": 0.072,
      "grad_norm": 0.3806588351726532,
      "learning_rate": 0.0001856004,
      "loss": 0.373,
      "step": 36000
    },
    {
      "epoch": 0.0722,
      "grad_norm": 0.25435832142829895,
      "learning_rate": 0.0001855604,
      "loss": 0.3789,
      "step": 36100
    },
    {
      "epoch": 0.0724,
      "grad_norm": 0.26886075735092163,
      "learning_rate": 0.00018552040000000003,
      "loss": 0.3837,
      "step": 36200
    },
    {
      "epoch": 0.0726,
      "grad_norm": 0.22618788480758667,
      "learning_rate": 0.0001854804,
      "loss": 0.3775,
      "step": 36300
    },
    {
      "epoch": 0.0728,
      "grad_norm": 0.1857277899980545,
      "learning_rate": 0.00018544040000000001,
      "loss": 0.3759,
      "step": 36400
    },
    {
      "epoch": 0.073,
      "grad_norm": 0.34357950091362,
      "learning_rate": 0.0001854004,
      "loss": 0.3698,
      "step": 36500
    },
    {
      "epoch": 0.0732,
      "grad_norm": 0.26498422026634216,
      "learning_rate": 0.0001853604,
      "loss": 0.3702,
      "step": 36600
    },
    {
      "epoch": 0.0734,
      "grad_norm": 0.21762557327747345,
      "learning_rate": 0.00018532040000000002,
      "loss": 0.3756,
      "step": 36700
    },
    {
      "epoch": 0.0736,
      "grad_norm": 0.2890067398548126,
      "learning_rate": 0.0001852804,
      "loss": 0.3707,
      "step": 36800
    },
    {
      "epoch": 0.0738,
      "grad_norm": 0.24609336256980896,
      "learning_rate": 0.0001852404,
      "loss": 0.3766,
      "step": 36900
    },
    {
      "epoch": 0.074,
      "grad_norm": 0.262658953666687,
      "learning_rate": 0.0001852004,
      "loss": 0.3736,
      "step": 37000
    },
    {
      "epoch": 0.0742,
      "grad_norm": 0.20918047428131104,
      "learning_rate": 0.00018516040000000002,
      "loss": 0.3756,
      "step": 37100
    },
    {
      "epoch": 0.0744,
      "grad_norm": 0.324136883020401,
      "learning_rate": 0.00018512040000000002,
      "loss": 0.3737,
      "step": 37200
    },
    {
      "epoch": 0.0746,
      "grad_norm": 0.3206271827220917,
      "learning_rate": 0.0001850804,
      "loss": 0.3725,
      "step": 37300
    },
    {
      "epoch": 0.0748,
      "grad_norm": 0.24515919387340546,
      "learning_rate": 0.0001850404,
      "loss": 0.369,
      "step": 37400
    },
    {
      "epoch": 0.075,
      "grad_norm": 0.441785991191864,
      "learning_rate": 0.0001850004,
      "loss": 0.3843,
      "step": 37500
    },
    {
      "epoch": 0.0752,
      "grad_norm": 0.20712384581565857,
      "learning_rate": 0.00018496040000000002,
      "loss": 0.3758,
      "step": 37600
    },
    {
      "epoch": 0.0754,
      "grad_norm": 0.2728515565395355,
      "learning_rate": 0.0001849204,
      "loss": 0.3674,
      "step": 37700
    },
    {
      "epoch": 0.0756,
      "grad_norm": 0.34958529472351074,
      "learning_rate": 0.0001848804,
      "loss": 0.3713,
      "step": 37800
    },
    {
      "epoch": 0.0758,
      "grad_norm": 0.2223168909549713,
      "learning_rate": 0.0001848404,
      "loss": 0.3724,
      "step": 37900
    },
    {
      "epoch": 0.076,
      "grad_norm": 0.22961725294589996,
      "learning_rate": 0.0001848004,
      "loss": 0.3768,
      "step": 38000
    },
    {
      "epoch": 0.0762,
      "grad_norm": 0.23851755261421204,
      "learning_rate": 0.00018476040000000001,
      "loss": 0.3715,
      "step": 38100
    },
    {
      "epoch": 0.0764,
      "grad_norm": 0.3652637004852295,
      "learning_rate": 0.0001847204,
      "loss": 0.3585,
      "step": 38200
    },
    {
      "epoch": 0.0766,
      "grad_norm": 0.2556198835372925,
      "learning_rate": 0.0001846804,
      "loss": 0.3682,
      "step": 38300
    },
    {
      "epoch": 0.0768,
      "grad_norm": 0.2592809498310089,
      "learning_rate": 0.0001846404,
      "loss": 0.3697,
      "step": 38400
    },
    {
      "epoch": 0.077,
      "grad_norm": 0.217251256108284,
      "learning_rate": 0.00018460040000000002,
      "loss": 0.3657,
      "step": 38500
    },
    {
      "epoch": 0.0772,
      "grad_norm": 0.22854512929916382,
      "learning_rate": 0.0001845604,
      "loss": 0.3669,
      "step": 38600
    },
    {
      "epoch": 0.0774,
      "grad_norm": 0.4855661690235138,
      "learning_rate": 0.0001845204,
      "loss": 0.3672,
      "step": 38700
    },
    {
      "epoch": 0.0776,
      "grad_norm": 0.26462578773498535,
      "learning_rate": 0.00018448040000000002,
      "loss": 0.3642,
      "step": 38800
    },
    {
      "epoch": 0.0778,
      "grad_norm": 0.24063749611377716,
      "learning_rate": 0.0001844404,
      "loss": 0.3682,
      "step": 38900
    },
    {
      "epoch": 0.078,
      "grad_norm": 0.21736368536949158,
      "learning_rate": 0.0001844004,
      "loss": 0.3813,
      "step": 39000
    },
    {
      "epoch": 0.0782,
      "grad_norm": 0.2315133512020111,
      "learning_rate": 0.0001843604,
      "loss": 0.3735,
      "step": 39100
    },
    {
      "epoch": 0.0784,
      "grad_norm": 0.23136012256145477,
      "learning_rate": 0.00018432040000000003,
      "loss": 0.3603,
      "step": 39200
    },
    {
      "epoch": 0.0786,
      "grad_norm": 0.2924899756908417,
      "learning_rate": 0.00018428040000000002,
      "loss": 0.3632,
      "step": 39300
    },
    {
      "epoch": 0.0788,
      "grad_norm": 0.307910293340683,
      "learning_rate": 0.0001842404,
      "loss": 0.3671,
      "step": 39400
    },
    {
      "epoch": 0.079,
      "grad_norm": 0.22759227454662323,
      "learning_rate": 0.0001842004,
      "loss": 0.3707,
      "step": 39500
    },
    {
      "epoch": 0.0792,
      "grad_norm": 0.21660542488098145,
      "learning_rate": 0.0001841604,
      "loss": 0.3728,
      "step": 39600
    },
    {
      "epoch": 0.0794,
      "grad_norm": 0.2659781575202942,
      "learning_rate": 0.00018412040000000002,
      "loss": 0.3729,
      "step": 39700
    },
    {
      "epoch": 0.0796,
      "grad_norm": 0.25910884141921997,
      "learning_rate": 0.00018408040000000001,
      "loss": 0.3681,
      "step": 39800
    },
    {
      "epoch": 0.0798,
      "grad_norm": 0.2730049788951874,
      "learning_rate": 0.0001840404,
      "loss": 0.3716,
      "step": 39900
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.23555901646614075,
      "learning_rate": 0.0001840004,
      "loss": 0.3683,
      "step": 40000
    },
    {
      "epoch": 0.0802,
      "grad_norm": 0.22227054834365845,
      "learning_rate": 0.00018396040000000002,
      "loss": 0.371,
      "step": 40100
    },
    {
      "epoch": 0.0804,
      "grad_norm": 0.21306991577148438,
      "learning_rate": 0.00018392040000000002,
      "loss": 0.3666,
      "step": 40200
    },
    {
      "epoch": 0.0806,
      "grad_norm": 0.37939247488975525,
      "learning_rate": 0.0001838804,
      "loss": 0.3782,
      "step": 40300
    },
    {
      "epoch": 0.0808,
      "grad_norm": 0.26984167098999023,
      "learning_rate": 0.0001838404,
      "loss": 0.3717,
      "step": 40400
    },
    {
      "epoch": 0.081,
      "grad_norm": 0.22309809923171997,
      "learning_rate": 0.0001838004,
      "loss": 0.3761,
      "step": 40500
    },
    {
      "epoch": 0.0812,
      "grad_norm": 0.22144988179206848,
      "learning_rate": 0.00018376040000000002,
      "loss": 0.3609,
      "step": 40600
    },
    {
      "epoch": 0.0814,
      "grad_norm": 0.23661202192306519,
      "learning_rate": 0.0001837204,
      "loss": 0.364,
      "step": 40700
    },
    {
      "epoch": 0.0816,
      "grad_norm": 0.23225858807563782,
      "learning_rate": 0.00018368040000000003,
      "loss": 0.3608,
      "step": 40800
    },
    {
      "epoch": 0.0818,
      "grad_norm": 0.21633361279964447,
      "learning_rate": 0.0001836404,
      "loss": 0.3677,
      "step": 40900
    },
    {
      "epoch": 0.082,
      "grad_norm": 0.20861107110977173,
      "learning_rate": 0.00018360040000000002,
      "loss": 0.3586,
      "step": 41000
    },
    {
      "epoch": 0.0822,
      "grad_norm": 0.22681254148483276,
      "learning_rate": 0.0001835604,
      "loss": 0.3503,
      "step": 41100
    },
    {
      "epoch": 0.0824,
      "grad_norm": 0.29530957341194153,
      "learning_rate": 0.0001835204,
      "loss": 0.3546,
      "step": 41200
    },
    {
      "epoch": 0.0826,
      "grad_norm": 0.22965680062770844,
      "learning_rate": 0.00018348040000000003,
      "loss": 0.3581,
      "step": 41300
    },
    {
      "epoch": 0.0828,
      "grad_norm": 0.24958209693431854,
      "learning_rate": 0.0001834404,
      "loss": 0.3601,
      "step": 41400
    },
    {
      "epoch": 0.083,
      "grad_norm": 0.34926220774650574,
      "learning_rate": 0.0001834004,
      "loss": 0.3546,
      "step": 41500
    },
    {
      "epoch": 0.0832,
      "grad_norm": 0.23350180685520172,
      "learning_rate": 0.0001833604,
      "loss": 0.3591,
      "step": 41600
    },
    {
      "epoch": 0.0834,
      "grad_norm": 0.28490149974823,
      "learning_rate": 0.00018332040000000003,
      "loss": 0.3595,
      "step": 41700
    },
    {
      "epoch": 0.0836,
      "grad_norm": 0.37000980973243713,
      "learning_rate": 0.00018328040000000002,
      "loss": 0.357,
      "step": 41800
    },
    {
      "epoch": 0.0838,
      "grad_norm": 0.25736093521118164,
      "learning_rate": 0.0001832404,
      "loss": 0.3584,
      "step": 41900
    },
    {
      "epoch": 0.084,
      "grad_norm": 0.3347998559474945,
      "learning_rate": 0.0001832004,
      "loss": 0.3668,
      "step": 42000
    },
    {
      "epoch": 0.0842,
      "grad_norm": 0.259397029876709,
      "learning_rate": 0.0001831604,
      "loss": 0.358,
      "step": 42100
    },
    {
      "epoch": 0.0844,
      "grad_norm": 0.26411810517311096,
      "learning_rate": 0.00018312040000000002,
      "loss": 0.3678,
      "step": 42200
    },
    {
      "epoch": 0.0846,
      "grad_norm": 0.24818368256092072,
      "learning_rate": 0.00018308040000000002,
      "loss": 0.3563,
      "step": 42300
    },
    {
      "epoch": 0.0848,
      "grad_norm": 0.27950188517570496,
      "learning_rate": 0.0001830404,
      "loss": 0.3561,
      "step": 42400
    },
    {
      "epoch": 0.085,
      "grad_norm": 0.21346434950828552,
      "learning_rate": 0.0001830004,
      "loss": 0.3608,
      "step": 42500
    },
    {
      "epoch": 0.0852,
      "grad_norm": 0.20952267944812775,
      "learning_rate": 0.0001829604,
      "loss": 0.3565,
      "step": 42600
    },
    {
      "epoch": 0.0854,
      "grad_norm": 0.3930506706237793,
      "learning_rate": 0.00018292040000000002,
      "loss": 0.3591,
      "step": 42700
    },
    {
      "epoch": 0.0856,
      "grad_norm": 0.19581203162670135,
      "learning_rate": 0.0001828804,
      "loss": 0.3575,
      "step": 42800
    },
    {
      "epoch": 0.0858,
      "grad_norm": 0.20917409658432007,
      "learning_rate": 0.0001828404,
      "loss": 0.3609,
      "step": 42900
    },
    {
      "epoch": 0.086,
      "grad_norm": 0.47267964482307434,
      "learning_rate": 0.0001828004,
      "loss": 0.3629,
      "step": 43000
    },
    {
      "epoch": 0.0862,
      "grad_norm": 0.3048167824745178,
      "learning_rate": 0.00018276040000000002,
      "loss": 0.3535,
      "step": 43100
    },
    {
      "epoch": 0.0864,
      "grad_norm": 0.23811852931976318,
      "learning_rate": 0.0001827204,
      "loss": 0.362,
      "step": 43200
    },
    {
      "epoch": 0.0866,
      "grad_norm": 0.21190153062343597,
      "learning_rate": 0.0001826804,
      "loss": 0.3856,
      "step": 43300
    },
    {
      "epoch": 0.0868,
      "grad_norm": 0.2275034487247467,
      "learning_rate": 0.0001826404,
      "loss": 0.3586,
      "step": 43400
    },
    {
      "epoch": 0.087,
      "grad_norm": 0.19929227232933044,
      "learning_rate": 0.0001826004,
      "loss": 0.3518,
      "step": 43500
    },
    {
      "epoch": 0.0872,
      "grad_norm": 0.25546959042549133,
      "learning_rate": 0.00018256040000000001,
      "loss": 0.3543,
      "step": 43600
    },
    {
      "epoch": 0.0874,
      "grad_norm": 0.20379209518432617,
      "learning_rate": 0.0001825204,
      "loss": 0.3599,
      "step": 43700
    },
    {
      "epoch": 0.0876,
      "grad_norm": 0.27397245168685913,
      "learning_rate": 0.00018248040000000003,
      "loss": 0.3573,
      "step": 43800
    },
    {
      "epoch": 0.0878,
      "grad_norm": 0.23391596972942352,
      "learning_rate": 0.0001824404,
      "loss": 0.3538,
      "step": 43900
    },
    {
      "epoch": 0.088,
      "grad_norm": 0.23923568427562714,
      "learning_rate": 0.00018240040000000002,
      "loss": 0.3529,
      "step": 44000
    },
    {
      "epoch": 0.0882,
      "grad_norm": 0.2521684467792511,
      "learning_rate": 0.0001823604,
      "loss": 0.3691,
      "step": 44100
    },
    {
      "epoch": 0.0884,
      "grad_norm": 0.2728547155857086,
      "learning_rate": 0.0001823204,
      "loss": 0.3596,
      "step": 44200
    },
    {
      "epoch": 0.0886,
      "grad_norm": 0.28586891293525696,
      "learning_rate": 0.00018228040000000002,
      "loss": 0.3634,
      "step": 44300
    },
    {
      "epoch": 0.0888,
      "grad_norm": 0.2821336090564728,
      "learning_rate": 0.0001822404,
      "loss": 0.3554,
      "step": 44400
    },
    {
      "epoch": 0.089,
      "grad_norm": 0.38594385981559753,
      "learning_rate": 0.0001822004,
      "loss": 0.3559,
      "step": 44500
    },
    {
      "epoch": 0.0892,
      "grad_norm": 0.235568568110466,
      "learning_rate": 0.0001821604,
      "loss": 0.3514,
      "step": 44600
    },
    {
      "epoch": 0.0894,
      "grad_norm": 0.2375953644514084,
      "learning_rate": 0.00018212040000000003,
      "loss": 0.3576,
      "step": 44700
    },
    {
      "epoch": 0.0896,
      "grad_norm": 0.20530065894126892,
      "learning_rate": 0.00018208040000000002,
      "loss": 0.3574,
      "step": 44800
    },
    {
      "epoch": 0.0898,
      "grad_norm": 0.1840105950832367,
      "learning_rate": 0.0001820404,
      "loss": 0.351,
      "step": 44900
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.25125375390052795,
      "learning_rate": 0.0001820004,
      "loss": 0.3533,
      "step": 45000
    },
    {
      "epoch": 0.0902,
      "grad_norm": 0.2656700313091278,
      "learning_rate": 0.0001819604,
      "loss": 0.3604,
      "step": 45100
    },
    {
      "epoch": 0.0904,
      "grad_norm": 0.27208247780799866,
      "learning_rate": 0.00018192040000000002,
      "loss": 0.3552,
      "step": 45200
    },
    {
      "epoch": 0.0906,
      "grad_norm": 0.2230006903409958,
      "learning_rate": 0.00018188040000000001,
      "loss": 0.3515,
      "step": 45300
    },
    {
      "epoch": 0.0908,
      "grad_norm": 0.28219157457351685,
      "learning_rate": 0.0001818404,
      "loss": 0.357,
      "step": 45400
    },
    {
      "epoch": 0.091,
      "grad_norm": 0.21550622582435608,
      "learning_rate": 0.0001818004,
      "loss": 0.3512,
      "step": 45500
    },
    {
      "epoch": 0.0912,
      "grad_norm": 0.23631860315799713,
      "learning_rate": 0.0001817604,
      "loss": 0.3471,
      "step": 45600
    },
    {
      "epoch": 0.0914,
      "grad_norm": 0.1854236125946045,
      "learning_rate": 0.00018172040000000002,
      "loss": 0.3509,
      "step": 45700
    },
    {
      "epoch": 0.0916,
      "grad_norm": 0.2366824895143509,
      "learning_rate": 0.0001816804,
      "loss": 0.3583,
      "step": 45800
    },
    {
      "epoch": 0.0918,
      "grad_norm": 0.37892627716064453,
      "learning_rate": 0.0001816404,
      "loss": 0.3479,
      "step": 45900
    },
    {
      "epoch": 0.092,
      "grad_norm": 0.2608516216278076,
      "learning_rate": 0.0001816004,
      "loss": 0.3478,
      "step": 46000
    },
    {
      "epoch": 0.0922,
      "grad_norm": 0.23194310069084167,
      "learning_rate": 0.00018156040000000002,
      "loss": 0.3654,
      "step": 46100
    },
    {
      "epoch": 0.0924,
      "grad_norm": 0.2503475248813629,
      "learning_rate": 0.0001815204,
      "loss": 0.3508,
      "step": 46200
    },
    {
      "epoch": 0.0926,
      "grad_norm": 0.25393983721733093,
      "learning_rate": 0.0001814804,
      "loss": 0.3512,
      "step": 46300
    },
    {
      "epoch": 0.0928,
      "grad_norm": 0.19703878462314606,
      "learning_rate": 0.0001814404,
      "loss": 0.3468,
      "step": 46400
    },
    {
      "epoch": 0.093,
      "grad_norm": 0.18077781796455383,
      "learning_rate": 0.0001814004,
      "loss": 0.3529,
      "step": 46500
    },
    {
      "epoch": 0.0932,
      "grad_norm": 0.21559618413448334,
      "learning_rate": 0.0001813604,
      "loss": 0.3454,
      "step": 46600
    },
    {
      "epoch": 0.0934,
      "grad_norm": 0.223658949136734,
      "learning_rate": 0.0001813204,
      "loss": 0.3545,
      "step": 46700
    },
    {
      "epoch": 0.0936,
      "grad_norm": 0.2453765571117401,
      "learning_rate": 0.00018128040000000003,
      "loss": 0.3493,
      "step": 46800
    },
    {
      "epoch": 0.0938,
      "grad_norm": 0.2784949541091919,
      "learning_rate": 0.0001812404,
      "loss": 0.3432,
      "step": 46900
    },
    {
      "epoch": 0.094,
      "grad_norm": 0.21154820919036865,
      "learning_rate": 0.00018120040000000001,
      "loss": 0.347,
      "step": 47000
    },
    {
      "epoch": 0.0942,
      "grad_norm": 0.2772074341773987,
      "learning_rate": 0.0001811604,
      "loss": 0.3495,
      "step": 47100
    },
    {
      "epoch": 0.0944,
      "grad_norm": 0.20805895328521729,
      "learning_rate": 0.0001811204,
      "loss": 0.361,
      "step": 47200
    },
    {
      "epoch": 0.0946,
      "grad_norm": 0.1839285045862198,
      "learning_rate": 0.00018108040000000002,
      "loss": 0.3557,
      "step": 47300
    },
    {
      "epoch": 0.0948,
      "grad_norm": 0.23907071352005005,
      "learning_rate": 0.0001810404,
      "loss": 0.3416,
      "step": 47400
    },
    {
      "epoch": 0.095,
      "grad_norm": 0.3368641138076782,
      "learning_rate": 0.0001810004,
      "loss": 0.3483,
      "step": 47500
    },
    {
      "epoch": 0.0952,
      "grad_norm": 0.24331678450107574,
      "learning_rate": 0.0001809604,
      "loss": 0.3494,
      "step": 47600
    },
    {
      "epoch": 0.0954,
      "grad_norm": 0.20715878903865814,
      "learning_rate": 0.00018092040000000002,
      "loss": 0.3457,
      "step": 47700
    },
    {
      "epoch": 0.0956,
      "grad_norm": 0.1883363276720047,
      "learning_rate": 0.00018088040000000002,
      "loss": 0.3556,
      "step": 47800
    },
    {
      "epoch": 0.0958,
      "grad_norm": 0.24465380609035492,
      "learning_rate": 0.0001808404,
      "loss": 0.3447,
      "step": 47900
    },
    {
      "epoch": 0.096,
      "grad_norm": 0.24823075532913208,
      "learning_rate": 0.0001808004,
      "loss": 0.3431,
      "step": 48000
    },
    {
      "epoch": 0.0962,
      "grad_norm": 0.3323017358779907,
      "learning_rate": 0.0001807604,
      "loss": 0.3515,
      "step": 48100
    },
    {
      "epoch": 0.0964,
      "grad_norm": 0.2804560363292694,
      "learning_rate": 0.00018072040000000002,
      "loss": 0.351,
      "step": 48200
    },
    {
      "epoch": 0.0966,
      "grad_norm": 0.26708027720451355,
      "learning_rate": 0.0001806804,
      "loss": 0.338,
      "step": 48300
    },
    {
      "epoch": 0.0968,
      "grad_norm": 0.20081064105033875,
      "learning_rate": 0.0001806404,
      "loss": 0.3424,
      "step": 48400
    },
    {
      "epoch": 0.097,
      "grad_norm": 0.19008444249629974,
      "learning_rate": 0.0001806004,
      "loss": 0.3449,
      "step": 48500
    },
    {
      "epoch": 0.0972,
      "grad_norm": 0.2081468105316162,
      "learning_rate": 0.00018056040000000002,
      "loss": 0.3419,
      "step": 48600
    },
    {
      "epoch": 0.0974,
      "grad_norm": 0.41435888409614563,
      "learning_rate": 0.00018052040000000001,
      "loss": 0.345,
      "step": 48700
    },
    {
      "epoch": 0.0976,
      "grad_norm": 0.30877628922462463,
      "learning_rate": 0.0001804804,
      "loss": 0.3457,
      "step": 48800
    },
    {
      "epoch": 0.0978,
      "grad_norm": 0.25084757804870605,
      "learning_rate": 0.0001804404,
      "loss": 0.3436,
      "step": 48900
    },
    {
      "epoch": 0.098,
      "grad_norm": 0.26010265946388245,
      "learning_rate": 0.0001804004,
      "loss": 0.3522,
      "step": 49000
    },
    {
      "epoch": 0.0982,
      "grad_norm": 0.45441487431526184,
      "learning_rate": 0.00018036040000000002,
      "loss": 0.3366,
      "step": 49100
    },
    {
      "epoch": 0.0984,
      "grad_norm": 0.2062908411026001,
      "learning_rate": 0.0001803204,
      "loss": 0.3415,
      "step": 49200
    },
    {
      "epoch": 0.0986,
      "grad_norm": 0.24297113716602325,
      "learning_rate": 0.0001802804,
      "loss": 0.3403,
      "step": 49300
    },
    {
      "epoch": 0.0988,
      "grad_norm": 0.2683410942554474,
      "learning_rate": 0.0001802404,
      "loss": 0.3234,
      "step": 49400
    },
    {
      "epoch": 0.099,
      "grad_norm": 0.2585706114768982,
      "learning_rate": 0.0001802004,
      "loss": 0.3402,
      "step": 49500
    },
    {
      "epoch": 0.0992,
      "grad_norm": 0.2423756718635559,
      "learning_rate": 0.0001801604,
      "loss": 0.3394,
      "step": 49600
    },
    {
      "epoch": 0.0994,
      "grad_norm": 0.20874212682247162,
      "learning_rate": 0.0001801204,
      "loss": 0.3469,
      "step": 49700
    },
    {
      "epoch": 0.0996,
      "grad_norm": 0.26174086332321167,
      "learning_rate": 0.00018008040000000003,
      "loss": 0.3424,
      "step": 49800
    },
    {
      "epoch": 0.0998,
      "grad_norm": 0.258007675409317,
      "learning_rate": 0.0001800404,
      "loss": 0.333,
      "step": 49900
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.26558035612106323,
      "learning_rate": 0.0001800004,
      "loss": 0.3371,
      "step": 50000
    },
    {
      "epoch": 0.1002,
      "grad_norm": 0.2108493149280548,
      "learning_rate": 0.0001799604,
      "loss": 0.3439,
      "step": 50100
    },
    {
      "epoch": 0.1004,
      "grad_norm": 0.2356238216161728,
      "learning_rate": 0.0001799204,
      "loss": 0.3479,
      "step": 50200
    },
    {
      "epoch": 0.1006,
      "grad_norm": 0.22705116868019104,
      "learning_rate": 0.00017988040000000002,
      "loss": 0.3384,
      "step": 50300
    },
    {
      "epoch": 0.1008,
      "grad_norm": 0.22554484009742737,
      "learning_rate": 0.0001798404,
      "loss": 0.3398,
      "step": 50400
    },
    {
      "epoch": 0.101,
      "grad_norm": 0.23426580429077148,
      "learning_rate": 0.0001798004,
      "loss": 0.3463,
      "step": 50500
    },
    {
      "epoch": 0.1012,
      "grad_norm": 0.22517123818397522,
      "learning_rate": 0.0001797604,
      "loss": 0.3386,
      "step": 50600
    },
    {
      "epoch": 0.1014,
      "grad_norm": 0.24184340238571167,
      "learning_rate": 0.00017972040000000002,
      "loss": 0.3437,
      "step": 50700
    },
    {
      "epoch": 0.1016,
      "grad_norm": 0.3470735251903534,
      "learning_rate": 0.00017968040000000002,
      "loss": 0.3407,
      "step": 50800
    },
    {
      "epoch": 0.1018,
      "grad_norm": 0.35720571875572205,
      "learning_rate": 0.0001796404,
      "loss": 0.3379,
      "step": 50900
    },
    {
      "epoch": 0.102,
      "grad_norm": 0.25200462341308594,
      "learning_rate": 0.0001796004,
      "loss": 0.3454,
      "step": 51000
    },
    {
      "epoch": 0.1022,
      "grad_norm": 0.19085712730884552,
      "learning_rate": 0.0001795604,
      "loss": 0.3406,
      "step": 51100
    },
    {
      "epoch": 0.1024,
      "grad_norm": 0.21733929216861725,
      "learning_rate": 0.00017952040000000002,
      "loss": 0.3361,
      "step": 51200
    },
    {
      "epoch": 0.1026,
      "grad_norm": 0.2199040800333023,
      "learning_rate": 0.0001794804,
      "loss": 0.3381,
      "step": 51300
    },
    {
      "epoch": 0.1028,
      "grad_norm": 0.21806757152080536,
      "learning_rate": 0.00017944040000000003,
      "loss": 0.3391,
      "step": 51400
    },
    {
      "epoch": 0.103,
      "grad_norm": 0.27425751090049744,
      "learning_rate": 0.0001794004,
      "loss": 0.3336,
      "step": 51500
    },
    {
      "epoch": 0.1032,
      "grad_norm": 0.24505355954170227,
      "learning_rate": 0.00017936040000000002,
      "loss": 0.3384,
      "step": 51600
    },
    {
      "epoch": 0.1034,
      "grad_norm": 0.19927924871444702,
      "learning_rate": 0.0001793204,
      "loss": 0.3559,
      "step": 51700
    },
    {
      "epoch": 0.1036,
      "grad_norm": 0.31233856081962585,
      "learning_rate": 0.0001792804,
      "loss": 0.3369,
      "step": 51800
    },
    {
      "epoch": 0.1038,
      "grad_norm": 0.30651381611824036,
      "learning_rate": 0.00017924040000000003,
      "loss": 0.3579,
      "step": 51900
    },
    {
      "epoch": 0.104,
      "grad_norm": 0.18664880096912384,
      "learning_rate": 0.0001792004,
      "loss": 0.343,
      "step": 52000
    },
    {
      "epoch": 0.1042,
      "grad_norm": 0.48104581236839294,
      "learning_rate": 0.0001791604,
      "loss": 0.3705,
      "step": 52100
    },
    {
      "epoch": 0.1044,
      "grad_norm": 0.24804629385471344,
      "learning_rate": 0.0001791204,
      "loss": 0.3563,
      "step": 52200
    },
    {
      "epoch": 0.1046,
      "grad_norm": 0.2167651504278183,
      "learning_rate": 0.00017908040000000003,
      "loss": 0.3396,
      "step": 52300
    },
    {
      "epoch": 0.1048,
      "grad_norm": 0.5007449984550476,
      "learning_rate": 0.00017904040000000002,
      "loss": 0.3436,
      "step": 52400
    },
    {
      "epoch": 0.105,
      "grad_norm": 0.2310677170753479,
      "learning_rate": 0.0001790004,
      "loss": 0.3355,
      "step": 52500
    },
    {
      "epoch": 0.1052,
      "grad_norm": 0.26683372259140015,
      "learning_rate": 0.0001789604,
      "loss": 0.3472,
      "step": 52600
    },
    {
      "epoch": 0.1054,
      "grad_norm": 0.22053886950016022,
      "learning_rate": 0.0001789204,
      "loss": 0.3393,
      "step": 52700
    },
    {
      "epoch": 0.1056,
      "grad_norm": 0.24362322688102722,
      "learning_rate": 0.00017888040000000002,
      "loss": 0.3478,
      "step": 52800
    },
    {
      "epoch": 0.1058,
      "grad_norm": 0.28843867778778076,
      "learning_rate": 0.00017884040000000002,
      "loss": 0.3347,
      "step": 52900
    },
    {
      "epoch": 0.106,
      "grad_norm": 0.24268370866775513,
      "learning_rate": 0.0001788004,
      "loss": 0.3479,
      "step": 53000
    },
    {
      "epoch": 0.1062,
      "grad_norm": 0.25353243947029114,
      "learning_rate": 0.0001787604,
      "loss": 0.3503,
      "step": 53100
    },
    {
      "epoch": 0.1064,
      "grad_norm": 0.34019261598587036,
      "learning_rate": 0.0001787204,
      "loss": 0.3564,
      "step": 53200
    },
    {
      "epoch": 0.1066,
      "grad_norm": 0.28113237023353577,
      "learning_rate": 0.00017868040000000002,
      "loss": 0.3409,
      "step": 53300
    },
    {
      "epoch": 0.1068,
      "grad_norm": 0.33453458547592163,
      "learning_rate": 0.0001786404,
      "loss": 0.3393,
      "step": 53400
    },
    {
      "epoch": 0.107,
      "grad_norm": 0.25770464539527893,
      "learning_rate": 0.0001786004,
      "loss": 0.3448,
      "step": 53500
    },
    {
      "epoch": 0.1072,
      "grad_norm": 0.2797645628452301,
      "learning_rate": 0.0001785604,
      "loss": 0.3463,
      "step": 53600
    },
    {
      "epoch": 0.1074,
      "grad_norm": 0.26500073075294495,
      "learning_rate": 0.00017852040000000002,
      "loss": 0.3391,
      "step": 53700
    },
    {
      "epoch": 0.1076,
      "grad_norm": 0.23151659965515137,
      "learning_rate": 0.0001784804,
      "loss": 0.3369,
      "step": 53800
    },
    {
      "epoch": 0.1078,
      "grad_norm": 0.27178940176963806,
      "learning_rate": 0.0001784404,
      "loss": 0.3364,
      "step": 53900
    },
    {
      "epoch": 0.108,
      "grad_norm": 0.2767798900604248,
      "learning_rate": 0.0001784004,
      "loss": 0.3379,
      "step": 54000
    },
    {
      "epoch": 0.1082,
      "grad_norm": 0.21509261429309845,
      "learning_rate": 0.0001783604,
      "loss": 0.3382,
      "step": 54100
    },
    {
      "epoch": 0.1084,
      "grad_norm": 0.24864695966243744,
      "learning_rate": 0.00017832040000000001,
      "loss": 0.3369,
      "step": 54200
    },
    {
      "epoch": 0.1086,
      "grad_norm": 0.2107497751712799,
      "learning_rate": 0.0001782804,
      "loss": 0.333,
      "step": 54300
    },
    {
      "epoch": 0.1088,
      "grad_norm": 0.20573535561561584,
      "learning_rate": 0.00017824040000000003,
      "loss": 0.3397,
      "step": 54400
    },
    {
      "epoch": 0.109,
      "grad_norm": 0.25302717089653015,
      "learning_rate": 0.0001782004,
      "loss": 0.3384,
      "step": 54500
    },
    {
      "epoch": 0.1092,
      "grad_norm": 0.22738224267959595,
      "learning_rate": 0.00017816040000000002,
      "loss": 0.3398,
      "step": 54600
    },
    {
      "epoch": 0.1094,
      "grad_norm": 0.28848886489868164,
      "learning_rate": 0.0001781204,
      "loss": 0.3361,
      "step": 54700
    },
    {
      "epoch": 0.1096,
      "grad_norm": 0.23477931320667267,
      "learning_rate": 0.0001780804,
      "loss": 0.3414,
      "step": 54800
    },
    {
      "epoch": 0.1098,
      "grad_norm": 0.24742473661899567,
      "learning_rate": 0.00017804040000000002,
      "loss": 0.3359,
      "step": 54900
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.24106016755104065,
      "learning_rate": 0.0001780004,
      "loss": 0.3428,
      "step": 55000
    },
    {
      "epoch": 0.1102,
      "grad_norm": 0.22427873313426971,
      "learning_rate": 0.0001779604,
      "loss": 0.346,
      "step": 55100
    },
    {
      "epoch": 0.1104,
      "grad_norm": 0.31361493468284607,
      "learning_rate": 0.0001779204,
      "loss": 0.3454,
      "step": 55200
    },
    {
      "epoch": 0.1106,
      "grad_norm": 0.2146797627210617,
      "learning_rate": 0.00017788040000000003,
      "loss": 0.3368,
      "step": 55300
    },
    {
      "epoch": 0.1108,
      "grad_norm": 0.2894279658794403,
      "learning_rate": 0.00017784040000000002,
      "loss": 0.339,
      "step": 55400
    },
    {
      "epoch": 0.111,
      "grad_norm": 0.3044085204601288,
      "learning_rate": 0.0001778004,
      "loss": 0.3411,
      "step": 55500
    },
    {
      "epoch": 0.1112,
      "grad_norm": 0.18939179182052612,
      "learning_rate": 0.0001777604,
      "loss": 0.3471,
      "step": 55600
    },
    {
      "epoch": 0.1114,
      "grad_norm": 0.24099594354629517,
      "learning_rate": 0.0001777204,
      "loss": 0.3296,
      "step": 55700
    },
    {
      "epoch": 0.1116,
      "grad_norm": 0.2313549667596817,
      "learning_rate": 0.00017768040000000002,
      "loss": 0.3384,
      "step": 55800
    },
    {
      "epoch": 0.1118,
      "grad_norm": 0.20545341074466705,
      "learning_rate": 0.00017764040000000001,
      "loss": 0.331,
      "step": 55900
    },
    {
      "epoch": 0.112,
      "grad_norm": 0.22012929618358612,
      "learning_rate": 0.0001776004,
      "loss": 0.3383,
      "step": 56000
    },
    {
      "epoch": 0.1122,
      "grad_norm": 0.2095547914505005,
      "learning_rate": 0.0001775604,
      "loss": 0.3355,
      "step": 56100
    },
    {
      "epoch": 0.1124,
      "grad_norm": 0.2438439428806305,
      "learning_rate": 0.00017752040000000002,
      "loss": 0.3305,
      "step": 56200
    },
    {
      "epoch": 0.1126,
      "grad_norm": 0.5170906186103821,
      "learning_rate": 0.00017748040000000002,
      "loss": 0.3385,
      "step": 56300
    },
    {
      "epoch": 0.1128,
      "grad_norm": 0.1860545575618744,
      "learning_rate": 0.0001774404,
      "loss": 0.3336,
      "step": 56400
    },
    {
      "epoch": 0.113,
      "grad_norm": 0.24801748991012573,
      "learning_rate": 0.0001774004,
      "loss": 0.3286,
      "step": 56500
    },
    {
      "epoch": 0.1132,
      "grad_norm": 0.25407829880714417,
      "learning_rate": 0.0001773604,
      "loss": 0.3402,
      "step": 56600
    },
    {
      "epoch": 0.1134,
      "grad_norm": 0.20441265404224396,
      "learning_rate": 0.00017732040000000002,
      "loss": 0.3336,
      "step": 56700
    },
    {
      "epoch": 0.1136,
      "grad_norm": 0.2255195528268814,
      "learning_rate": 0.0001772804,
      "loss": 0.3283,
      "step": 56800
    },
    {
      "epoch": 0.1138,
      "grad_norm": 0.3453531265258789,
      "learning_rate": 0.0001772404,
      "loss": 0.3325,
      "step": 56900
    },
    {
      "epoch": 0.114,
      "grad_norm": 0.1892947405576706,
      "learning_rate": 0.0001772004,
      "loss": 0.3298,
      "step": 57000
    },
    {
      "epoch": 0.1142,
      "grad_norm": 0.3136577308177948,
      "learning_rate": 0.0001771604,
      "loss": 0.3324,
      "step": 57100
    },
    {
      "epoch": 0.1144,
      "grad_norm": 0.2848574221134186,
      "learning_rate": 0.0001771204,
      "loss": 0.3266,
      "step": 57200
    },
    {
      "epoch": 0.1146,
      "grad_norm": 0.3240659832954407,
      "learning_rate": 0.0001770804,
      "loss": 0.3413,
      "step": 57300
    },
    {
      "epoch": 0.1148,
      "grad_norm": 0.2988292872905731,
      "learning_rate": 0.00017704040000000003,
      "loss": 0.3335,
      "step": 57400
    },
    {
      "epoch": 0.115,
      "grad_norm": 0.31577932834625244,
      "learning_rate": 0.0001770004,
      "loss": 0.3294,
      "step": 57500
    },
    {
      "epoch": 0.1152,
      "grad_norm": 0.32162484526634216,
      "learning_rate": 0.00017696040000000001,
      "loss": 0.3521,
      "step": 57600
    },
    {
      "epoch": 0.1154,
      "grad_norm": 0.661695122718811,
      "learning_rate": 0.0001769204,
      "loss": 0.3324,
      "step": 57700
    },
    {
      "epoch": 0.1156,
      "grad_norm": 0.25184112787246704,
      "learning_rate": 0.0001768804,
      "loss": 0.3438,
      "step": 57800
    },
    {
      "epoch": 0.1158,
      "grad_norm": 0.24458423256874084,
      "learning_rate": 0.00017684040000000002,
      "loss": 0.3297,
      "step": 57900
    },
    {
      "epoch": 0.116,
      "grad_norm": 0.23055191338062286,
      "learning_rate": 0.0001768004,
      "loss": 0.328,
      "step": 58000
    },
    {
      "epoch": 0.1162,
      "grad_norm": 0.22799059748649597,
      "learning_rate": 0.0001767604,
      "loss": 0.3305,
      "step": 58100
    },
    {
      "epoch": 0.1164,
      "grad_norm": 0.17628814280033112,
      "learning_rate": 0.0001767204,
      "loss": 0.3351,
      "step": 58200
    },
    {
      "epoch": 0.1166,
      "grad_norm": 0.29074835777282715,
      "learning_rate": 0.00017668040000000002,
      "loss": 0.3267,
      "step": 58300
    },
    {
      "epoch": 0.1168,
      "grad_norm": 0.26170602440834045,
      "learning_rate": 0.00017664040000000002,
      "loss": 0.332,
      "step": 58400
    },
    {
      "epoch": 0.117,
      "grad_norm": 0.18410225212574005,
      "learning_rate": 0.0001766004,
      "loss": 0.3291,
      "step": 58500
    },
    {
      "epoch": 0.1172,
      "grad_norm": 0.3183979392051697,
      "learning_rate": 0.0001765604,
      "loss": 0.3311,
      "step": 58600
    },
    {
      "epoch": 0.1174,
      "grad_norm": 0.21326270699501038,
      "learning_rate": 0.0001765204,
      "loss": 0.3279,
      "step": 58700
    },
    {
      "epoch": 0.1176,
      "grad_norm": 0.2310231775045395,
      "learning_rate": 0.00017648040000000002,
      "loss": 0.3267,
      "step": 58800
    },
    {
      "epoch": 0.1178,
      "grad_norm": 0.30372264981269836,
      "learning_rate": 0.0001764404,
      "loss": 0.3411,
      "step": 58900
    },
    {
      "epoch": 0.118,
      "grad_norm": 0.22354651987552643,
      "learning_rate": 0.0001764004,
      "loss": 0.3323,
      "step": 59000
    },
    {
      "epoch": 0.1182,
      "grad_norm": 0.22137747704982758,
      "learning_rate": 0.0001763604,
      "loss": 0.3261,
      "step": 59100
    },
    {
      "epoch": 0.1184,
      "grad_norm": 0.20340581238269806,
      "learning_rate": 0.00017632040000000002,
      "loss": 0.3291,
      "step": 59200
    },
    {
      "epoch": 0.1186,
      "grad_norm": 0.21525894105434418,
      "learning_rate": 0.00017628040000000001,
      "loss": 0.324,
      "step": 59300
    },
    {
      "epoch": 0.1188,
      "grad_norm": 0.21832208335399628,
      "learning_rate": 0.0001762404,
      "loss": 0.3319,
      "step": 59400
    },
    {
      "epoch": 0.119,
      "grad_norm": 0.2369861602783203,
      "learning_rate": 0.0001762004,
      "loss": 0.3298,
      "step": 59500
    },
    {
      "epoch": 0.1192,
      "grad_norm": 0.4107494652271271,
      "learning_rate": 0.0001761604,
      "loss": 0.3306,
      "step": 59600
    },
    {
      "epoch": 0.1194,
      "grad_norm": 0.23435452580451965,
      "learning_rate": 0.00017612040000000002,
      "loss": 0.3239,
      "step": 59700
    },
    {
      "epoch": 0.1196,
      "grad_norm": 0.20295970141887665,
      "learning_rate": 0.0001760804,
      "loss": 0.335,
      "step": 59800
    },
    {
      "epoch": 0.1198,
      "grad_norm": 0.29473334550857544,
      "learning_rate": 0.00017604040000000003,
      "loss": 0.332,
      "step": 59900
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.22703693807125092,
      "learning_rate": 0.0001760004,
      "loss": 0.3291,
      "step": 60000
    },
    {
      "epoch": 0.1202,
      "grad_norm": 0.2531866729259491,
      "learning_rate": 0.0001759604,
      "loss": 0.3326,
      "step": 60100
    },
    {
      "epoch": 0.1204,
      "grad_norm": 0.20017871260643005,
      "learning_rate": 0.0001759204,
      "loss": 0.3304,
      "step": 60200
    },
    {
      "epoch": 0.1206,
      "grad_norm": 0.22206462919712067,
      "learning_rate": 0.0001758804,
      "loss": 0.3442,
      "step": 60300
    },
    {
      "epoch": 0.1208,
      "grad_norm": 0.24844832718372345,
      "learning_rate": 0.00017584040000000002,
      "loss": 0.3301,
      "step": 60400
    },
    {
      "epoch": 0.121,
      "grad_norm": 0.19882357120513916,
      "learning_rate": 0.0001758004,
      "loss": 0.3296,
      "step": 60500
    },
    {
      "epoch": 0.1212,
      "grad_norm": 0.2944953441619873,
      "learning_rate": 0.0001757604,
      "loss": 0.3283,
      "step": 60600
    },
    {
      "epoch": 0.1214,
      "grad_norm": 0.21407614648342133,
      "learning_rate": 0.0001757204,
      "loss": 0.328,
      "step": 60700
    },
    {
      "epoch": 0.1216,
      "grad_norm": 0.19046856462955475,
      "learning_rate": 0.0001756804,
      "loss": 0.3292,
      "step": 60800
    },
    {
      "epoch": 0.1218,
      "grad_norm": 0.22615882754325867,
      "learning_rate": 0.00017564040000000002,
      "loss": 0.3316,
      "step": 60900
    },
    {
      "epoch": 0.122,
      "grad_norm": 0.1709730625152588,
      "learning_rate": 0.0001756004,
      "loss": 0.3318,
      "step": 61000
    },
    {
      "epoch": 0.1222,
      "grad_norm": 0.1718514859676361,
      "learning_rate": 0.0001755604,
      "loss": 0.3201,
      "step": 61100
    },
    {
      "epoch": 0.1224,
      "grad_norm": 0.23949329555034637,
      "learning_rate": 0.0001755204,
      "loss": 0.3329,
      "step": 61200
    },
    {
      "epoch": 0.1226,
      "grad_norm": 0.2090582698583603,
      "learning_rate": 0.00017548040000000002,
      "loss": 0.3354,
      "step": 61300
    },
    {
      "epoch": 0.1228,
      "grad_norm": 0.2083326280117035,
      "learning_rate": 0.00017544040000000002,
      "loss": 0.3293,
      "step": 61400
    },
    {
      "epoch": 0.123,
      "grad_norm": 0.21610133349895477,
      "learning_rate": 0.0001754004,
      "loss": 0.3252,
      "step": 61500
    },
    {
      "epoch": 0.1232,
      "grad_norm": 0.2022552639245987,
      "learning_rate": 0.0001753604,
      "loss": 0.3322,
      "step": 61600
    },
    {
      "epoch": 0.1234,
      "grad_norm": 0.24811524152755737,
      "learning_rate": 0.0001753204,
      "loss": 0.3405,
      "step": 61700
    },
    {
      "epoch": 0.1236,
      "grad_norm": 0.25753867626190186,
      "learning_rate": 0.00017528040000000002,
      "loss": 0.3359,
      "step": 61800
    },
    {
      "epoch": 0.1238,
      "grad_norm": 0.32304272055625916,
      "learning_rate": 0.0001752404,
      "loss": 0.3254,
      "step": 61900
    },
    {
      "epoch": 0.124,
      "grad_norm": 0.18287380039691925,
      "learning_rate": 0.0001752004,
      "loss": 0.3298,
      "step": 62000
    },
    {
      "epoch": 0.1242,
      "grad_norm": 0.19973132014274597,
      "learning_rate": 0.0001751604,
      "loss": 0.3363,
      "step": 62100
    },
    {
      "epoch": 0.1244,
      "grad_norm": 0.23819290101528168,
      "learning_rate": 0.00017512040000000002,
      "loss": 0.3238,
      "step": 62200
    },
    {
      "epoch": 0.1246,
      "grad_norm": 0.27806952595710754,
      "learning_rate": 0.0001750804,
      "loss": 0.3286,
      "step": 62300
    },
    {
      "epoch": 0.1248,
      "grad_norm": 0.17436453700065613,
      "learning_rate": 0.0001750404,
      "loss": 0.3266,
      "step": 62400
    },
    {
      "epoch": 0.125,
      "grad_norm": 0.2046598643064499,
      "learning_rate": 0.0001750004,
      "loss": 0.32,
      "step": 62500
    },
    {
      "epoch": 0.1252,
      "grad_norm": 0.23056712746620178,
      "learning_rate": 0.0001749604,
      "loss": 0.325,
      "step": 62600
    },
    {
      "epoch": 0.1254,
      "grad_norm": 0.32476136088371277,
      "learning_rate": 0.0001749204,
      "loss": 0.3301,
      "step": 62700
    },
    {
      "epoch": 0.1256,
      "grad_norm": 0.24279266595840454,
      "learning_rate": 0.0001748804,
      "loss": 0.3253,
      "step": 62800
    },
    {
      "epoch": 0.1258,
      "grad_norm": 0.1884712129831314,
      "learning_rate": 0.00017484040000000003,
      "loss": 0.3309,
      "step": 62900
    },
    {
      "epoch": 0.126,
      "grad_norm": 0.21700552105903625,
      "learning_rate": 0.0001748004,
      "loss": 0.3299,
      "step": 63000
    },
    {
      "epoch": 0.1262,
      "grad_norm": 0.22113928198814392,
      "learning_rate": 0.00017476040000000001,
      "loss": 0.3283,
      "step": 63100
    },
    {
      "epoch": 0.1264,
      "grad_norm": 0.2137170284986496,
      "learning_rate": 0.0001747204,
      "loss": 0.3243,
      "step": 63200
    },
    {
      "epoch": 0.1266,
      "grad_norm": 0.273173063993454,
      "learning_rate": 0.0001746804,
      "loss": 0.3337,
      "step": 63300
    },
    {
      "epoch": 0.1268,
      "grad_norm": 0.2576025724411011,
      "learning_rate": 0.00017464040000000002,
      "loss": 0.329,
      "step": 63400
    },
    {
      "epoch": 0.127,
      "grad_norm": 0.3533582389354706,
      "learning_rate": 0.0001746004,
      "loss": 0.3322,
      "step": 63500
    },
    {
      "epoch": 0.1272,
      "grad_norm": 0.373040109872818,
      "learning_rate": 0.0001745604,
      "loss": 0.3262,
      "step": 63600
    },
    {
      "epoch": 0.1274,
      "grad_norm": 0.2242802381515503,
      "learning_rate": 0.0001745204,
      "loss": 0.328,
      "step": 63700
    },
    {
      "epoch": 0.1276,
      "grad_norm": 0.4904221296310425,
      "learning_rate": 0.0001744804,
      "loss": 0.3253,
      "step": 63800
    },
    {
      "epoch": 0.1278,
      "grad_norm": 0.20114104449748993,
      "learning_rate": 0.00017444040000000002,
      "loss": 0.3272,
      "step": 63900
    },
    {
      "epoch": 0.128,
      "grad_norm": 0.18482904136180878,
      "learning_rate": 0.0001744004,
      "loss": 0.3254,
      "step": 64000
    },
    {
      "epoch": 0.1282,
      "grad_norm": 0.19531016051769257,
      "learning_rate": 0.0001743604,
      "loss": 0.3206,
      "step": 64100
    },
    {
      "epoch": 0.1284,
      "grad_norm": 0.2820129990577698,
      "learning_rate": 0.0001743204,
      "loss": 0.3324,
      "step": 64200
    },
    {
      "epoch": 0.1286,
      "grad_norm": 0.23063960671424866,
      "learning_rate": 0.00017428040000000002,
      "loss": 0.3262,
      "step": 64300
    },
    {
      "epoch": 0.1288,
      "grad_norm": 0.2425381988286972,
      "learning_rate": 0.0001742404,
      "loss": 0.3221,
      "step": 64400
    },
    {
      "epoch": 0.129,
      "grad_norm": 0.1685548573732376,
      "learning_rate": 0.0001742004,
      "loss": 0.3225,
      "step": 64500
    },
    {
      "epoch": 0.1292,
      "grad_norm": 0.23022134602069855,
      "learning_rate": 0.0001741604,
      "loss": 0.3297,
      "step": 64600
    },
    {
      "epoch": 0.1294,
      "grad_norm": 0.2650277018547058,
      "learning_rate": 0.0001741204,
      "loss": 0.3293,
      "step": 64700
    },
    {
      "epoch": 0.1296,
      "grad_norm": 0.2417951077222824,
      "learning_rate": 0.00017408040000000001,
      "loss": 0.328,
      "step": 64800
    },
    {
      "epoch": 0.1298,
      "grad_norm": 0.22391964495182037,
      "learning_rate": 0.0001740404,
      "loss": 0.3266,
      "step": 64900
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.2012777179479599,
      "learning_rate": 0.00017400040000000003,
      "loss": 0.3278,
      "step": 65000
    },
    {
      "epoch": 0.1302,
      "grad_norm": 0.18540987372398376,
      "learning_rate": 0.0001739604,
      "loss": 0.337,
      "step": 65100
    },
    {
      "epoch": 0.1304,
      "grad_norm": 0.19150865077972412,
      "learning_rate": 0.00017392040000000002,
      "loss": 0.3319,
      "step": 65200
    },
    {
      "epoch": 0.1306,
      "grad_norm": 0.24447111785411835,
      "learning_rate": 0.0001738804,
      "loss": 0.3312,
      "step": 65300
    },
    {
      "epoch": 0.1308,
      "grad_norm": 0.20017050206661224,
      "learning_rate": 0.0001738404,
      "loss": 0.3375,
      "step": 65400
    },
    {
      "epoch": 0.131,
      "grad_norm": 0.2571297585964203,
      "learning_rate": 0.00017380040000000002,
      "loss": 0.3268,
      "step": 65500
    },
    {
      "epoch": 0.1312,
      "grad_norm": 0.18792451918125153,
      "learning_rate": 0.0001737604,
      "loss": 0.3261,
      "step": 65600
    },
    {
      "epoch": 0.1314,
      "grad_norm": 0.4205167889595032,
      "learning_rate": 0.0001737204,
      "loss": 0.33,
      "step": 65700
    },
    {
      "epoch": 0.1316,
      "grad_norm": 0.2724183201789856,
      "learning_rate": 0.0001736804,
      "loss": 0.3273,
      "step": 65800
    },
    {
      "epoch": 0.1318,
      "grad_norm": 0.21143561601638794,
      "learning_rate": 0.00017364040000000003,
      "loss": 0.3281,
      "step": 65900
    },
    {
      "epoch": 0.132,
      "grad_norm": 0.1903173327445984,
      "learning_rate": 0.00017360040000000002,
      "loss": 0.3185,
      "step": 66000
    },
    {
      "epoch": 0.1322,
      "grad_norm": 0.3626430928707123,
      "learning_rate": 0.0001735604,
      "loss": 0.3411,
      "step": 66100
    },
    {
      "epoch": 0.1324,
      "grad_norm": 0.1814277172088623,
      "learning_rate": 0.0001735204,
      "loss": 0.3301,
      "step": 66200
    },
    {
      "epoch": 0.1326,
      "grad_norm": 0.22784367203712463,
      "learning_rate": 0.0001734804,
      "loss": 0.3244,
      "step": 66300
    },
    {
      "epoch": 0.1328,
      "grad_norm": 0.2449122965335846,
      "learning_rate": 0.00017344040000000002,
      "loss": 0.3281,
      "step": 66400
    },
    {
      "epoch": 0.133,
      "grad_norm": 0.17180882394313812,
      "learning_rate": 0.00017340040000000001,
      "loss": 0.3237,
      "step": 66500
    },
    {
      "epoch": 0.1332,
      "grad_norm": 0.24725696444511414,
      "learning_rate": 0.0001733604,
      "loss": 0.3281,
      "step": 66600
    },
    {
      "epoch": 0.1334,
      "grad_norm": 0.31062814593315125,
      "learning_rate": 0.0001733204,
      "loss": 0.3312,
      "step": 66700
    },
    {
      "epoch": 0.1336,
      "grad_norm": 0.20241935551166534,
      "learning_rate": 0.00017328040000000002,
      "loss": 0.3214,
      "step": 66800
    },
    {
      "epoch": 0.1338,
      "grad_norm": 0.319897323846817,
      "learning_rate": 0.00017324040000000002,
      "loss": 0.3262,
      "step": 66900
    },
    {
      "epoch": 0.134,
      "grad_norm": 0.36917874217033386,
      "learning_rate": 0.0001732004,
      "loss": 0.3232,
      "step": 67000
    },
    {
      "epoch": 0.1342,
      "grad_norm": 0.24121050536632538,
      "learning_rate": 0.0001731604,
      "loss": 0.3333,
      "step": 67100
    },
    {
      "epoch": 0.1344,
      "grad_norm": 0.21649840474128723,
      "learning_rate": 0.0001731204,
      "loss": 0.3246,
      "step": 67200
    },
    {
      "epoch": 0.1346,
      "grad_norm": 0.21219566464424133,
      "learning_rate": 0.00017308040000000002,
      "loss": 0.3263,
      "step": 67300
    },
    {
      "epoch": 0.1348,
      "grad_norm": 0.20997853577136993,
      "learning_rate": 0.0001730404,
      "loss": 0.321,
      "step": 67400
    },
    {
      "epoch": 0.135,
      "grad_norm": 0.20444467663764954,
      "learning_rate": 0.0001730004,
      "loss": 0.3242,
      "step": 67500
    },
    {
      "epoch": 0.1352,
      "grad_norm": 0.28162842988967896,
      "learning_rate": 0.0001729604,
      "loss": 0.3191,
      "step": 67600
    },
    {
      "epoch": 0.1354,
      "grad_norm": 0.266450971364975,
      "learning_rate": 0.0001729204,
      "loss": 0.3201,
      "step": 67700
    },
    {
      "epoch": 0.1356,
      "grad_norm": 0.22291947901248932,
      "learning_rate": 0.0001728804,
      "loss": 0.3171,
      "step": 67800
    },
    {
      "epoch": 0.1358,
      "grad_norm": 0.46510088443756104,
      "learning_rate": 0.0001728404,
      "loss": 0.3276,
      "step": 67900
    },
    {
      "epoch": 0.136,
      "grad_norm": 0.2612142562866211,
      "learning_rate": 0.00017280040000000003,
      "loss": 0.3313,
      "step": 68000
    },
    {
      "epoch": 0.1362,
      "grad_norm": 0.19849452376365662,
      "learning_rate": 0.0001727604,
      "loss": 0.3155,
      "step": 68100
    },
    {
      "epoch": 0.1364,
      "grad_norm": 0.32336533069610596,
      "learning_rate": 0.00017272040000000001,
      "loss": 0.3225,
      "step": 68200
    },
    {
      "epoch": 0.1366,
      "grad_norm": 0.26454541087150574,
      "learning_rate": 0.0001726804,
      "loss": 0.333,
      "step": 68300
    },
    {
      "epoch": 0.1368,
      "grad_norm": 0.2315870225429535,
      "learning_rate": 0.0001726404,
      "loss": 0.3297,
      "step": 68400
    },
    {
      "epoch": 0.137,
      "grad_norm": 0.3298985958099365,
      "learning_rate": 0.00017260040000000002,
      "loss": 0.3261,
      "step": 68500
    },
    {
      "epoch": 0.1372,
      "grad_norm": 0.2369878590106964,
      "learning_rate": 0.0001725604,
      "loss": 0.3287,
      "step": 68600
    },
    {
      "epoch": 0.1374,
      "grad_norm": 0.18581421673297882,
      "learning_rate": 0.0001725204,
      "loss": 0.3221,
      "step": 68700
    },
    {
      "epoch": 0.1376,
      "grad_norm": 0.3875245153903961,
      "learning_rate": 0.0001724804,
      "loss": 0.3181,
      "step": 68800
    },
    {
      "epoch": 0.1378,
      "grad_norm": 0.216262087225914,
      "learning_rate": 0.00017244040000000002,
      "loss": 0.3169,
      "step": 68900
    },
    {
      "epoch": 0.138,
      "grad_norm": 0.21017438173294067,
      "learning_rate": 0.00017240040000000002,
      "loss": 0.3184,
      "step": 69000
    },
    {
      "epoch": 0.1382,
      "grad_norm": 0.20590710639953613,
      "learning_rate": 0.0001723604,
      "loss": 0.3287,
      "step": 69100
    },
    {
      "epoch": 0.1384,
      "grad_norm": 0.2776513695716858,
      "learning_rate": 0.0001723204,
      "loss": 0.3195,
      "step": 69200
    },
    {
      "epoch": 0.1386,
      "grad_norm": 0.21307912468910217,
      "learning_rate": 0.0001722804,
      "loss": 0.3277,
      "step": 69300
    },
    {
      "epoch": 0.1388,
      "grad_norm": 0.2287779301404953,
      "learning_rate": 0.00017224040000000002,
      "loss": 0.3199,
      "step": 69400
    },
    {
      "epoch": 0.139,
      "grad_norm": 0.33188334107398987,
      "learning_rate": 0.0001722004,
      "loss": 0.3167,
      "step": 69500
    },
    {
      "epoch": 0.1392,
      "grad_norm": 0.2860109508037567,
      "learning_rate": 0.0001721604,
      "loss": 0.3151,
      "step": 69600
    },
    {
      "epoch": 0.1394,
      "grad_norm": 0.19062092900276184,
      "learning_rate": 0.0001721204,
      "loss": 0.3237,
      "step": 69700
    },
    {
      "epoch": 0.1396,
      "grad_norm": 0.21721702814102173,
      "learning_rate": 0.00017208040000000002,
      "loss": 0.3173,
      "step": 69800
    },
    {
      "epoch": 0.1398,
      "grad_norm": 0.20423859357833862,
      "learning_rate": 0.00017204040000000001,
      "loss": 0.3197,
      "step": 69900
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.21749666333198547,
      "learning_rate": 0.0001720004,
      "loss": 0.3274,
      "step": 70000
    },
    {
      "epoch": 0.1402,
      "grad_norm": 0.2770256996154785,
      "learning_rate": 0.0001719604,
      "loss": 0.3189,
      "step": 70100
    },
    {
      "epoch": 0.1404,
      "grad_norm": 0.300563782453537,
      "learning_rate": 0.0001719204,
      "loss": 0.3158,
      "step": 70200
    },
    {
      "epoch": 0.1406,
      "grad_norm": 0.21529285609722137,
      "learning_rate": 0.00017188040000000002,
      "loss": 0.3234,
      "step": 70300
    },
    {
      "epoch": 0.1408,
      "grad_norm": 0.30113551020622253,
      "learning_rate": 0.0001718404,
      "loss": 0.3188,
      "step": 70400
    },
    {
      "epoch": 0.141,
      "grad_norm": 0.22355583310127258,
      "learning_rate": 0.00017180040000000003,
      "loss": 0.321,
      "step": 70500
    },
    {
      "epoch": 0.1412,
      "grad_norm": 0.23251134157180786,
      "learning_rate": 0.0001717604,
      "loss": 0.3215,
      "step": 70600
    },
    {
      "epoch": 0.1414,
      "grad_norm": 0.2621996998786926,
      "learning_rate": 0.00017172040000000002,
      "loss": 0.328,
      "step": 70700
    },
    {
      "epoch": 0.1416,
      "grad_norm": 0.24400697648525238,
      "learning_rate": 0.0001716804,
      "loss": 0.3182,
      "step": 70800
    },
    {
      "epoch": 0.1418,
      "grad_norm": 0.182724729180336,
      "learning_rate": 0.0001716404,
      "loss": 0.3104,
      "step": 70900
    },
    {
      "epoch": 0.142,
      "grad_norm": 0.2504395544528961,
      "learning_rate": 0.00017160040000000002,
      "loss": 0.3186,
      "step": 71000
    },
    {
      "epoch": 0.1422,
      "grad_norm": 0.3078869581222534,
      "learning_rate": 0.0001715604,
      "loss": 0.3175,
      "step": 71100
    },
    {
      "epoch": 0.1424,
      "grad_norm": 0.2186155915260315,
      "learning_rate": 0.0001715204,
      "loss": 0.32,
      "step": 71200
    },
    {
      "epoch": 0.1426,
      "grad_norm": 0.18938790261745453,
      "learning_rate": 0.0001714804,
      "loss": 0.3138,
      "step": 71300
    },
    {
      "epoch": 0.1428,
      "grad_norm": 0.21720623970031738,
      "learning_rate": 0.0001714404,
      "loss": 0.3192,
      "step": 71400
    },
    {
      "epoch": 0.143,
      "grad_norm": 0.18728028237819672,
      "learning_rate": 0.00017140040000000002,
      "loss": 0.3261,
      "step": 71500
    },
    {
      "epoch": 0.1432,
      "grad_norm": 0.23028717935085297,
      "learning_rate": 0.0001713604,
      "loss": 0.3167,
      "step": 71600
    },
    {
      "epoch": 0.1434,
      "grad_norm": 0.22258538007736206,
      "learning_rate": 0.0001713204,
      "loss": 0.3138,
      "step": 71700
    },
    {
      "epoch": 0.1436,
      "grad_norm": 0.3108178675174713,
      "learning_rate": 0.0001712804,
      "loss": 0.3261,
      "step": 71800
    },
    {
      "epoch": 0.1438,
      "grad_norm": 0.2817421853542328,
      "learning_rate": 0.00017124040000000002,
      "loss": 0.3153,
      "step": 71900
    },
    {
      "epoch": 0.144,
      "grad_norm": 0.2490064799785614,
      "learning_rate": 0.00017120040000000002,
      "loss": 0.314,
      "step": 72000
    },
    {
      "epoch": 0.1442,
      "grad_norm": 0.18532483279705048,
      "learning_rate": 0.0001711604,
      "loss": 0.3162,
      "step": 72100
    },
    {
      "epoch": 0.1444,
      "grad_norm": 0.23727016150951385,
      "learning_rate": 0.0001711204,
      "loss": 0.3278,
      "step": 72200
    },
    {
      "epoch": 0.1446,
      "grad_norm": 0.18576467037200928,
      "learning_rate": 0.0001710804,
      "loss": 0.3186,
      "step": 72300
    },
    {
      "epoch": 0.1448,
      "grad_norm": 0.21316267549991608,
      "learning_rate": 0.00017104040000000002,
      "loss": 0.3198,
      "step": 72400
    },
    {
      "epoch": 0.145,
      "grad_norm": 0.21670052409172058,
      "learning_rate": 0.0001710004,
      "loss": 0.3187,
      "step": 72500
    },
    {
      "epoch": 0.1452,
      "grad_norm": 0.260593056678772,
      "learning_rate": 0.0001709604,
      "loss": 0.3119,
      "step": 72600
    },
    {
      "epoch": 0.1454,
      "grad_norm": 0.24345095455646515,
      "learning_rate": 0.0001709204,
      "loss": 0.3081,
      "step": 72700
    },
    {
      "epoch": 0.1456,
      "grad_norm": 0.21282540261745453,
      "learning_rate": 0.00017088040000000002,
      "loss": 0.3222,
      "step": 72800
    },
    {
      "epoch": 0.1458,
      "grad_norm": 0.22302696108818054,
      "learning_rate": 0.0001708404,
      "loss": 0.3133,
      "step": 72900
    },
    {
      "epoch": 0.146,
      "grad_norm": 0.24244046211242676,
      "learning_rate": 0.0001708004,
      "loss": 0.3215,
      "step": 73000
    },
    {
      "epoch": 0.1462,
      "grad_norm": 0.651138961315155,
      "learning_rate": 0.0001707604,
      "loss": 0.3123,
      "step": 73100
    },
    {
      "epoch": 0.1464,
      "grad_norm": 0.27869105339050293,
      "learning_rate": 0.0001707204,
      "loss": 0.3184,
      "step": 73200
    },
    {
      "epoch": 0.1466,
      "grad_norm": 0.2049490511417389,
      "learning_rate": 0.0001706804,
      "loss": 0.3157,
      "step": 73300
    },
    {
      "epoch": 0.1468,
      "grad_norm": 0.2697393298149109,
      "learning_rate": 0.0001706404,
      "loss": 0.334,
      "step": 73400
    },
    {
      "epoch": 0.147,
      "grad_norm": 0.17334036529064178,
      "learning_rate": 0.00017060040000000003,
      "loss": 0.3145,
      "step": 73500
    },
    {
      "epoch": 0.1472,
      "grad_norm": 0.2551487386226654,
      "learning_rate": 0.0001705604,
      "loss": 0.3138,
      "step": 73600
    },
    {
      "epoch": 0.1474,
      "grad_norm": 0.23012271523475647,
      "learning_rate": 0.00017052040000000001,
      "loss": 0.3133,
      "step": 73700
    },
    {
      "epoch": 0.1476,
      "grad_norm": 0.21023279428482056,
      "learning_rate": 0.0001704804,
      "loss": 0.3096,
      "step": 73800
    },
    {
      "epoch": 0.1478,
      "grad_norm": 0.24528542160987854,
      "learning_rate": 0.0001704404,
      "loss": 0.313,
      "step": 73900
    },
    {
      "epoch": 0.148,
      "grad_norm": 0.24039429426193237,
      "learning_rate": 0.00017040040000000002,
      "loss": 0.3155,
      "step": 74000
    },
    {
      "epoch": 0.1482,
      "grad_norm": 0.20780912041664124,
      "learning_rate": 0.0001703604,
      "loss": 0.3131,
      "step": 74100
    },
    {
      "epoch": 0.1484,
      "grad_norm": 0.18914741277694702,
      "learning_rate": 0.0001703204,
      "loss": 0.3125,
      "step": 74200
    },
    {
      "epoch": 0.1486,
      "grad_norm": 0.18693341314792633,
      "learning_rate": 0.0001702804,
      "loss": 0.3185,
      "step": 74300
    },
    {
      "epoch": 0.1488,
      "grad_norm": 0.2793961465358734,
      "learning_rate": 0.00017024040000000002,
      "loss": 0.3209,
      "step": 74400
    },
    {
      "epoch": 0.149,
      "grad_norm": 0.3118375539779663,
      "learning_rate": 0.00017020040000000002,
      "loss": 0.3173,
      "step": 74500
    },
    {
      "epoch": 0.1492,
      "grad_norm": 0.23574410378932953,
      "learning_rate": 0.00017016039999999998,
      "loss": 0.3167,
      "step": 74600
    },
    {
      "epoch": 0.1494,
      "grad_norm": 0.33424112200737,
      "learning_rate": 0.0001701204,
      "loss": 0.3196,
      "step": 74700
    },
    {
      "epoch": 0.1496,
      "grad_norm": 0.24209052324295044,
      "learning_rate": 0.0001700804,
      "loss": 0.3201,
      "step": 74800
    },
    {
      "epoch": 0.1498,
      "grad_norm": 0.21331624686717987,
      "learning_rate": 0.00017004040000000002,
      "loss": 0.3163,
      "step": 74900
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.20186734199523926,
      "learning_rate": 0.0001700004,
      "loss": 0.3174,
      "step": 75000
    },
    {
      "epoch": 0.1502,
      "grad_norm": 0.17939358949661255,
      "learning_rate": 0.0001699604,
      "loss": 0.3185,
      "step": 75100
    },
    {
      "epoch": 0.1504,
      "grad_norm": 0.23708830773830414,
      "learning_rate": 0.0001699204,
      "loss": 0.3239,
      "step": 75200
    },
    {
      "epoch": 0.1506,
      "grad_norm": 0.23293326795101166,
      "learning_rate": 0.0001698804,
      "loss": 0.3211,
      "step": 75300
    },
    {
      "epoch": 0.1508,
      "grad_norm": 0.18440894782543182,
      "learning_rate": 0.00016984040000000001,
      "loss": 0.3118,
      "step": 75400
    },
    {
      "epoch": 0.151,
      "grad_norm": 0.2222815454006195,
      "learning_rate": 0.0001698004,
      "loss": 0.3108,
      "step": 75500
    },
    {
      "epoch": 0.1512,
      "grad_norm": 0.24183017015457153,
      "learning_rate": 0.0001697604,
      "loss": 0.3145,
      "step": 75600
    },
    {
      "epoch": 0.1514,
      "grad_norm": 0.2108313888311386,
      "learning_rate": 0.0001697204,
      "loss": 0.3128,
      "step": 75700
    },
    {
      "epoch": 0.1516,
      "grad_norm": 0.25066909193992615,
      "learning_rate": 0.00016968040000000002,
      "loss": 0.3223,
      "step": 75800
    },
    {
      "epoch": 0.1518,
      "grad_norm": 0.2033839225769043,
      "learning_rate": 0.0001696404,
      "loss": 0.3253,
      "step": 75900
    },
    {
      "epoch": 0.152,
      "grad_norm": 0.21795319020748138,
      "learning_rate": 0.0001696004,
      "loss": 0.3186,
      "step": 76000
    },
    {
      "epoch": 0.1522,
      "grad_norm": 0.2567816972732544,
      "learning_rate": 0.0001695604,
      "loss": 0.3184,
      "step": 76100
    },
    {
      "epoch": 0.1524,
      "grad_norm": 0.2667364478111267,
      "learning_rate": 0.0001695204,
      "loss": 0.3174,
      "step": 76200
    },
    {
      "epoch": 0.1526,
      "grad_norm": 0.24945732951164246,
      "learning_rate": 0.0001694804,
      "loss": 0.3166,
      "step": 76300
    },
    {
      "epoch": 0.1528,
      "grad_norm": 0.3093933165073395,
      "learning_rate": 0.0001694404,
      "loss": 0.3089,
      "step": 76400
    },
    {
      "epoch": 0.153,
      "grad_norm": 0.2426934689283371,
      "learning_rate": 0.00016940040000000003,
      "loss": 0.3128,
      "step": 76500
    },
    {
      "epoch": 0.1532,
      "grad_norm": 0.20812727510929108,
      "learning_rate": 0.00016936040000000002,
      "loss": 0.3093,
      "step": 76600
    },
    {
      "epoch": 0.1534,
      "grad_norm": 0.4342077672481537,
      "learning_rate": 0.0001693204,
      "loss": 0.3295,
      "step": 76700
    },
    {
      "epoch": 0.1536,
      "grad_norm": 0.2487853765487671,
      "learning_rate": 0.0001692804,
      "loss": 0.3202,
      "step": 76800
    },
    {
      "epoch": 0.1538,
      "grad_norm": 0.21004439890384674,
      "learning_rate": 0.0001692404,
      "loss": 0.3136,
      "step": 76900
    },
    {
      "epoch": 0.154,
      "grad_norm": 0.20426279306411743,
      "learning_rate": 0.00016920040000000002,
      "loss": 0.3154,
      "step": 77000
    },
    {
      "epoch": 0.1542,
      "grad_norm": 0.2324894517660141,
      "learning_rate": 0.00016916040000000001,
      "loss": 0.3124,
      "step": 77100
    },
    {
      "epoch": 0.1544,
      "grad_norm": 0.24114584922790527,
      "learning_rate": 0.0001691204,
      "loss": 0.3218,
      "step": 77200
    },
    {
      "epoch": 0.1546,
      "grad_norm": 0.21619604527950287,
      "learning_rate": 0.0001690804,
      "loss": 0.3081,
      "step": 77300
    },
    {
      "epoch": 0.1548,
      "grad_norm": 0.3266576826572418,
      "learning_rate": 0.00016904040000000002,
      "loss": 0.3154,
      "step": 77400
    },
    {
      "epoch": 0.155,
      "grad_norm": 0.19163742661476135,
      "learning_rate": 0.00016900040000000002,
      "loss": 0.3112,
      "step": 77500
    },
    {
      "epoch": 0.1552,
      "grad_norm": 0.24338412284851074,
      "learning_rate": 0.0001689604,
      "loss": 0.309,
      "step": 77600
    },
    {
      "epoch": 0.1554,
      "grad_norm": 0.2039482295513153,
      "learning_rate": 0.0001689204,
      "loss": 0.3196,
      "step": 77700
    },
    {
      "epoch": 0.1556,
      "grad_norm": 0.2261604517698288,
      "learning_rate": 0.0001688804,
      "loss": 0.3157,
      "step": 77800
    },
    {
      "epoch": 0.1558,
      "grad_norm": 0.23354065418243408,
      "learning_rate": 0.00016884040000000002,
      "loss": 0.3091,
      "step": 77900
    },
    {
      "epoch": 0.156,
      "grad_norm": 0.3552130162715912,
      "learning_rate": 0.0001688004,
      "loss": 0.3138,
      "step": 78000
    },
    {
      "epoch": 0.1562,
      "grad_norm": 0.3127024471759796,
      "learning_rate": 0.00016876040000000003,
      "loss": 0.3134,
      "step": 78100
    },
    {
      "epoch": 0.1564,
      "grad_norm": 0.2056431621313095,
      "learning_rate": 0.0001687204,
      "loss": 0.3082,
      "step": 78200
    },
    {
      "epoch": 0.1566,
      "grad_norm": 0.22769121825695038,
      "learning_rate": 0.0001686804,
      "loss": 0.318,
      "step": 78300
    },
    {
      "epoch": 0.1568,
      "grad_norm": 0.1975649893283844,
      "learning_rate": 0.0001686404,
      "loss": 0.3284,
      "step": 78400
    },
    {
      "epoch": 0.157,
      "grad_norm": 0.25407645106315613,
      "learning_rate": 0.0001686004,
      "loss": 0.3124,
      "step": 78500
    },
    {
      "epoch": 0.1572,
      "grad_norm": 0.18915730714797974,
      "learning_rate": 0.00016856040000000003,
      "loss": 0.3125,
      "step": 78600
    },
    {
      "epoch": 0.1574,
      "grad_norm": 0.20961833000183105,
      "learning_rate": 0.0001685204,
      "loss": 0.3113,
      "step": 78700
    },
    {
      "epoch": 0.1576,
      "grad_norm": 0.20440556108951569,
      "learning_rate": 0.00016848040000000001,
      "loss": 0.3133,
      "step": 78800
    },
    {
      "epoch": 0.1578,
      "grad_norm": 0.20298384130001068,
      "learning_rate": 0.0001684404,
      "loss": 0.3195,
      "step": 78900
    },
    {
      "epoch": 0.158,
      "grad_norm": 0.2182042896747589,
      "learning_rate": 0.0001684004,
      "loss": 0.3091,
      "step": 79000
    },
    {
      "epoch": 0.1582,
      "grad_norm": 0.22923989593982697,
      "learning_rate": 0.00016836040000000002,
      "loss": 0.3181,
      "step": 79100
    },
    {
      "epoch": 0.1584,
      "grad_norm": 0.3075358271598816,
      "learning_rate": 0.0001683204,
      "loss": 0.3143,
      "step": 79200
    },
    {
      "epoch": 0.1586,
      "grad_norm": 0.2053437978029251,
      "learning_rate": 0.0001682804,
      "loss": 0.3086,
      "step": 79300
    },
    {
      "epoch": 0.1588,
      "grad_norm": 0.7424286007881165,
      "learning_rate": 0.0001682404,
      "loss": 0.3103,
      "step": 79400
    },
    {
      "epoch": 0.159,
      "grad_norm": 0.6188491582870483,
      "learning_rate": 0.00016820040000000002,
      "loss": 0.3183,
      "step": 79500
    },
    {
      "epoch": 0.1592,
      "grad_norm": 0.276163786649704,
      "learning_rate": 0.00016816040000000002,
      "loss": 0.3191,
      "step": 79600
    },
    {
      "epoch": 0.1594,
      "grad_norm": 0.18703459203243256,
      "learning_rate": 0.0001681204,
      "loss": 0.3139,
      "step": 79700
    },
    {
      "epoch": 0.1596,
      "grad_norm": 0.17597968876361847,
      "learning_rate": 0.0001680804,
      "loss": 0.3108,
      "step": 79800
    },
    {
      "epoch": 0.1598,
      "grad_norm": 0.24276955425739288,
      "learning_rate": 0.0001680404,
      "loss": 0.3068,
      "step": 79900
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.21838606894016266,
      "learning_rate": 0.00016800040000000002,
      "loss": 0.3132,
      "step": 80000
    },
    {
      "epoch": 0.1602,
      "grad_norm": 0.2652193009853363,
      "learning_rate": 0.0001679604,
      "loss": 0.3107,
      "step": 80100
    },
    {
      "epoch": 0.1604,
      "grad_norm": 0.22899410128593445,
      "learning_rate": 0.0001679204,
      "loss": 0.3124,
      "step": 80200
    },
    {
      "epoch": 0.1606,
      "grad_norm": 0.19513286650180817,
      "learning_rate": 0.0001678804,
      "loss": 0.3125,
      "step": 80300
    },
    {
      "epoch": 0.1608,
      "grad_norm": 0.25317785143852234,
      "learning_rate": 0.00016784040000000002,
      "loss": 0.3075,
      "step": 80400
    },
    {
      "epoch": 0.161,
      "grad_norm": 0.21707969903945923,
      "learning_rate": 0.00016780040000000001,
      "loss": 0.3203,
      "step": 80500
    },
    {
      "epoch": 0.1612,
      "grad_norm": 0.22994515299797058,
      "learning_rate": 0.0001677604,
      "loss": 0.3136,
      "step": 80600
    },
    {
      "epoch": 0.1614,
      "grad_norm": 0.29428768157958984,
      "learning_rate": 0.0001677204,
      "loss": 0.3201,
      "step": 80700
    },
    {
      "epoch": 0.1616,
      "grad_norm": 0.24848338961601257,
      "learning_rate": 0.0001676804,
      "loss": 0.3169,
      "step": 80800
    },
    {
      "epoch": 0.1618,
      "grad_norm": 0.2289322018623352,
      "learning_rate": 0.00016764040000000002,
      "loss": 0.3055,
      "step": 80900
    },
    {
      "epoch": 0.162,
      "grad_norm": 0.20457616448402405,
      "learning_rate": 0.0001676004,
      "loss": 0.3115,
      "step": 81000
    },
    {
      "epoch": 0.1622,
      "grad_norm": 0.21032114326953888,
      "learning_rate": 0.00016756040000000003,
      "loss": 0.3116,
      "step": 81100
    },
    {
      "epoch": 0.1624,
      "grad_norm": 0.20674587786197662,
      "learning_rate": 0.0001675204,
      "loss": 0.3184,
      "step": 81200
    },
    {
      "epoch": 0.1626,
      "grad_norm": 0.4695895314216614,
      "learning_rate": 0.00016748040000000002,
      "loss": 0.3165,
      "step": 81300
    },
    {
      "epoch": 0.1628,
      "grad_norm": 0.26248136162757874,
      "learning_rate": 0.0001674404,
      "loss": 0.3092,
      "step": 81400
    },
    {
      "epoch": 0.163,
      "grad_norm": 0.20731079578399658,
      "learning_rate": 0.0001674004,
      "loss": 0.3141,
      "step": 81500
    },
    {
      "epoch": 0.1632,
      "grad_norm": 0.2757965922355652,
      "learning_rate": 0.00016736040000000002,
      "loss": 0.3124,
      "step": 81600
    },
    {
      "epoch": 0.1634,
      "grad_norm": 0.23029279708862305,
      "learning_rate": 0.0001673204,
      "loss": 0.3079,
      "step": 81700
    },
    {
      "epoch": 0.1636,
      "grad_norm": 0.22975629568099976,
      "learning_rate": 0.0001672804,
      "loss": 0.3157,
      "step": 81800
    },
    {
      "epoch": 0.1638,
      "grad_norm": 0.27681633830070496,
      "learning_rate": 0.0001672404,
      "loss": 0.31,
      "step": 81900
    },
    {
      "epoch": 0.164,
      "grad_norm": 0.20509104430675507,
      "learning_rate": 0.0001672004,
      "loss": 0.3141,
      "step": 82000
    },
    {
      "epoch": 0.1642,
      "grad_norm": 0.22510622441768646,
      "learning_rate": 0.00016716040000000002,
      "loss": 0.3046,
      "step": 82100
    },
    {
      "epoch": 0.1644,
      "grad_norm": 0.24418208003044128,
      "learning_rate": 0.0001671204,
      "loss": 0.324,
      "step": 82200
    },
    {
      "epoch": 0.1646,
      "grad_norm": 0.2196940779685974,
      "learning_rate": 0.0001670804,
      "loss": 0.3079,
      "step": 82300
    },
    {
      "epoch": 0.1648,
      "grad_norm": 0.19491960108280182,
      "learning_rate": 0.0001670404,
      "loss": 0.3114,
      "step": 82400
    },
    {
      "epoch": 0.165,
      "grad_norm": 0.18604259192943573,
      "learning_rate": 0.00016700040000000002,
      "loss": 0.3087,
      "step": 82500
    },
    {
      "epoch": 0.1652,
      "grad_norm": 0.22354719042778015,
      "learning_rate": 0.00016696040000000002,
      "loss": 0.3059,
      "step": 82600
    },
    {
      "epoch": 0.1654,
      "grad_norm": 0.19884774088859558,
      "learning_rate": 0.0001669204,
      "loss": 0.3094,
      "step": 82700
    },
    {
      "epoch": 0.1656,
      "grad_norm": 0.17430974543094635,
      "learning_rate": 0.0001668804,
      "loss": 0.3062,
      "step": 82800
    },
    {
      "epoch": 0.1658,
      "grad_norm": 0.32356834411621094,
      "learning_rate": 0.0001668404,
      "loss": 0.3127,
      "step": 82900
    },
    {
      "epoch": 0.166,
      "grad_norm": 0.19034580886363983,
      "learning_rate": 0.00016680040000000002,
      "loss": 0.308,
      "step": 83000
    },
    {
      "epoch": 0.1662,
      "grad_norm": 0.2106025516986847,
      "learning_rate": 0.0001667604,
      "loss": 0.3065,
      "step": 83100
    },
    {
      "epoch": 0.1664,
      "grad_norm": 0.22621573507785797,
      "learning_rate": 0.0001667204,
      "loss": 0.327,
      "step": 83200
    },
    {
      "epoch": 0.1666,
      "grad_norm": 0.19813230633735657,
      "learning_rate": 0.0001666804,
      "loss": 0.3081,
      "step": 83300
    },
    {
      "epoch": 0.1668,
      "grad_norm": 0.20063921809196472,
      "learning_rate": 0.00016664040000000002,
      "loss": 0.3069,
      "step": 83400
    },
    {
      "epoch": 0.167,
      "grad_norm": 0.1986655443906784,
      "learning_rate": 0.0001666004,
      "loss": 0.316,
      "step": 83500
    },
    {
      "epoch": 0.1672,
      "grad_norm": 0.22707831859588623,
      "learning_rate": 0.0001665604,
      "loss": 0.3103,
      "step": 83600
    },
    {
      "epoch": 0.1674,
      "grad_norm": 0.233692929148674,
      "learning_rate": 0.0001665204,
      "loss": 0.3196,
      "step": 83700
    },
    {
      "epoch": 0.1676,
      "grad_norm": 0.35872554779052734,
      "learning_rate": 0.0001664804,
      "loss": 0.3179,
      "step": 83800
    },
    {
      "epoch": 0.1678,
      "grad_norm": 0.3032999336719513,
      "learning_rate": 0.0001664404,
      "loss": 0.3123,
      "step": 83900
    },
    {
      "epoch": 0.168,
      "grad_norm": 0.27240657806396484,
      "learning_rate": 0.0001664004,
      "loss": 0.312,
      "step": 84000
    },
    {
      "epoch": 0.1682,
      "grad_norm": 0.20436140894889832,
      "learning_rate": 0.00016636040000000003,
      "loss": 0.3079,
      "step": 84100
    },
    {
      "epoch": 0.1684,
      "grad_norm": 0.26683861017227173,
      "learning_rate": 0.0001663204,
      "loss": 0.3107,
      "step": 84200
    },
    {
      "epoch": 0.1686,
      "grad_norm": 0.3275362253189087,
      "learning_rate": 0.00016628040000000001,
      "loss": 0.3066,
      "step": 84300
    },
    {
      "epoch": 0.1688,
      "grad_norm": 0.24865290522575378,
      "learning_rate": 0.0001662404,
      "loss": 0.3103,
      "step": 84400
    },
    {
      "epoch": 0.169,
      "grad_norm": 0.24508464336395264,
      "learning_rate": 0.0001662004,
      "loss": 0.3084,
      "step": 84500
    },
    {
      "epoch": 0.1692,
      "grad_norm": 0.19160589575767517,
      "learning_rate": 0.00016616040000000002,
      "loss": 0.3086,
      "step": 84600
    },
    {
      "epoch": 0.1694,
      "grad_norm": 0.21846391260623932,
      "learning_rate": 0.0001661204,
      "loss": 0.3038,
      "step": 84700
    },
    {
      "epoch": 0.1696,
      "grad_norm": 0.3486625850200653,
      "learning_rate": 0.0001660804,
      "loss": 0.3157,
      "step": 84800
    },
    {
      "epoch": 0.1698,
      "grad_norm": 0.281613826751709,
      "learning_rate": 0.0001660404,
      "loss": 0.3051,
      "step": 84900
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.2179608792066574,
      "learning_rate": 0.00016600040000000002,
      "loss": 0.3258,
      "step": 85000
    },
    {
      "epoch": 0.1702,
      "grad_norm": 0.24830220639705658,
      "learning_rate": 0.00016596040000000002,
      "loss": 0.312,
      "step": 85100
    },
    {
      "epoch": 0.1704,
      "grad_norm": 0.2119312733411789,
      "learning_rate": 0.0001659204,
      "loss": 0.3119,
      "step": 85200
    },
    {
      "epoch": 0.1706,
      "grad_norm": 0.34956997632980347,
      "learning_rate": 0.0001658804,
      "loss": 0.3004,
      "step": 85300
    },
    {
      "epoch": 0.1708,
      "grad_norm": 0.2335246205329895,
      "learning_rate": 0.0001658404,
      "loss": 0.3102,
      "step": 85400
    },
    {
      "epoch": 0.171,
      "grad_norm": 0.2509993612766266,
      "learning_rate": 0.00016580040000000002,
      "loss": 0.3118,
      "step": 85500
    },
    {
      "epoch": 0.1712,
      "grad_norm": 0.23735851049423218,
      "learning_rate": 0.0001657604,
      "loss": 0.3073,
      "step": 85600
    },
    {
      "epoch": 0.1714,
      "grad_norm": 0.1960500031709671,
      "learning_rate": 0.0001657204,
      "loss": 0.3061,
      "step": 85700
    },
    {
      "epoch": 0.1716,
      "grad_norm": 0.17898212373256683,
      "learning_rate": 0.0001656804,
      "loss": 0.3056,
      "step": 85800
    },
    {
      "epoch": 0.1718,
      "grad_norm": 0.207002654671669,
      "learning_rate": 0.0001656404,
      "loss": 0.3114,
      "step": 85900
    },
    {
      "epoch": 0.172,
      "grad_norm": 0.19042788445949554,
      "learning_rate": 0.00016560040000000001,
      "loss": 0.3109,
      "step": 86000
    },
    {
      "epoch": 0.1722,
      "grad_norm": 0.2865552604198456,
      "learning_rate": 0.0001655604,
      "loss": 0.3185,
      "step": 86100
    },
    {
      "epoch": 0.1724,
      "grad_norm": 0.19299517571926117,
      "learning_rate": 0.0001655204,
      "loss": 0.3329,
      "step": 86200
    },
    {
      "epoch": 0.1726,
      "grad_norm": 0.22495882213115692,
      "learning_rate": 0.0001654804,
      "loss": 0.3175,
      "step": 86300
    },
    {
      "epoch": 0.1728,
      "grad_norm": 0.23224885761737823,
      "learning_rate": 0.00016544040000000002,
      "loss": 0.3076,
      "step": 86400
    },
    {
      "epoch": 0.173,
      "grad_norm": 0.24318882822990417,
      "learning_rate": 0.0001654004,
      "loss": 0.3123,
      "step": 86500
    },
    {
      "epoch": 0.1732,
      "grad_norm": 0.24434731900691986,
      "learning_rate": 0.0001653604,
      "loss": 0.3092,
      "step": 86600
    },
    {
      "epoch": 0.1734,
      "grad_norm": 0.23535099625587463,
      "learning_rate": 0.0001653204,
      "loss": 0.3091,
      "step": 86700
    },
    {
      "epoch": 0.1736,
      "grad_norm": 0.3537319600582123,
      "learning_rate": 0.0001652804,
      "loss": 0.3187,
      "step": 86800
    },
    {
      "epoch": 0.1738,
      "grad_norm": 0.2087029367685318,
      "learning_rate": 0.0001652404,
      "loss": 0.3096,
      "step": 86900
    },
    {
      "epoch": 0.174,
      "grad_norm": 0.20568256080150604,
      "learning_rate": 0.0001652004,
      "loss": 0.3114,
      "step": 87000
    },
    {
      "epoch": 0.1742,
      "grad_norm": 0.19157211482524872,
      "learning_rate": 0.00016516040000000003,
      "loss": 0.3124,
      "step": 87100
    },
    {
      "epoch": 0.1744,
      "grad_norm": 0.20994393527507782,
      "learning_rate": 0.0001651204,
      "loss": 0.3047,
      "step": 87200
    },
    {
      "epoch": 0.1746,
      "grad_norm": 0.4066809117794037,
      "learning_rate": 0.0001650804,
      "loss": 0.3105,
      "step": 87300
    },
    {
      "epoch": 0.1748,
      "grad_norm": 0.21257740259170532,
      "learning_rate": 0.0001650404,
      "loss": 0.3084,
      "step": 87400
    },
    {
      "epoch": 0.175,
      "grad_norm": 0.21162879467010498,
      "learning_rate": 0.0001650004,
      "loss": 0.3105,
      "step": 87500
    },
    {
      "epoch": 0.1752,
      "grad_norm": 0.21106883883476257,
      "learning_rate": 0.00016496040000000002,
      "loss": 0.3098,
      "step": 87600
    },
    {
      "epoch": 0.1754,
      "grad_norm": 0.21054783463478088,
      "learning_rate": 0.0001649204,
      "loss": 0.3018,
      "step": 87700
    },
    {
      "epoch": 0.1756,
      "grad_norm": 0.1708284318447113,
      "learning_rate": 0.0001648804,
      "loss": 0.3049,
      "step": 87800
    },
    {
      "epoch": 0.1758,
      "grad_norm": 0.20066872239112854,
      "learning_rate": 0.0001648404,
      "loss": 0.3055,
      "step": 87900
    },
    {
      "epoch": 0.176,
      "grad_norm": 0.3257821798324585,
      "learning_rate": 0.00016480040000000002,
      "loss": 0.3102,
      "step": 88000
    },
    {
      "epoch": 0.1762,
      "grad_norm": 0.1875178962945938,
      "learning_rate": 0.00016476040000000002,
      "loss": 0.311,
      "step": 88100
    },
    {
      "epoch": 0.1764,
      "grad_norm": 0.3387661874294281,
      "learning_rate": 0.0001647204,
      "loss": 0.3101,
      "step": 88200
    },
    {
      "epoch": 0.1766,
      "grad_norm": 0.3336031436920166,
      "learning_rate": 0.0001646804,
      "loss": 0.3154,
      "step": 88300
    },
    {
      "epoch": 0.1768,
      "grad_norm": 0.21378177404403687,
      "learning_rate": 0.0001646404,
      "loss": 0.3195,
      "step": 88400
    },
    {
      "epoch": 0.177,
      "grad_norm": 0.2274891436100006,
      "learning_rate": 0.00016460040000000002,
      "loss": 0.3096,
      "step": 88500
    },
    {
      "epoch": 0.1772,
      "grad_norm": 0.23640379309654236,
      "learning_rate": 0.0001645604,
      "loss": 0.3089,
      "step": 88600
    },
    {
      "epoch": 0.1774,
      "grad_norm": 0.19889909029006958,
      "learning_rate": 0.00016452040000000003,
      "loss": 0.3058,
      "step": 88700
    },
    {
      "epoch": 0.1776,
      "grad_norm": 0.20005521178245544,
      "learning_rate": 0.0001644804,
      "loss": 0.3059,
      "step": 88800
    },
    {
      "epoch": 0.1778,
      "grad_norm": 0.18955092132091522,
      "learning_rate": 0.00016444040000000002,
      "loss": 0.3026,
      "step": 88900
    },
    {
      "epoch": 0.178,
      "grad_norm": 0.23005805909633636,
      "learning_rate": 0.0001644004,
      "loss": 0.3067,
      "step": 89000
    },
    {
      "epoch": 0.1782,
      "grad_norm": 0.1776682585477829,
      "learning_rate": 0.0001643604,
      "loss": 0.3029,
      "step": 89100
    },
    {
      "epoch": 0.1784,
      "grad_norm": 0.29056572914123535,
      "learning_rate": 0.00016432040000000003,
      "loss": 0.3025,
      "step": 89200
    },
    {
      "epoch": 0.1786,
      "grad_norm": 0.2611645460128784,
      "learning_rate": 0.0001642804,
      "loss": 0.3022,
      "step": 89300
    },
    {
      "epoch": 0.1788,
      "grad_norm": 0.22479821741580963,
      "learning_rate": 0.00016424040000000001,
      "loss": 0.3102,
      "step": 89400
    },
    {
      "epoch": 0.179,
      "grad_norm": 0.3284640908241272,
      "learning_rate": 0.0001642004,
      "loss": 0.3078,
      "step": 89500
    },
    {
      "epoch": 0.1792,
      "grad_norm": 0.19706085324287415,
      "learning_rate": 0.0001641604,
      "loss": 0.3005,
      "step": 89600
    },
    {
      "epoch": 0.1794,
      "grad_norm": 0.1871166229248047,
      "learning_rate": 0.00016412040000000002,
      "loss": 0.3067,
      "step": 89700
    },
    {
      "epoch": 0.1796,
      "grad_norm": 0.18414820730686188,
      "learning_rate": 0.0001640804,
      "loss": 0.3065,
      "step": 89800
    },
    {
      "epoch": 0.1798,
      "grad_norm": 0.22224344313144684,
      "learning_rate": 0.0001640404,
      "loss": 0.3023,
      "step": 89900
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.1839105784893036,
      "learning_rate": 0.0001640004,
      "loss": 0.3044,
      "step": 90000
    },
    {
      "epoch": 0.1802,
      "grad_norm": 0.3029592037200928,
      "learning_rate": 0.00016396040000000002,
      "loss": 0.3002,
      "step": 90100
    },
    {
      "epoch": 0.1804,
      "grad_norm": 0.21777479350566864,
      "learning_rate": 0.00016392040000000002,
      "loss": 0.3055,
      "step": 90200
    },
    {
      "epoch": 0.1806,
      "grad_norm": 0.1881372481584549,
      "learning_rate": 0.0001638804,
      "loss": 0.3015,
      "step": 90300
    },
    {
      "epoch": 0.1808,
      "grad_norm": 0.20605388283729553,
      "learning_rate": 0.0001638404,
      "loss": 0.3086,
      "step": 90400
    },
    {
      "epoch": 0.181,
      "grad_norm": 0.21422967314720154,
      "learning_rate": 0.0001638004,
      "loss": 0.3107,
      "step": 90500
    },
    {
      "epoch": 0.1812,
      "grad_norm": 0.18776142597198486,
      "learning_rate": 0.00016376040000000002,
      "loss": 0.297,
      "step": 90600
    },
    {
      "epoch": 0.1814,
      "grad_norm": 0.22885388135910034,
      "learning_rate": 0.0001637204,
      "loss": 0.2897,
      "step": 90700
    },
    {
      "epoch": 0.1816,
      "grad_norm": 0.16700296103954315,
      "learning_rate": 0.0001636804,
      "loss": 0.298,
      "step": 90800
    },
    {
      "epoch": 0.1818,
      "grad_norm": 0.2172263115644455,
      "learning_rate": 0.0001636404,
      "loss": 0.2995,
      "step": 90900
    },
    {
      "epoch": 0.182,
      "grad_norm": 0.25213050842285156,
      "learning_rate": 0.00016360040000000002,
      "loss": 0.3077,
      "step": 91000
    },
    {
      "epoch": 0.1822,
      "grad_norm": 0.22334252297878265,
      "learning_rate": 0.00016356040000000001,
      "loss": 0.3077,
      "step": 91100
    },
    {
      "epoch": 0.1824,
      "grad_norm": 0.25537583231925964,
      "learning_rate": 0.0001635204,
      "loss": 0.3074,
      "step": 91200
    },
    {
      "epoch": 0.1826,
      "grad_norm": 0.24679423868656158,
      "learning_rate": 0.0001634804,
      "loss": 0.3139,
      "step": 91300
    },
    {
      "epoch": 0.1828,
      "grad_norm": 0.26473841071128845,
      "learning_rate": 0.0001634404,
      "loss": 0.3155,
      "step": 91400
    },
    {
      "epoch": 0.183,
      "grad_norm": 0.23032037913799286,
      "learning_rate": 0.00016340040000000002,
      "loss": 0.3191,
      "step": 91500
    },
    {
      "epoch": 0.1832,
      "grad_norm": 0.4667571485042572,
      "learning_rate": 0.0001633604,
      "loss": 0.3006,
      "step": 91600
    },
    {
      "epoch": 0.1834,
      "grad_norm": 0.1938803791999817,
      "learning_rate": 0.00016332040000000003,
      "loss": 0.3034,
      "step": 91700
    },
    {
      "epoch": 0.1836,
      "grad_norm": 0.1822306215763092,
      "learning_rate": 0.0001632804,
      "loss": 0.306,
      "step": 91800
    },
    {
      "epoch": 0.1838,
      "grad_norm": 0.29107385873794556,
      "learning_rate": 0.00016324040000000002,
      "loss": 0.304,
      "step": 91900
    },
    {
      "epoch": 0.184,
      "grad_norm": 0.3339613676071167,
      "learning_rate": 0.0001632004,
      "loss": 0.2963,
      "step": 92000
    },
    {
      "epoch": 0.1842,
      "grad_norm": 0.2274635285139084,
      "learning_rate": 0.0001631604,
      "loss": 0.2997,
      "step": 92100
    },
    {
      "epoch": 0.1844,
      "grad_norm": 0.19457891583442688,
      "learning_rate": 0.00016312040000000002,
      "loss": 0.306,
      "step": 92200
    },
    {
      "epoch": 0.1846,
      "grad_norm": 0.2524142265319824,
      "learning_rate": 0.0001630804,
      "loss": 0.3096,
      "step": 92300
    },
    {
      "epoch": 0.1848,
      "grad_norm": 0.25092560052871704,
      "learning_rate": 0.0001630404,
      "loss": 0.2999,
      "step": 92400
    },
    {
      "epoch": 0.185,
      "grad_norm": 0.323590487241745,
      "learning_rate": 0.0001630004,
      "loss": 0.2996,
      "step": 92500
    },
    {
      "epoch": 0.1852,
      "grad_norm": 0.4235788881778717,
      "learning_rate": 0.00016296040000000003,
      "loss": 0.3031,
      "step": 92600
    },
    {
      "epoch": 0.1854,
      "grad_norm": 0.19289268553256989,
      "learning_rate": 0.00016292040000000002,
      "loss": 0.3026,
      "step": 92700
    },
    {
      "epoch": 0.1856,
      "grad_norm": 0.1985141932964325,
      "learning_rate": 0.00016288039999999999,
      "loss": 0.302,
      "step": 92800
    },
    {
      "epoch": 0.1858,
      "grad_norm": 0.3089956045150757,
      "learning_rate": 0.0001628404,
      "loss": 0.3002,
      "step": 92900
    },
    {
      "epoch": 0.186,
      "grad_norm": 0.2129218727350235,
      "learning_rate": 0.0001628004,
      "loss": 0.3036,
      "step": 93000
    },
    {
      "epoch": 0.1862,
      "grad_norm": 0.3660809397697449,
      "learning_rate": 0.00016276040000000002,
      "loss": 0.3017,
      "step": 93100
    },
    {
      "epoch": 0.1864,
      "grad_norm": 0.22182439267635345,
      "learning_rate": 0.00016272040000000002,
      "loss": 0.3012,
      "step": 93200
    },
    {
      "epoch": 0.1866,
      "grad_norm": 0.22655236721038818,
      "learning_rate": 0.0001626804,
      "loss": 0.3054,
      "step": 93300
    },
    {
      "epoch": 0.1868,
      "grad_norm": 0.19816626608371735,
      "learning_rate": 0.0001626404,
      "loss": 0.2978,
      "step": 93400
    },
    {
      "epoch": 0.187,
      "grad_norm": 0.19407741725444794,
      "learning_rate": 0.0001626004,
      "loss": 0.3013,
      "step": 93500
    },
    {
      "epoch": 0.1872,
      "grad_norm": 0.15464486181735992,
      "learning_rate": 0.00016256040000000002,
      "loss": 0.3073,
      "step": 93600
    },
    {
      "epoch": 0.1874,
      "grad_norm": 0.34153929352760315,
      "learning_rate": 0.0001625204,
      "loss": 0.3023,
      "step": 93700
    },
    {
      "epoch": 0.1876,
      "grad_norm": 0.35526177287101746,
      "learning_rate": 0.0001624804,
      "loss": 0.311,
      "step": 93800
    },
    {
      "epoch": 0.1878,
      "grad_norm": 0.2251167744398117,
      "learning_rate": 0.0001624404,
      "loss": 0.3127,
      "step": 93900
    },
    {
      "epoch": 0.188,
      "grad_norm": 0.25468969345092773,
      "learning_rate": 0.00016240040000000002,
      "loss": 0.3004,
      "step": 94000
    },
    {
      "epoch": 0.1882,
      "grad_norm": 0.17790032923221588,
      "learning_rate": 0.0001623604,
      "loss": 0.3096,
      "step": 94100
    },
    {
      "epoch": 0.1884,
      "grad_norm": 0.23429210484027863,
      "learning_rate": 0.0001623204,
      "loss": 0.3023,
      "step": 94200
    },
    {
      "epoch": 0.1886,
      "grad_norm": 0.3114053010940552,
      "learning_rate": 0.0001622804,
      "loss": 0.3043,
      "step": 94300
    },
    {
      "epoch": 0.1888,
      "grad_norm": 0.19942690432071686,
      "learning_rate": 0.0001622404,
      "loss": 0.2985,
      "step": 94400
    },
    {
      "epoch": 0.189,
      "grad_norm": 0.22419995069503784,
      "learning_rate": 0.0001622004,
      "loss": 0.3014,
      "step": 94500
    },
    {
      "epoch": 0.1892,
      "grad_norm": 0.19408832490444183,
      "learning_rate": 0.0001621604,
      "loss": 0.3031,
      "step": 94600
    },
    {
      "epoch": 0.1894,
      "grad_norm": 0.24643263220787048,
      "learning_rate": 0.00016212040000000003,
      "loss": 0.3042,
      "step": 94700
    },
    {
      "epoch": 0.1896,
      "grad_norm": 0.22279608249664307,
      "learning_rate": 0.0001620804,
      "loss": 0.3022,
      "step": 94800
    },
    {
      "epoch": 0.1898,
      "grad_norm": 0.340925931930542,
      "learning_rate": 0.00016204040000000001,
      "loss": 0.2986,
      "step": 94900
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.1957314908504486,
      "learning_rate": 0.0001620004,
      "loss": 0.3014,
      "step": 95000
    },
    {
      "epoch": 0.1902,
      "grad_norm": 0.23643195629119873,
      "learning_rate": 0.0001619604,
      "loss": 0.3103,
      "step": 95100
    },
    {
      "epoch": 0.1904,
      "grad_norm": 0.2049199789762497,
      "learning_rate": 0.00016192040000000002,
      "loss": 0.3039,
      "step": 95200
    },
    {
      "epoch": 0.1906,
      "grad_norm": 0.19225084781646729,
      "learning_rate": 0.0001618804,
      "loss": 0.2963,
      "step": 95300
    },
    {
      "epoch": 0.1908,
      "grad_norm": 0.2920241355895996,
      "learning_rate": 0.0001618404,
      "loss": 0.2996,
      "step": 95400
    },
    {
      "epoch": 0.191,
      "grad_norm": 0.21930663287639618,
      "learning_rate": 0.0001618004,
      "loss": 0.3014,
      "step": 95500
    },
    {
      "epoch": 0.1912,
      "grad_norm": 0.2423180788755417,
      "learning_rate": 0.00016176040000000002,
      "loss": 0.3032,
      "step": 95600
    },
    {
      "epoch": 0.1914,
      "grad_norm": 0.37270796298980713,
      "learning_rate": 0.00016172040000000002,
      "loss": 0.3054,
      "step": 95700
    },
    {
      "epoch": 0.1916,
      "grad_norm": 0.26363104581832886,
      "learning_rate": 0.0001616804,
      "loss": 0.3049,
      "step": 95800
    },
    {
      "epoch": 0.1918,
      "grad_norm": 0.2359706610441208,
      "learning_rate": 0.0001616404,
      "loss": 0.3016,
      "step": 95900
    },
    {
      "epoch": 0.192,
      "grad_norm": 0.2633230984210968,
      "learning_rate": 0.0001616004,
      "loss": 0.3027,
      "step": 96000
    },
    {
      "epoch": 0.1922,
      "grad_norm": 0.16779562830924988,
      "learning_rate": 0.00016156040000000002,
      "loss": 0.3006,
      "step": 96100
    },
    {
      "epoch": 0.1924,
      "grad_norm": 0.17921751737594604,
      "learning_rate": 0.0001615204,
      "loss": 0.2997,
      "step": 96200
    },
    {
      "epoch": 0.1926,
      "grad_norm": 0.1813753992319107,
      "learning_rate": 0.0001614804,
      "loss": 0.2989,
      "step": 96300
    },
    {
      "epoch": 0.1928,
      "grad_norm": 0.2046331912279129,
      "learning_rate": 0.0001614404,
      "loss": 0.2992,
      "step": 96400
    },
    {
      "epoch": 0.193,
      "grad_norm": 0.1848832368850708,
      "learning_rate": 0.0001614004,
      "loss": 0.301,
      "step": 96500
    },
    {
      "epoch": 0.1932,
      "grad_norm": 0.27407926321029663,
      "learning_rate": 0.00016136040000000001,
      "loss": 0.3084,
      "step": 96600
    },
    {
      "epoch": 0.1934,
      "grad_norm": 0.17408868670463562,
      "learning_rate": 0.0001613204,
      "loss": 0.2984,
      "step": 96700
    },
    {
      "epoch": 0.1936,
      "grad_norm": 0.3170103132724762,
      "learning_rate": 0.0001612804,
      "loss": 0.3098,
      "step": 96800
    },
    {
      "epoch": 0.1938,
      "grad_norm": 0.1906721144914627,
      "learning_rate": 0.0001612404,
      "loss": 0.3002,
      "step": 96900
    },
    {
      "epoch": 0.194,
      "grad_norm": 0.22694748640060425,
      "learning_rate": 0.00016120040000000002,
      "loss": 0.3056,
      "step": 97000
    },
    {
      "epoch": 0.1942,
      "grad_norm": 0.2381366342306137,
      "learning_rate": 0.0001611604,
      "loss": 0.3112,
      "step": 97100
    },
    {
      "epoch": 0.1944,
      "grad_norm": 0.2619655132293701,
      "learning_rate": 0.0001611204,
      "loss": 0.3044,
      "step": 97200
    },
    {
      "epoch": 0.1946,
      "grad_norm": 0.22410011291503906,
      "learning_rate": 0.0001610804,
      "loss": 0.2976,
      "step": 97300
    },
    {
      "epoch": 0.1948,
      "grad_norm": 0.17326219379901886,
      "learning_rate": 0.0001610404,
      "loss": 0.2977,
      "step": 97400
    },
    {
      "epoch": 0.195,
      "grad_norm": 0.16517135500907898,
      "learning_rate": 0.0001610004,
      "loss": 0.3074,
      "step": 97500
    },
    {
      "epoch": 0.1952,
      "grad_norm": 0.2561172544956207,
      "learning_rate": 0.0001609604,
      "loss": 0.3074,
      "step": 97600
    },
    {
      "epoch": 0.1954,
      "grad_norm": 0.19643421471118927,
      "learning_rate": 0.00016092040000000003,
      "loss": 0.3031,
      "step": 97700
    },
    {
      "epoch": 0.1956,
      "grad_norm": 0.26884031295776367,
      "learning_rate": 0.0001608804,
      "loss": 0.298,
      "step": 97800
    },
    {
      "epoch": 0.1958,
      "grad_norm": 0.1689986288547516,
      "learning_rate": 0.0001608404,
      "loss": 0.303,
      "step": 97900
    },
    {
      "epoch": 0.196,
      "grad_norm": 0.22177661955356598,
      "learning_rate": 0.0001608004,
      "loss": 0.3113,
      "step": 98000
    },
    {
      "epoch": 0.1962,
      "grad_norm": 0.23559048771858215,
      "learning_rate": 0.0001607604,
      "loss": 0.3019,
      "step": 98100
    },
    {
      "epoch": 0.1964,
      "grad_norm": 0.2539811432361603,
      "learning_rate": 0.00016072040000000002,
      "loss": 0.3017,
      "step": 98200
    },
    {
      "epoch": 0.1966,
      "grad_norm": 0.2040734440088272,
      "learning_rate": 0.0001606804,
      "loss": 0.294,
      "step": 98300
    },
    {
      "epoch": 0.1968,
      "grad_norm": 0.18892860412597656,
      "learning_rate": 0.0001606404,
      "loss": 0.3082,
      "step": 98400
    },
    {
      "epoch": 0.197,
      "grad_norm": 0.23716984689235687,
      "learning_rate": 0.0001606004,
      "loss": 0.3005,
      "step": 98500
    },
    {
      "epoch": 0.1972,
      "grad_norm": 0.28783920407295227,
      "learning_rate": 0.00016056040000000002,
      "loss": 0.303,
      "step": 98600
    },
    {
      "epoch": 0.1974,
      "grad_norm": 0.31491243839263916,
      "learning_rate": 0.00016052040000000002,
      "loss": 0.2989,
      "step": 98700
    },
    {
      "epoch": 0.1976,
      "grad_norm": 0.2184073030948639,
      "learning_rate": 0.0001604804,
      "loss": 0.2992,
      "step": 98800
    },
    {
      "epoch": 0.1978,
      "grad_norm": 0.19414541125297546,
      "learning_rate": 0.0001604404,
      "loss": 0.3067,
      "step": 98900
    },
    {
      "epoch": 0.198,
      "grad_norm": 0.18975822627544403,
      "learning_rate": 0.0001604004,
      "loss": 0.2994,
      "step": 99000
    },
    {
      "epoch": 0.1982,
      "grad_norm": 0.24121256172657013,
      "learning_rate": 0.00016036040000000002,
      "loss": 0.2967,
      "step": 99100
    },
    {
      "epoch": 0.1984,
      "grad_norm": 0.19773977994918823,
      "learning_rate": 0.0001603204,
      "loss": 0.2955,
      "step": 99200
    },
    {
      "epoch": 0.1986,
      "grad_norm": 0.22694402933120728,
      "learning_rate": 0.0001602804,
      "loss": 0.2977,
      "step": 99300
    },
    {
      "epoch": 0.1988,
      "grad_norm": 0.3497297465801239,
      "learning_rate": 0.0001602404,
      "loss": 0.3044,
      "step": 99400
    },
    {
      "epoch": 0.199,
      "grad_norm": 0.18374694883823395,
      "learning_rate": 0.00016020040000000002,
      "loss": 0.2965,
      "step": 99500
    },
    {
      "epoch": 0.1992,
      "grad_norm": 0.33236002922058105,
      "learning_rate": 0.0001601604,
      "loss": 0.3007,
      "step": 99600
    },
    {
      "epoch": 0.1994,
      "grad_norm": 0.17745955288410187,
      "learning_rate": 0.0001601204,
      "loss": 0.2924,
      "step": 99700
    },
    {
      "epoch": 0.1996,
      "grad_norm": 0.3132670223712921,
      "learning_rate": 0.0001600804,
      "loss": 0.3021,
      "step": 99800
    },
    {
      "epoch": 0.1998,
      "grad_norm": 0.2375771552324295,
      "learning_rate": 0.0001600404,
      "loss": 0.3016,
      "step": 99900
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.31381866335868835,
      "learning_rate": 0.00016000040000000001,
      "loss": 0.2965,
      "step": 100000
    },
    {
      "epoch": 0.0002,
      "grad_norm": 0.24663914740085602,
      "learning_rate": 0.0001599604,
      "loss": 0.3381,
      "step": 100100
    },
    {
      "epoch": 0.0004,
      "grad_norm": 0.6509382128715515,
      "learning_rate": 0.0001599204,
      "loss": 0.3687,
      "step": 100200
    },
    {
      "epoch": 0.0006,
      "grad_norm": 0.8887121081352234,
      "learning_rate": 0.0001598804,
      "loss": 0.3573,
      "step": 100300
    },
    {
      "epoch": 0.0008,
      "grad_norm": 0.366519033908844,
      "learning_rate": 0.0001598404,
      "loss": 0.3624,
      "step": 100400
    },
    {
      "epoch": 0.001,
      "grad_norm": 0.2282172590494156,
      "learning_rate": 0.0001598004,
      "loss": 0.3336,
      "step": 100500
    },
    {
      "epoch": 0.0012,
      "grad_norm": 0.46760937571525574,
      "learning_rate": 0.0001597604,
      "loss": 0.3279,
      "step": 100600
    },
    {
      "epoch": 0.0014,
      "grad_norm": 0.38649675250053406,
      "learning_rate": 0.00015972040000000002,
      "loss": 0.3385,
      "step": 100700
    },
    {
      "epoch": 0.0016,
      "grad_norm": 0.4251730740070343,
      "learning_rate": 0.0001596804,
      "loss": 0.335,
      "step": 100800
    },
    {
      "epoch": 0.0018,
      "grad_norm": 0.3479570150375366,
      "learning_rate": 0.0001596404,
      "loss": 0.3238,
      "step": 100900
    },
    {
      "epoch": 0.002,
      "grad_norm": 0.3096480667591095,
      "learning_rate": 0.0001596004,
      "loss": 0.3159,
      "step": 101000
    },
    {
      "epoch": 0.0022,
      "grad_norm": 0.36655187606811523,
      "learning_rate": 0.0001595604,
      "loss": 0.3457,
      "step": 101100
    },
    {
      "epoch": 0.0024,
      "grad_norm": 0.6139711141586304,
      "learning_rate": 0.00015952040000000002,
      "loss": 0.3249,
      "step": 101200
    },
    {
      "epoch": 0.0026,
      "grad_norm": 0.6844519376754761,
      "learning_rate": 0.0001594804,
      "loss": 0.3306,
      "step": 101300
    },
    {
      "epoch": 0.0028,
      "grad_norm": 0.16951784491539001,
      "learning_rate": 0.0001594404,
      "loss": 0.3109,
      "step": 101400
    },
    {
      "epoch": 0.003,
      "grad_norm": 0.4543248414993286,
      "learning_rate": 0.0001594004,
      "loss": 0.3241,
      "step": 101500
    },
    {
      "epoch": 0.0032,
      "grad_norm": 0.6316892504692078,
      "learning_rate": 0.00015936040000000002,
      "loss": 0.3479,
      "step": 101600
    },
    {
      "epoch": 0.0034,
      "grad_norm": 0.8181749582290649,
      "learning_rate": 0.00015932040000000001,
      "loss": 0.3215,
      "step": 101700
    },
    {
      "epoch": 0.0036,
      "grad_norm": 0.33392763137817383,
      "learning_rate": 0.0001592804,
      "loss": 0.3253,
      "step": 101800
    },
    {
      "epoch": 0.0038,
      "grad_norm": 0.27364447712898254,
      "learning_rate": 0.0001592404,
      "loss": 0.3252,
      "step": 101900
    },
    {
      "epoch": 0.004,
      "grad_norm": 0.443889319896698,
      "learning_rate": 0.0001592004,
      "loss": 0.3214,
      "step": 102000
    },
    {
      "epoch": 0.0042,
      "grad_norm": 0.23544713854789734,
      "learning_rate": 0.00015916040000000002,
      "loss": 0.3227,
      "step": 102100
    },
    {
      "epoch": 0.0044,
      "grad_norm": 0.24548670649528503,
      "learning_rate": 0.0001591204,
      "loss": 0.3491,
      "step": 102200
    },
    {
      "epoch": 0.0046,
      "grad_norm": 0.4170477092266083,
      "learning_rate": 0.00015908040000000003,
      "loss": 0.3318,
      "step": 102300
    },
    {
      "epoch": 0.0048,
      "grad_norm": 0.24510768055915833,
      "learning_rate": 0.0001590404,
      "loss": 0.3116,
      "step": 102400
    },
    {
      "epoch": 0.005,
      "grad_norm": 0.2921444773674011,
      "learning_rate": 0.00015900040000000002,
      "loss": 0.3049,
      "step": 102500
    },
    {
      "epoch": 0.0052,
      "grad_norm": 0.2453673630952835,
      "learning_rate": 0.0001589604,
      "loss": 0.3052,
      "step": 102600
    },
    {
      "epoch": 0.0054,
      "grad_norm": 0.3682481348514557,
      "learning_rate": 0.0001589204,
      "loss": 0.3123,
      "step": 102700
    },
    {
      "epoch": 0.0056,
      "grad_norm": 0.6065927743911743,
      "learning_rate": 0.00015888040000000002,
      "loss": 0.3278,
      "step": 102800
    },
    {
      "epoch": 0.0058,
      "grad_norm": 0.27878880500793457,
      "learning_rate": 0.0001588404,
      "loss": 0.3095,
      "step": 102900
    },
    {
      "epoch": 0.006,
      "grad_norm": 0.3071708083152771,
      "learning_rate": 0.0001588004,
      "loss": 0.3729,
      "step": 103000
    },
    {
      "epoch": 0.0062,
      "grad_norm": 0.27451980113983154,
      "learning_rate": 0.0001587604,
      "loss": 0.3234,
      "step": 103100
    },
    {
      "epoch": 0.0064,
      "grad_norm": 0.1992095708847046,
      "learning_rate": 0.00015872040000000003,
      "loss": 0.3142,
      "step": 103200
    },
    {
      "epoch": 0.0066,
      "grad_norm": 1.1765793561935425,
      "learning_rate": 0.00015868040000000002,
      "loss": 0.3365,
      "step": 103300
    },
    {
      "epoch": 0.0068,
      "grad_norm": 0.22701187431812286,
      "learning_rate": 0.0001586404,
      "loss": 0.3165,
      "step": 103400
    },
    {
      "epoch": 0.007,
      "grad_norm": 0.30635154247283936,
      "learning_rate": 0.0001586004,
      "loss": 0.3271,
      "step": 103500
    },
    {
      "epoch": 0.0072,
      "grad_norm": 0.3161128759384155,
      "learning_rate": 0.0001585604,
      "loss": 0.3246,
      "step": 103600
    },
    {
      "epoch": 0.0074,
      "grad_norm": 0.2176142781972885,
      "learning_rate": 0.00015852040000000002,
      "loss": 0.3342,
      "step": 103700
    },
    {
      "epoch": 0.0076,
      "grad_norm": 0.15987540781497955,
      "learning_rate": 0.00015848040000000001,
      "loss": 0.3345,
      "step": 103800
    },
    {
      "epoch": 0.0078,
      "grad_norm": 0.2533618211746216,
      "learning_rate": 0.0001584404,
      "loss": 0.3176,
      "step": 103900
    },
    {
      "epoch": 0.008,
      "grad_norm": 0.28857430815696716,
      "learning_rate": 0.0001584004,
      "loss": 0.3287,
      "step": 104000
    },
    {
      "epoch": 0.0082,
      "grad_norm": 0.4161205291748047,
      "learning_rate": 0.0001583604,
      "loss": 0.3342,
      "step": 104100
    },
    {
      "epoch": 0.0084,
      "grad_norm": 0.851681649684906,
      "learning_rate": 0.00015832040000000002,
      "loss": 0.318,
      "step": 104200
    },
    {
      "epoch": 0.0086,
      "grad_norm": 0.21251384913921356,
      "learning_rate": 0.0001582804,
      "loss": 0.3125,
      "step": 104300
    },
    {
      "epoch": 0.0088,
      "grad_norm": 0.19260574877262115,
      "learning_rate": 0.0001582404,
      "loss": 0.3137,
      "step": 104400
    },
    {
      "epoch": 0.009,
      "grad_norm": 0.3808644413948059,
      "learning_rate": 0.0001582004,
      "loss": 0.33,
      "step": 104500
    },
    {
      "epoch": 0.0092,
      "grad_norm": 0.23479130864143372,
      "learning_rate": 0.00015816040000000002,
      "loss": 0.3368,
      "step": 104600
    },
    {
      "epoch": 0.0094,
      "grad_norm": 0.3280342221260071,
      "learning_rate": 0.0001581204,
      "loss": 0.3242,
      "step": 104700
    },
    {
      "epoch": 0.0096,
      "grad_norm": 0.3940763473510742,
      "learning_rate": 0.0001580804,
      "loss": 0.3362,
      "step": 104800
    },
    {
      "epoch": 0.0098,
      "grad_norm": 0.23537495732307434,
      "learning_rate": 0.0001580404,
      "loss": 0.3342,
      "step": 104900
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.3302271366119385,
      "learning_rate": 0.0001580004,
      "loss": 0.3262,
      "step": 105000
    },
    {
      "epoch": 0.0102,
      "grad_norm": 0.3009653389453888,
      "learning_rate": 0.0001579604,
      "loss": 0.3353,
      "step": 105100
    },
    {
      "epoch": 0.0104,
      "grad_norm": 0.28532806038856506,
      "learning_rate": 0.0001579204,
      "loss": 0.3156,
      "step": 105200
    },
    {
      "epoch": 0.0106,
      "grad_norm": 0.24950698018074036,
      "learning_rate": 0.00015788040000000003,
      "loss": 0.3294,
      "step": 105300
    },
    {
      "epoch": 0.0108,
      "grad_norm": 0.26212894916534424,
      "learning_rate": 0.0001578404,
      "loss": 0.3329,
      "step": 105400
    },
    {
      "epoch": 0.011,
      "grad_norm": 0.22611194849014282,
      "learning_rate": 0.00015780040000000001,
      "loss": 0.3311,
      "step": 105500
    },
    {
      "epoch": 0.0112,
      "grad_norm": 0.2197837382555008,
      "learning_rate": 0.0001577604,
      "loss": 0.327,
      "step": 105600
    },
    {
      "epoch": 0.0114,
      "grad_norm": 0.31932878494262695,
      "learning_rate": 0.0001577204,
      "loss": 0.316,
      "step": 105700
    },
    {
      "epoch": 0.0116,
      "grad_norm": 0.2394053041934967,
      "learning_rate": 0.00015768040000000002,
      "loss": 0.331,
      "step": 105800
    },
    {
      "epoch": 0.0118,
      "grad_norm": 0.20572052896022797,
      "learning_rate": 0.0001576404,
      "loss": 0.3324,
      "step": 105900
    },
    {
      "epoch": 0.012,
      "grad_norm": 0.19541539251804352,
      "learning_rate": 0.0001576004,
      "loss": 0.3201,
      "step": 106000
    },
    {
      "epoch": 0.0122,
      "grad_norm": 0.2593480348587036,
      "learning_rate": 0.0001575604,
      "loss": 0.3545,
      "step": 106100
    },
    {
      "epoch": 0.0124,
      "grad_norm": 0.2632744014263153,
      "learning_rate": 0.00015752040000000002,
      "loss": 0.3217,
      "step": 106200
    },
    {
      "epoch": 0.0126,
      "grad_norm": 0.2666124701499939,
      "learning_rate": 0.00015748040000000002,
      "loss": 0.3302,
      "step": 106300
    },
    {
      "epoch": 0.0128,
      "grad_norm": 0.17219951748847961,
      "learning_rate": 0.0001574404,
      "loss": 0.3136,
      "step": 106400
    },
    {
      "epoch": 0.013,
      "grad_norm": 0.22360770404338837,
      "learning_rate": 0.0001574004,
      "loss": 0.3171,
      "step": 106500
    },
    {
      "epoch": 0.0132,
      "grad_norm": 0.24799229204654694,
      "learning_rate": 0.0001573604,
      "loss": 0.3081,
      "step": 106600
    },
    {
      "epoch": 0.0134,
      "grad_norm": 0.20889222621917725,
      "learning_rate": 0.00015732040000000002,
      "loss": 0.3466,
      "step": 106700
    },
    {
      "epoch": 0.0136,
      "grad_norm": 0.2627999186515808,
      "learning_rate": 0.0001572804,
      "loss": 0.3139,
      "step": 106800
    },
    {
      "epoch": 0.0138,
      "grad_norm": 0.3904314339160919,
      "learning_rate": 0.0001572404,
      "loss": 0.3438,
      "step": 106900
    },
    {
      "epoch": 0.014,
      "grad_norm": 0.2342156618833542,
      "learning_rate": 0.0001572004,
      "loss": 0.335,
      "step": 107000
    },
    {
      "epoch": 0.0142,
      "grad_norm": 0.17536994814872742,
      "learning_rate": 0.00015716040000000002,
      "loss": 0.3221,
      "step": 107100
    },
    {
      "epoch": 0.0144,
      "grad_norm": 0.3388653099536896,
      "learning_rate": 0.00015712040000000001,
      "loss": 0.3233,
      "step": 107200
    },
    {
      "epoch": 0.0146,
      "grad_norm": 0.1892201155424118,
      "learning_rate": 0.0001570804,
      "loss": 0.3233,
      "step": 107300
    },
    {
      "epoch": 0.0148,
      "grad_norm": 0.4015616178512573,
      "learning_rate": 0.0001570404,
      "loss": 0.3135,
      "step": 107400
    },
    {
      "epoch": 0.015,
      "grad_norm": 0.2883925139904022,
      "learning_rate": 0.0001570004,
      "loss": 0.3252,
      "step": 107500
    },
    {
      "epoch": 0.0152,
      "grad_norm": 0.27161914110183716,
      "learning_rate": 0.00015696040000000002,
      "loss": 0.3202,
      "step": 107600
    },
    {
      "epoch": 0.0154,
      "grad_norm": 0.1823039948940277,
      "learning_rate": 0.0001569204,
      "loss": 0.3132,
      "step": 107700
    },
    {
      "epoch": 0.0156,
      "grad_norm": 0.23866082727909088,
      "learning_rate": 0.0001568804,
      "loss": 0.3158,
      "step": 107800
    },
    {
      "epoch": 0.0158,
      "grad_norm": 0.3076813519001007,
      "learning_rate": 0.0001568404,
      "loss": 0.3097,
      "step": 107900
    },
    {
      "epoch": 0.016,
      "grad_norm": 0.3990881145000458,
      "learning_rate": 0.0001568004,
      "loss": 0.3117,
      "step": 108000
    },
    {
      "epoch": 0.0162,
      "grad_norm": 0.2077111005783081,
      "learning_rate": 0.0001567604,
      "loss": 0.3304,
      "step": 108100
    },
    {
      "epoch": 0.0164,
      "grad_norm": 0.20977669954299927,
      "learning_rate": 0.0001567204,
      "loss": 0.3177,
      "step": 108200
    },
    {
      "epoch": 0.0166,
      "grad_norm": 0.33695122599601746,
      "learning_rate": 0.00015668040000000003,
      "loss": 0.3127,
      "step": 108300
    },
    {
      "epoch": 0.0168,
      "grad_norm": 0.3933875262737274,
      "learning_rate": 0.0001566404,
      "loss": 0.3187,
      "step": 108400
    },
    {
      "epoch": 0.017,
      "grad_norm": 0.24768926203250885,
      "learning_rate": 0.0001566004,
      "loss": 0.3279,
      "step": 108500
    },
    {
      "epoch": 0.0172,
      "grad_norm": 0.23914366960525513,
      "learning_rate": 0.0001565604,
      "loss": 0.3389,
      "step": 108600
    },
    {
      "epoch": 0.0174,
      "grad_norm": 0.2776049077510834,
      "learning_rate": 0.0001565204,
      "loss": 0.3202,
      "step": 108700
    },
    {
      "epoch": 0.0176,
      "grad_norm": 0.2476639598608017,
      "learning_rate": 0.00015648040000000002,
      "loss": 0.3203,
      "step": 108800
    },
    {
      "epoch": 0.0178,
      "grad_norm": 0.25279584527015686,
      "learning_rate": 0.0001564404,
      "loss": 0.3031,
      "step": 108900
    },
    {
      "epoch": 0.018,
      "grad_norm": 0.19141827523708344,
      "learning_rate": 0.0001564004,
      "loss": 0.3146,
      "step": 109000
    },
    {
      "epoch": 0.0182,
      "grad_norm": 0.1972874402999878,
      "learning_rate": 0.0001563604,
      "loss": 0.3188,
      "step": 109100
    },
    {
      "epoch": 0.0184,
      "grad_norm": 0.21893855929374695,
      "learning_rate": 0.00015632040000000002,
      "loss": 0.3208,
      "step": 109200
    },
    {
      "epoch": 0.0186,
      "grad_norm": 0.2527053654193878,
      "learning_rate": 0.00015628040000000002,
      "loss": 0.3135,
      "step": 109300
    },
    {
      "epoch": 0.0188,
      "grad_norm": 0.21851500868797302,
      "learning_rate": 0.0001562404,
      "loss": 0.2979,
      "step": 109400
    },
    {
      "epoch": 0.019,
      "grad_norm": 0.41574743390083313,
      "learning_rate": 0.0001562004,
      "loss": 0.3395,
      "step": 109500
    },
    {
      "epoch": 0.0192,
      "grad_norm": 0.23641952872276306,
      "learning_rate": 0.0001561604,
      "loss": 0.3052,
      "step": 109600
    },
    {
      "epoch": 0.0194,
      "grad_norm": 0.2123720645904541,
      "learning_rate": 0.00015612040000000002,
      "loss": 0.3294,
      "step": 109700
    },
    {
      "epoch": 0.0196,
      "grad_norm": 0.2654692530632019,
      "learning_rate": 0.0001560804,
      "loss": 0.3133,
      "step": 109800
    },
    {
      "epoch": 0.0198,
      "grad_norm": 0.7307865023612976,
      "learning_rate": 0.0001560404,
      "loss": 0.3231,
      "step": 109900
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.23842667043209076,
      "learning_rate": 0.0001560004,
      "loss": 0.3278,
      "step": 110000
    },
    {
      "epoch": 0.0202,
      "grad_norm": 0.16572842001914978,
      "learning_rate": 0.00015596040000000002,
      "loss": 0.3166,
      "step": 110100
    },
    {
      "epoch": 0.0204,
      "grad_norm": 1.4259867668151855,
      "learning_rate": 0.0001559204,
      "loss": 0.3491,
      "step": 110200
    },
    {
      "epoch": 0.0206,
      "grad_norm": 0.2205478399991989,
      "learning_rate": 0.0001558804,
      "loss": 0.3302,
      "step": 110300
    },
    {
      "epoch": 0.0208,
      "grad_norm": 0.20075049996376038,
      "learning_rate": 0.0001558404,
      "loss": 0.3316,
      "step": 110400
    },
    {
      "epoch": 0.021,
      "grad_norm": 0.2577056884765625,
      "learning_rate": 0.0001558004,
      "loss": 0.2999,
      "step": 110500
    },
    {
      "epoch": 0.0212,
      "grad_norm": 0.2297847867012024,
      "learning_rate": 0.00015576040000000001,
      "loss": 0.3269,
      "step": 110600
    },
    {
      "epoch": 0.0214,
      "grad_norm": 0.21519100666046143,
      "learning_rate": 0.0001557204,
      "loss": 0.3017,
      "step": 110700
    },
    {
      "epoch": 0.0216,
      "grad_norm": 0.3626148998737335,
      "learning_rate": 0.00015568040000000003,
      "loss": 0.3238,
      "step": 110800
    },
    {
      "epoch": 0.0218,
      "grad_norm": 0.20477981865406036,
      "learning_rate": 0.0001556404,
      "loss": 0.3233,
      "step": 110900
    },
    {
      "epoch": 0.022,
      "grad_norm": 0.593943178653717,
      "learning_rate": 0.0001556004,
      "loss": 0.3187,
      "step": 111000
    },
    {
      "epoch": 0.0222,
      "grad_norm": 0.22717879712581635,
      "learning_rate": 0.0001555604,
      "loss": 0.329,
      "step": 111100
    },
    {
      "epoch": 0.0224,
      "grad_norm": 0.24597914516925812,
      "learning_rate": 0.0001555204,
      "loss": 0.3204,
      "step": 111200
    },
    {
      "epoch": 0.0226,
      "grad_norm": 0.6169671416282654,
      "learning_rate": 0.00015548040000000002,
      "loss": 0.3318,
      "step": 111300
    },
    {
      "epoch": 0.0228,
      "grad_norm": 0.512597918510437,
      "learning_rate": 0.0001554404,
      "loss": 0.3495,
      "step": 111400
    },
    {
      "epoch": 0.023,
      "grad_norm": 0.1982726901769638,
      "learning_rate": 0.0001554004,
      "loss": 0.3256,
      "step": 111500
    },
    {
      "epoch": 0.0232,
      "grad_norm": 0.24446536600589752,
      "learning_rate": 0.0001553604,
      "loss": 0.3182,
      "step": 111600
    },
    {
      "epoch": 0.0234,
      "grad_norm": 0.1990106999874115,
      "learning_rate": 0.0001553204,
      "loss": 0.3103,
      "step": 111700
    },
    {
      "epoch": 0.0236,
      "grad_norm": 0.32062387466430664,
      "learning_rate": 0.00015528040000000002,
      "loss": 0.3336,
      "step": 111800
    },
    {
      "epoch": 0.0238,
      "grad_norm": 0.24834854900836945,
      "learning_rate": 0.00015524039999999999,
      "loss": 0.3718,
      "step": 111900
    },
    {
      "epoch": 0.024,
      "grad_norm": 0.431271493434906,
      "learning_rate": 0.0001552004,
      "loss": 0.3148,
      "step": 112000
    },
    {
      "epoch": 0.0242,
      "grad_norm": 0.24593783915042877,
      "learning_rate": 0.0001551604,
      "loss": 0.3206,
      "step": 112100
    },
    {
      "epoch": 0.0244,
      "grad_norm": 0.22774197161197662,
      "learning_rate": 0.00015512040000000002,
      "loss": 0.323,
      "step": 112200
    },
    {
      "epoch": 0.0246,
      "grad_norm": 0.28193041682243347,
      "learning_rate": 0.00015508040000000001,
      "loss": 0.321,
      "step": 112300
    },
    {
      "epoch": 0.0248,
      "grad_norm": 0.27978020906448364,
      "learning_rate": 0.0001550404,
      "loss": 0.3293,
      "step": 112400
    },
    {
      "epoch": 0.025,
      "grad_norm": 0.33567145466804504,
      "learning_rate": 0.0001550004,
      "loss": 0.3398,
      "step": 112500
    },
    {
      "epoch": 0.0252,
      "grad_norm": 0.31685078144073486,
      "learning_rate": 0.0001549604,
      "loss": 0.3468,
      "step": 112600
    },
    {
      "epoch": 0.0254,
      "grad_norm": 0.23281008005142212,
      "learning_rate": 0.00015492040000000002,
      "loss": 0.3415,
      "step": 112700
    },
    {
      "epoch": 0.0256,
      "grad_norm": 0.29617008566856384,
      "learning_rate": 0.0001548804,
      "loss": 0.3215,
      "step": 112800
    },
    {
      "epoch": 0.0258,
      "grad_norm": 0.28596845269203186,
      "learning_rate": 0.0001548404,
      "loss": 0.3273,
      "step": 112900
    },
    {
      "epoch": 0.026,
      "grad_norm": 0.2520228922367096,
      "learning_rate": 0.0001548004,
      "loss": 0.3656,
      "step": 113000
    },
    {
      "epoch": 0.0262,
      "grad_norm": 0.1868925839662552,
      "learning_rate": 0.00015476040000000002,
      "loss": 0.3305,
      "step": 113100
    },
    {
      "epoch": 0.0264,
      "grad_norm": 0.2512012720108032,
      "learning_rate": 0.0001547204,
      "loss": 0.3331,
      "step": 113200
    },
    {
      "epoch": 0.0266,
      "grad_norm": 0.24037250876426697,
      "learning_rate": 0.0001546804,
      "loss": 0.3265,
      "step": 113300
    },
    {
      "epoch": 0.0268,
      "grad_norm": 0.44156473875045776,
      "learning_rate": 0.0001546404,
      "loss": 0.325,
      "step": 113400
    },
    {
      "epoch": 0.027,
      "grad_norm": 0.212955042719841,
      "learning_rate": 0.0001546004,
      "loss": 0.3103,
      "step": 113500
    },
    {
      "epoch": 0.0272,
      "grad_norm": 0.3389400839805603,
      "learning_rate": 0.0001545604,
      "loss": 0.2959,
      "step": 113600
    },
    {
      "epoch": 0.0274,
      "grad_norm": 0.21144139766693115,
      "learning_rate": 0.0001545204,
      "loss": 0.298,
      "step": 113700
    },
    {
      "epoch": 0.0276,
      "grad_norm": 0.2982296049594879,
      "learning_rate": 0.00015448040000000003,
      "loss": 0.2967,
      "step": 113800
    },
    {
      "epoch": 0.0278,
      "grad_norm": 0.23678374290466309,
      "learning_rate": 0.00015444040000000002,
      "loss": 0.2932,
      "step": 113900
    },
    {
      "epoch": 0.028,
      "grad_norm": 0.18785041570663452,
      "learning_rate": 0.0001544004,
      "loss": 0.3059,
      "step": 114000
    },
    {
      "epoch": 0.0282,
      "grad_norm": 0.27205973863601685,
      "learning_rate": 0.0001543604,
      "loss": 0.2817,
      "step": 114100
    },
    {
      "epoch": 0.0284,
      "grad_norm": 0.18749140202999115,
      "learning_rate": 0.0001543204,
      "loss": 0.2911,
      "step": 114200
    },
    {
      "epoch": 0.0286,
      "grad_norm": 0.21401409804821014,
      "learning_rate": 0.00015428040000000002,
      "loss": 0.3041,
      "step": 114300
    },
    {
      "epoch": 0.0288,
      "grad_norm": 0.20857536792755127,
      "learning_rate": 0.00015424040000000001,
      "loss": 0.2904,
      "step": 114400
    },
    {
      "epoch": 0.029,
      "grad_norm": 0.3695431053638458,
      "learning_rate": 0.0001542004,
      "loss": 0.3039,
      "step": 114500
    },
    {
      "epoch": 0.0292,
      "grad_norm": 0.42280298471450806,
      "learning_rate": 0.0001541604,
      "loss": 0.2926,
      "step": 114600
    },
    {
      "epoch": 0.0294,
      "grad_norm": 0.3620811998844147,
      "learning_rate": 0.0001541204,
      "loss": 0.29,
      "step": 114700
    },
    {
      "epoch": 0.0296,
      "grad_norm": 0.23109310865402222,
      "learning_rate": 0.00015408040000000002,
      "loss": 0.3008,
      "step": 114800
    },
    {
      "epoch": 0.0298,
      "grad_norm": 0.17469018697738647,
      "learning_rate": 0.0001540404,
      "loss": 0.2955,
      "step": 114900
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.25855448842048645,
      "learning_rate": 0.0001540004,
      "loss": 0.3033,
      "step": 115000
    },
    {
      "epoch": 0.0302,
      "grad_norm": 0.2488468438386917,
      "learning_rate": 0.0001539604,
      "loss": 0.3022,
      "step": 115100
    },
    {
      "epoch": 0.0304,
      "grad_norm": 0.19968360662460327,
      "learning_rate": 0.00015392040000000002,
      "loss": 0.2934,
      "step": 115200
    },
    {
      "epoch": 0.0306,
      "grad_norm": 0.22593122720718384,
      "learning_rate": 0.0001538804,
      "loss": 0.292,
      "step": 115300
    },
    {
      "epoch": 0.0308,
      "grad_norm": 0.27579984068870544,
      "learning_rate": 0.0001538404,
      "loss": 0.295,
      "step": 115400
    },
    {
      "epoch": 0.031,
      "grad_norm": 0.25232911109924316,
      "learning_rate": 0.0001538004,
      "loss": 0.3041,
      "step": 115500
    },
    {
      "epoch": 0.0312,
      "grad_norm": 0.295006662607193,
      "learning_rate": 0.0001537604,
      "loss": 0.3088,
      "step": 115600
    },
    {
      "epoch": 0.0314,
      "grad_norm": 0.520390510559082,
      "learning_rate": 0.0001537204,
      "loss": 0.2972,
      "step": 115700
    },
    {
      "epoch": 0.0316,
      "grad_norm": 0.24330058693885803,
      "learning_rate": 0.0001536804,
      "loss": 0.2937,
      "step": 115800
    },
    {
      "epoch": 0.0318,
      "grad_norm": 0.4119510054588318,
      "learning_rate": 0.00015364040000000003,
      "loss": 0.298,
      "step": 115900
    },
    {
      "epoch": 0.032,
      "grad_norm": 0.309059202671051,
      "learning_rate": 0.0001536004,
      "loss": 0.3025,
      "step": 116000
    },
    {
      "epoch": 0.0322,
      "grad_norm": 0.30650094151496887,
      "learning_rate": 0.00015356040000000001,
      "loss": 0.3046,
      "step": 116100
    },
    {
      "epoch": 0.0324,
      "grad_norm": 0.24224722385406494,
      "learning_rate": 0.0001535204,
      "loss": 0.299,
      "step": 116200
    },
    {
      "epoch": 0.0326,
      "grad_norm": 0.32729941606521606,
      "learning_rate": 0.0001534804,
      "loss": 0.3008,
      "step": 116300
    },
    {
      "epoch": 0.0328,
      "grad_norm": 0.1765100657939911,
      "learning_rate": 0.00015344040000000002,
      "loss": 0.3152,
      "step": 116400
    },
    {
      "epoch": 0.033,
      "grad_norm": 0.25575095415115356,
      "learning_rate": 0.0001534004,
      "loss": 0.306,
      "step": 116500
    },
    {
      "epoch": 0.0332,
      "grad_norm": 0.6274484992027283,
      "learning_rate": 0.0001533604,
      "loss": 0.3092,
      "step": 116600
    },
    {
      "epoch": 0.0334,
      "grad_norm": 0.21123872697353363,
      "learning_rate": 0.0001533204,
      "loss": 0.3978,
      "step": 116700
    },
    {
      "epoch": 0.0336,
      "grad_norm": 0.29489830136299133,
      "learning_rate": 0.00015328040000000002,
      "loss": 0.3038,
      "step": 116800
    },
    {
      "epoch": 0.0338,
      "grad_norm": 0.2510158121585846,
      "learning_rate": 0.00015324040000000002,
      "loss": 0.3114,
      "step": 116900
    },
    {
      "epoch": 0.034,
      "grad_norm": 0.23414947092533112,
      "learning_rate": 0.0001532004,
      "loss": 0.3052,
      "step": 117000
    },
    {
      "epoch": 0.0342,
      "grad_norm": 0.19055521488189697,
      "learning_rate": 0.0001531604,
      "loss": 0.3013,
      "step": 117100
    },
    {
      "epoch": 0.0344,
      "grad_norm": 0.8105829954147339,
      "learning_rate": 0.0001531204,
      "loss": 0.2967,
      "step": 117200
    },
    {
      "epoch": 0.0346,
      "grad_norm": 0.4188934564590454,
      "learning_rate": 0.00015308040000000002,
      "loss": 0.2951,
      "step": 117300
    },
    {
      "epoch": 0.0348,
      "grad_norm": 0.21309562027454376,
      "learning_rate": 0.0001530404,
      "loss": 0.2981,
      "step": 117400
    },
    {
      "epoch": 0.035,
      "grad_norm": 0.18121813237667084,
      "learning_rate": 0.0001530004,
      "loss": 0.2931,
      "step": 117500
    },
    {
      "epoch": 0.0352,
      "grad_norm": 0.632888674736023,
      "learning_rate": 0.0001529604,
      "loss": 0.2981,
      "step": 117600
    },
    {
      "epoch": 0.0354,
      "grad_norm": 0.3604109287261963,
      "learning_rate": 0.00015292040000000002,
      "loss": 0.2895,
      "step": 117700
    },
    {
      "epoch": 0.0356,
      "grad_norm": 0.2515547573566437,
      "learning_rate": 0.00015288040000000001,
      "loss": 0.3134,
      "step": 117800
    },
    {
      "epoch": 0.0358,
      "grad_norm": 0.26086676120758057,
      "learning_rate": 0.0001528404,
      "loss": 0.3068,
      "step": 117900
    },
    {
      "epoch": 0.036,
      "grad_norm": 0.20093345642089844,
      "learning_rate": 0.0001528004,
      "loss": 0.2985,
      "step": 118000
    },
    {
      "epoch": 0.0362,
      "grad_norm": 0.43283843994140625,
      "learning_rate": 0.0001527604,
      "loss": 0.304,
      "step": 118100
    },
    {
      "epoch": 0.0364,
      "grad_norm": 0.2671961784362793,
      "learning_rate": 0.00015272040000000002,
      "loss": 0.2987,
      "step": 118200
    },
    {
      "epoch": 0.0366,
      "grad_norm": 0.2851164638996124,
      "learning_rate": 0.0001526804,
      "loss": 0.3012,
      "step": 118300
    },
    {
      "epoch": 0.0368,
      "grad_norm": 0.27135607600212097,
      "learning_rate": 0.00015264040000000003,
      "loss": 0.3035,
      "step": 118400
    },
    {
      "epoch": 0.037,
      "grad_norm": 0.2493288666009903,
      "learning_rate": 0.0001526004,
      "loss": 0.3087,
      "step": 118500
    },
    {
      "epoch": 0.0372,
      "grad_norm": 0.2523597478866577,
      "learning_rate": 0.0001525604,
      "loss": 0.2925,
      "step": 118600
    },
    {
      "epoch": 0.0374,
      "grad_norm": 2.79110050201416,
      "learning_rate": 0.0001525204,
      "loss": 0.3105,
      "step": 118700
    },
    {
      "epoch": 0.0376,
      "grad_norm": 0.20949222147464752,
      "learning_rate": 0.0001524804,
      "loss": 0.3069,
      "step": 118800
    },
    {
      "epoch": 0.0378,
      "grad_norm": 0.2129492163658142,
      "learning_rate": 0.00015244040000000003,
      "loss": 0.2957,
      "step": 118900
    },
    {
      "epoch": 0.038,
      "grad_norm": 0.2193364053964615,
      "learning_rate": 0.0001524004,
      "loss": 0.2939,
      "step": 119000
    },
    {
      "epoch": 0.0382,
      "grad_norm": 0.21624450385570526,
      "learning_rate": 0.0001523604,
      "loss": 0.3033,
      "step": 119100
    },
    {
      "epoch": 0.0384,
      "grad_norm": 0.16359418630599976,
      "learning_rate": 0.0001523204,
      "loss": 0.3051,
      "step": 119200
    },
    {
      "epoch": 0.0386,
      "grad_norm": 0.38990601897239685,
      "learning_rate": 0.0001522804,
      "loss": 0.2948,
      "step": 119300
    },
    {
      "epoch": 0.0388,
      "grad_norm": 0.25620323419570923,
      "learning_rate": 0.00015224040000000002,
      "loss": 0.3069,
      "step": 119400
    },
    {
      "epoch": 0.039,
      "grad_norm": 0.3383728563785553,
      "learning_rate": 0.0001522004,
      "loss": 0.3168,
      "step": 119500
    },
    {
      "epoch": 0.0392,
      "grad_norm": 0.296837717294693,
      "learning_rate": 0.0001521604,
      "loss": 0.3068,
      "step": 119600
    },
    {
      "epoch": 0.0394,
      "grad_norm": 0.17246387898921967,
      "learning_rate": 0.0001521204,
      "loss": 0.2959,
      "step": 119700
    },
    {
      "epoch": 0.0396,
      "grad_norm": 0.27589595317840576,
      "learning_rate": 0.00015208040000000002,
      "loss": 0.2961,
      "step": 119800
    },
    {
      "epoch": 0.0398,
      "grad_norm": 0.2941451668739319,
      "learning_rate": 0.00015204040000000002,
      "loss": 0.294,
      "step": 119900
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.18005476891994476,
      "learning_rate": 0.0001520004,
      "loss": 0.3013,
      "step": 120000
    },
    {
      "epoch": 0.0002,
      "grad_norm": 0.21327625215053558,
      "learning_rate": 0.0001519604,
      "loss": 0.2806,
      "step": 120100
    },
    {
      "epoch": 0.0004,
      "grad_norm": 0.4669700562953949,
      "learning_rate": 0.0001519204,
      "loss": 0.2849,
      "step": 120200
    },
    {
      "epoch": 0.0006,
      "grad_norm": 0.3588303029537201,
      "learning_rate": 0.00015188040000000002,
      "loss": 0.2861,
      "step": 120300
    },
    {
      "epoch": 0.0008,
      "grad_norm": 0.28102320432662964,
      "learning_rate": 0.0001518404,
      "loss": 0.2878,
      "step": 120400
    },
    {
      "epoch": 0.001,
      "grad_norm": 0.1880110502243042,
      "learning_rate": 0.0001518004,
      "loss": 0.2811,
      "step": 120500
    },
    {
      "epoch": 0.0012,
      "grad_norm": 0.35984283685684204,
      "learning_rate": 0.0001517604,
      "loss": 0.277,
      "step": 120600
    },
    {
      "epoch": 0.0014,
      "grad_norm": 0.46658045053482056,
      "learning_rate": 0.00015172040000000002,
      "loss": 0.2813,
      "step": 120700
    },
    {
      "epoch": 0.0016,
      "grad_norm": 0.3781064748764038,
      "learning_rate": 0.0001516804,
      "loss": 0.2833,
      "step": 120800
    },
    {
      "epoch": 0.0018,
      "grad_norm": 0.30914661288261414,
      "learning_rate": 0.0001516404,
      "loss": 0.2806,
      "step": 120900
    },
    {
      "epoch": 0.002,
      "grad_norm": 0.32875189185142517,
      "learning_rate": 0.0001516004,
      "loss": 0.2778,
      "step": 121000
    },
    {
      "epoch": 0.0022,
      "grad_norm": 0.21213066577911377,
      "learning_rate": 0.0001515604,
      "loss": 0.2951,
      "step": 121100
    },
    {
      "epoch": 0.0024,
      "grad_norm": 0.31731417775154114,
      "learning_rate": 0.00015152040000000001,
      "loss": 0.2819,
      "step": 121200
    },
    {
      "epoch": 0.0026,
      "grad_norm": 0.5732041597366333,
      "learning_rate": 0.0001514804,
      "loss": 0.2908,
      "step": 121300
    },
    {
      "epoch": 0.0028,
      "grad_norm": 0.1619558483362198,
      "learning_rate": 0.00015144040000000003,
      "loss": 0.2796,
      "step": 121400
    },
    {
      "epoch": 0.003,
      "grad_norm": 0.4144911766052246,
      "learning_rate": 0.0001514004,
      "loss": 0.2892,
      "step": 121500
    },
    {
      "epoch": 0.0032,
      "grad_norm": 0.2565133273601532,
      "learning_rate": 0.00015136040000000002,
      "loss": 0.3034,
      "step": 121600
    },
    {
      "epoch": 0.0034,
      "grad_norm": 0.5545548796653748,
      "learning_rate": 0.0001513204,
      "loss": 0.2899,
      "step": 121700
    },
    {
      "epoch": 0.0036,
      "grad_norm": 0.21982091665267944,
      "learning_rate": 0.0001512804,
      "loss": 0.2901,
      "step": 121800
    },
    {
      "epoch": 0.0038,
      "grad_norm": 0.20543520152568817,
      "learning_rate": 0.00015124040000000002,
      "loss": 0.2925,
      "step": 121900
    },
    {
      "epoch": 0.004,
      "grad_norm": 0.27426043152809143,
      "learning_rate": 0.0001512004,
      "loss": 0.2911,
      "step": 122000
    },
    {
      "epoch": 0.0042,
      "grad_norm": 0.20343656837940216,
      "learning_rate": 0.0001511604,
      "loss": 0.2918,
      "step": 122100
    },
    {
      "epoch": 0.0044,
      "grad_norm": 0.21032807230949402,
      "learning_rate": 0.0001511204,
      "loss": 0.3072,
      "step": 122200
    },
    {
      "epoch": 0.0046,
      "grad_norm": 0.3014376163482666,
      "learning_rate": 0.0001510804,
      "loss": 0.2942,
      "step": 122300
    },
    {
      "epoch": 0.0048,
      "grad_norm": 0.2240341305732727,
      "learning_rate": 0.00015104040000000002,
      "loss": 0.2799,
      "step": 122400
    },
    {
      "epoch": 0.005,
      "grad_norm": 0.2607118785381317,
      "learning_rate": 0.00015100039999999998,
      "loss": 0.2782,
      "step": 122500
    },
    {
      "epoch": 0.0052,
      "grad_norm": 0.2203282117843628,
      "learning_rate": 0.0001509604,
      "loss": 0.28,
      "step": 122600
    },
    {
      "epoch": 0.0054,
      "grad_norm": 0.275749534368515,
      "learning_rate": 0.0001509204,
      "loss": 0.2859,
      "step": 122700
    },
    {
      "epoch": 0.0056,
      "grad_norm": 0.2172224372625351,
      "learning_rate": 0.00015088040000000002,
      "loss": 0.2916,
      "step": 122800
    },
    {
      "epoch": 0.0058,
      "grad_norm": 0.24563243985176086,
      "learning_rate": 0.00015084040000000001,
      "loss": 0.2784,
      "step": 122900
    },
    {
      "epoch": 0.006,
      "grad_norm": 0.25981712341308594,
      "learning_rate": 0.0001508004,
      "loss": 0.3223,
      "step": 123000
    },
    {
      "epoch": 0.0062,
      "grad_norm": 0.22978562116622925,
      "learning_rate": 0.0001507604,
      "loss": 0.2904,
      "step": 123100
    },
    {
      "epoch": 0.0064,
      "grad_norm": 0.18407663702964783,
      "learning_rate": 0.0001507204,
      "loss": 0.2847,
      "step": 123200
    },
    {
      "epoch": 0.0066,
      "grad_norm": 0.697337806224823,
      "learning_rate": 0.00015068040000000002,
      "loss": 0.2996,
      "step": 123300
    },
    {
      "epoch": 0.0068,
      "grad_norm": 0.17664720118045807,
      "learning_rate": 0.0001506404,
      "loss": 0.2878,
      "step": 123400
    },
    {
      "epoch": 0.007,
      "grad_norm": 0.24168753623962402,
      "learning_rate": 0.0001506004,
      "loss": 0.2988,
      "step": 123500
    },
    {
      "epoch": 0.0072,
      "grad_norm": 0.20818035304546356,
      "learning_rate": 0.0001505604,
      "loss": 0.2945,
      "step": 123600
    },
    {
      "epoch": 0.0074,
      "grad_norm": 0.22152996063232422,
      "learning_rate": 0.00015052040000000002,
      "loss": 0.3007,
      "step": 123700
    },
    {
      "epoch": 0.0076,
      "grad_norm": 0.15268588066101074,
      "learning_rate": 0.0001504804,
      "loss": 0.3035,
      "step": 123800
    },
    {
      "epoch": 0.0078,
      "grad_norm": 0.19523349404335022,
      "learning_rate": 0.0001504404,
      "loss": 0.2886,
      "step": 123900
    },
    {
      "epoch": 0.008,
      "grad_norm": 0.25065121054649353,
      "learning_rate": 0.0001504004,
      "loss": 0.2931,
      "step": 124000
    },
    {
      "epoch": 0.0082,
      "grad_norm": 0.2577344477176666,
      "learning_rate": 0.0001503604,
      "loss": 0.3029,
      "step": 124100
    },
    {
      "epoch": 0.0084,
      "grad_norm": 0.49803975224494934,
      "learning_rate": 0.0001503204,
      "loss": 0.2879,
      "step": 124200
    },
    {
      "epoch": 0.0086,
      "grad_norm": 0.2067602127790451,
      "learning_rate": 0.0001502804,
      "loss": 0.2863,
      "step": 124300
    },
    {
      "epoch": 0.0088,
      "grad_norm": 0.14126543700695038,
      "learning_rate": 0.00015024040000000003,
      "loss": 0.2845,
      "step": 124400
    },
    {
      "epoch": 0.009,
      "grad_norm": 0.24404196441173553,
      "learning_rate": 0.0001502004,
      "loss": 0.2985,
      "step": 124500
    },
    {
      "epoch": 0.0092,
      "grad_norm": 0.2484026700258255,
      "learning_rate": 0.0001501604,
      "loss": 0.3004,
      "step": 124600
    },
    {
      "epoch": 0.0094,
      "grad_norm": 0.3001592457294464,
      "learning_rate": 0.0001501204,
      "loss": 0.2953,
      "step": 124700
    },
    {
      "epoch": 0.0096,
      "grad_norm": 0.2282063364982605,
      "learning_rate": 0.0001500804,
      "loss": 0.3028,
      "step": 124800
    },
    {
      "epoch": 0.0098,
      "grad_norm": 0.20628437399864197,
      "learning_rate": 0.00015004040000000002,
      "loss": 0.2982,
      "step": 124900
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.25722581148147583,
      "learning_rate": 0.0001500004,
      "loss": 0.2941,
      "step": 125000
    },
    {
      "epoch": 0.0102,
      "grad_norm": 0.23121199011802673,
      "learning_rate": 0.0001499604,
      "loss": 0.3042,
      "step": 125100
    },
    {
      "epoch": 0.0104,
      "grad_norm": 0.26357603073120117,
      "learning_rate": 0.0001499204,
      "loss": 0.2879,
      "step": 125200
    },
    {
      "epoch": 0.0106,
      "grad_norm": 0.24245673418045044,
      "learning_rate": 0.00014988040000000002,
      "loss": 0.2984,
      "step": 125300
    },
    {
      "epoch": 0.0108,
      "grad_norm": 0.23373626172542572,
      "learning_rate": 0.00014984040000000002,
      "loss": 0.3001,
      "step": 125400
    },
    {
      "epoch": 0.011,
      "grad_norm": 0.22621707618236542,
      "learning_rate": 0.00014980039999999998,
      "loss": 0.2982,
      "step": 125500
    },
    {
      "epoch": 0.0112,
      "grad_norm": 0.1624283492565155,
      "learning_rate": 0.0001497604,
      "loss": 0.2955,
      "step": 125600
    },
    {
      "epoch": 0.0114,
      "grad_norm": 0.2670327126979828,
      "learning_rate": 0.0001497204,
      "loss": 0.2883,
      "step": 125700
    },
    {
      "epoch": 0.0116,
      "grad_norm": 0.23716334998607635,
      "learning_rate": 0.00014968040000000002,
      "loss": 0.3013,
      "step": 125800
    },
    {
      "epoch": 0.0118,
      "grad_norm": 0.19362571835517883,
      "learning_rate": 0.0001496404,
      "loss": 0.3015,
      "step": 125900
    },
    {
      "epoch": 0.012,
      "grad_norm": 0.16306428611278534,
      "learning_rate": 0.0001496004,
      "loss": 0.2924,
      "step": 126000
    },
    {
      "epoch": 0.0122,
      "grad_norm": 0.25153741240501404,
      "learning_rate": 0.0001495604,
      "loss": 0.3193,
      "step": 126100
    },
    {
      "epoch": 0.0124,
      "grad_norm": 0.2298291176557541,
      "learning_rate": 0.0001495204,
      "loss": 0.2925,
      "step": 126200
    },
    {
      "epoch": 0.0126,
      "grad_norm": 0.2596379220485687,
      "learning_rate": 0.0001494804,
      "loss": 0.2992,
      "step": 126300
    },
    {
      "epoch": 0.0128,
      "grad_norm": 0.18484342098236084,
      "learning_rate": 0.0001494404,
      "loss": 0.2869,
      "step": 126400
    },
    {
      "epoch": 0.013,
      "grad_norm": 0.21150384843349457,
      "learning_rate": 0.00014940040000000003,
      "loss": 0.2917,
      "step": 126500
    },
    {
      "epoch": 0.0132,
      "grad_norm": 0.210966095328331,
      "learning_rate": 0.0001493604,
      "loss": 0.2858,
      "step": 126600
    },
    {
      "epoch": 0.0134,
      "grad_norm": 0.19671154022216797,
      "learning_rate": 0.00014932040000000001,
      "loss": 0.3121,
      "step": 126700
    },
    {
      "epoch": 0.0136,
      "grad_norm": 0.24427378177642822,
      "learning_rate": 0.0001492804,
      "loss": 0.2878,
      "step": 126800
    },
    {
      "epoch": 0.0138,
      "grad_norm": 0.278323233127594,
      "learning_rate": 0.0001492404,
      "loss": 0.3073,
      "step": 126900
    },
    {
      "epoch": 0.014,
      "grad_norm": 0.2476424276828766,
      "learning_rate": 0.00014920040000000002,
      "loss": 0.3057,
      "step": 127000
    },
    {
      "epoch": 0.0142,
      "grad_norm": 0.18065540492534637,
      "learning_rate": 0.0001491604,
      "loss": 0.2981,
      "step": 127100
    },
    {
      "epoch": 0.0144,
      "grad_norm": 0.28167441487312317,
      "learning_rate": 0.0001491204,
      "loss": 0.2962,
      "step": 127200
    },
    {
      "epoch": 0.0146,
      "grad_norm": 0.19970493018627167,
      "learning_rate": 0.0001490804,
      "loss": 0.2969,
      "step": 127300
    },
    {
      "epoch": 0.0148,
      "grad_norm": 0.40125706791877747,
      "learning_rate": 0.00014904040000000002,
      "loss": 0.2901,
      "step": 127400
    },
    {
      "epoch": 0.015,
      "grad_norm": 0.24552318453788757,
      "learning_rate": 0.00014900040000000002,
      "loss": 0.3009,
      "step": 127500
    },
    {
      "epoch": 0.0152,
      "grad_norm": 0.22551432251930237,
      "learning_rate": 0.0001489604,
      "loss": 0.2938,
      "step": 127600
    },
    {
      "epoch": 0.0154,
      "grad_norm": 0.21164937317371368,
      "learning_rate": 0.0001489204,
      "loss": 0.2914,
      "step": 127700
    },
    {
      "epoch": 0.0156,
      "grad_norm": 0.22755509614944458,
      "learning_rate": 0.0001488804,
      "loss": 0.2918,
      "step": 127800
    },
    {
      "epoch": 0.0158,
      "grad_norm": 0.28624242544174194,
      "learning_rate": 0.00014884040000000002,
      "loss": 0.2889,
      "step": 127900
    },
    {
      "epoch": 0.016,
      "grad_norm": 0.42901697754859924,
      "learning_rate": 0.0001488004,
      "loss": 0.2893,
      "step": 128000
    },
    {
      "epoch": 0.0162,
      "grad_norm": 0.2009563446044922,
      "learning_rate": 0.0001487604,
      "loss": 0.3049,
      "step": 128100
    },
    {
      "epoch": 0.0164,
      "grad_norm": 0.2092672884464264,
      "learning_rate": 0.0001487204,
      "loss": 0.2929,
      "step": 128200
    },
    {
      "epoch": 0.0166,
      "grad_norm": 0.2817251682281494,
      "learning_rate": 0.00014868040000000002,
      "loss": 0.2916,
      "step": 128300
    },
    {
      "epoch": 0.0168,
      "grad_norm": 0.3064669668674469,
      "learning_rate": 0.00014864040000000001,
      "loss": 0.2961,
      "step": 128400
    },
    {
      "epoch": 0.017,
      "grad_norm": 0.23615750670433044,
      "learning_rate": 0.0001486004,
      "loss": 0.3013,
      "step": 128500
    },
    {
      "epoch": 0.0172,
      "grad_norm": 0.2097761034965515,
      "learning_rate": 0.0001485604,
      "loss": 0.3128,
      "step": 128600
    },
    {
      "epoch": 0.0174,
      "grad_norm": 0.20526331663131714,
      "learning_rate": 0.0001485204,
      "loss": 0.2986,
      "step": 128700
    },
    {
      "epoch": 0.0176,
      "grad_norm": 0.24395328760147095,
      "learning_rate": 0.00014848040000000002,
      "loss": 0.2977,
      "step": 128800
    },
    {
      "epoch": 0.0178,
      "grad_norm": 0.31493526697158813,
      "learning_rate": 0.0001484404,
      "loss": 0.2832,
      "step": 128900
    },
    {
      "epoch": 0.018,
      "grad_norm": 0.19086699187755585,
      "learning_rate": 0.00014840040000000003,
      "loss": 0.2905,
      "step": 129000
    },
    {
      "epoch": 0.0182,
      "grad_norm": 0.16701176762580872,
      "learning_rate": 0.0001483604,
      "loss": 0.2968,
      "step": 129100
    },
    {
      "epoch": 0.0184,
      "grad_norm": 0.18984141945838928,
      "learning_rate": 0.0001483204,
      "loss": 0.2944,
      "step": 129200
    },
    {
      "epoch": 0.0186,
      "grad_norm": 0.21061408519744873,
      "learning_rate": 0.0001482804,
      "loss": 0.2907,
      "step": 129300
    },
    {
      "epoch": 0.0188,
      "grad_norm": 0.18577519059181213,
      "learning_rate": 0.0001482404,
      "loss": 0.2783,
      "step": 129400
    },
    {
      "epoch": 0.019,
      "grad_norm": 0.36443671584129333,
      "learning_rate": 0.00014820040000000003,
      "loss": 0.3125,
      "step": 129500
    },
    {
      "epoch": 0.0192,
      "grad_norm": 0.25582483410835266,
      "learning_rate": 0.0001481604,
      "loss": 0.2863,
      "step": 129600
    },
    {
      "epoch": 0.0194,
      "grad_norm": 0.16937601566314697,
      "learning_rate": 0.0001481204,
      "loss": 0.3062,
      "step": 129700
    },
    {
      "epoch": 0.0196,
      "grad_norm": 0.2256769835948944,
      "learning_rate": 0.0001480804,
      "loss": 0.2894,
      "step": 129800
    },
    {
      "epoch": 0.0198,
      "grad_norm": 0.5192826390266418,
      "learning_rate": 0.0001480404,
      "loss": 0.2953,
      "step": 129900
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.21238556504249573,
      "learning_rate": 0.00014800040000000002,
      "loss": 0.3024,
      "step": 130000
    },
    {
      "epoch": 0.0202,
      "grad_norm": 0.15953408181667328,
      "learning_rate": 0.0001479604,
      "loss": 0.2966,
      "step": 130100
    },
    {
      "epoch": 0.0204,
      "grad_norm": 0.6439893841743469,
      "learning_rate": 0.0001479204,
      "loss": 0.3231,
      "step": 130200
    },
    {
      "epoch": 0.0206,
      "grad_norm": 0.2107216715812683,
      "learning_rate": 0.0001478804,
      "loss": 0.3037,
      "step": 130300
    },
    {
      "epoch": 0.0208,
      "grad_norm": 0.2086646556854248,
      "learning_rate": 0.00014784040000000002,
      "loss": 0.3024,
      "step": 130400
    },
    {
      "epoch": 0.021,
      "grad_norm": 0.23454782366752625,
      "learning_rate": 0.00014780040000000002,
      "loss": 0.2821,
      "step": 130500
    },
    {
      "epoch": 0.0212,
      "grad_norm": 0.195747971534729,
      "learning_rate": 0.0001477604,
      "loss": 0.2976,
      "step": 130600
    },
    {
      "epoch": 0.0214,
      "grad_norm": 0.22037339210510254,
      "learning_rate": 0.0001477204,
      "loss": 0.2796,
      "step": 130700
    },
    {
      "epoch": 0.0216,
      "grad_norm": 0.28433117270469666,
      "learning_rate": 0.0001476804,
      "loss": 0.3009,
      "step": 130800
    },
    {
      "epoch": 0.0218,
      "grad_norm": 0.5101615786552429,
      "learning_rate": 0.00014764040000000002,
      "loss": 0.3018,
      "step": 130900
    },
    {
      "epoch": 0.022,
      "grad_norm": 0.4728710353374481,
      "learning_rate": 0.0001476004,
      "loss": 0.2966,
      "step": 131000
    },
    {
      "epoch": 0.0222,
      "grad_norm": 0.2941736876964569,
      "learning_rate": 0.0001475604,
      "loss": 0.3063,
      "step": 131100
    },
    {
      "epoch": 0.0224,
      "grad_norm": 0.26847612857818604,
      "learning_rate": 0.0001475204,
      "loss": 0.3034,
      "step": 131200
    },
    {
      "epoch": 0.0226,
      "grad_norm": 0.30397793650627136,
      "learning_rate": 0.00014748040000000002,
      "loss": 0.3112,
      "step": 131300
    },
    {
      "epoch": 0.0228,
      "grad_norm": 0.48798418045043945,
      "learning_rate": 0.0001474404,
      "loss": 0.3263,
      "step": 131400
    },
    {
      "epoch": 0.023,
      "grad_norm": 0.20954224467277527,
      "learning_rate": 0.0001474004,
      "loss": 0.3032,
      "step": 131500
    },
    {
      "epoch": 0.0232,
      "grad_norm": 0.26696014404296875,
      "learning_rate": 0.0001473604,
      "loss": 0.2978,
      "step": 131600
    },
    {
      "epoch": 0.0234,
      "grad_norm": 0.1869426965713501,
      "learning_rate": 0.0001473204,
      "loss": 0.2897,
      "step": 131700
    },
    {
      "epoch": 0.0236,
      "grad_norm": 0.2813512980937958,
      "learning_rate": 0.00014728040000000001,
      "loss": 0.3065,
      "step": 131800
    },
    {
      "epoch": 0.0238,
      "grad_norm": 0.2745366096496582,
      "learning_rate": 0.0001472404,
      "loss": 0.3405,
      "step": 131900
    },
    {
      "epoch": 0.024,
      "grad_norm": 0.3381493091583252,
      "learning_rate": 0.00014720040000000003,
      "loss": 0.2944,
      "step": 132000
    },
    {
      "epoch": 0.0242,
      "grad_norm": 0.20233681797981262,
      "learning_rate": 0.0001471604,
      "loss": 0.3009,
      "step": 132100
    },
    {
      "epoch": 0.0244,
      "grad_norm": 0.24026724696159363,
      "learning_rate": 0.00014712040000000002,
      "loss": 0.2996,
      "step": 132200
    },
    {
      "epoch": 0.0246,
      "grad_norm": 0.24572235345840454,
      "learning_rate": 0.0001470804,
      "loss": 0.2985,
      "step": 132300
    },
    {
      "epoch": 0.0248,
      "grad_norm": 0.26701977849006653,
      "learning_rate": 0.0001470404,
      "loss": 0.3037,
      "step": 132400
    },
    {
      "epoch": 0.025,
      "grad_norm": 0.24025647342205048,
      "learning_rate": 0.00014700040000000002,
      "loss": 0.3154,
      "step": 132500
    },
    {
      "epoch": 0.0252,
      "grad_norm": 0.3687751889228821,
      "learning_rate": 0.0001469604,
      "loss": 0.3188,
      "step": 132600
    },
    {
      "epoch": 0.0254,
      "grad_norm": 0.215996652841568,
      "learning_rate": 0.0001469204,
      "loss": 0.3143,
      "step": 132700
    },
    {
      "epoch": 0.0256,
      "grad_norm": 0.2471071183681488,
      "learning_rate": 0.0001468804,
      "loss": 0.3019,
      "step": 132800
    },
    {
      "epoch": 0.0258,
      "grad_norm": 0.21399539709091187,
      "learning_rate": 0.00014684040000000002,
      "loss": 0.3064,
      "step": 132900
    },
    {
      "epoch": 0.026,
      "grad_norm": 0.2229408621788025,
      "learning_rate": 0.00014680040000000002,
      "loss": 0.3383,
      "step": 133000
    },
    {
      "epoch": 0.0262,
      "grad_norm": 0.1846669614315033,
      "learning_rate": 0.00014676039999999998,
      "loss": 0.3073,
      "step": 133100
    },
    {
      "epoch": 0.0264,
      "grad_norm": 0.2286502569913864,
      "learning_rate": 0.0001467204,
      "loss": 0.3073,
      "step": 133200
    },
    {
      "epoch": 0.0266,
      "grad_norm": 0.21924042701721191,
      "learning_rate": 0.0001466804,
      "loss": 0.3055,
      "step": 133300
    },
    {
      "epoch": 0.0268,
      "grad_norm": 0.3561740517616272,
      "learning_rate": 0.00014664040000000002,
      "loss": 0.3023,
      "step": 133400
    },
    {
      "epoch": 0.027,
      "grad_norm": 0.2278672158718109,
      "learning_rate": 0.0001466004,
      "loss": 0.2862,
      "step": 133500
    },
    {
      "epoch": 0.0272,
      "grad_norm": 0.28851574659347534,
      "learning_rate": 0.0001465604,
      "loss": 0.2724,
      "step": 133600
    },
    {
      "epoch": 0.0274,
      "grad_norm": 0.20886564254760742,
      "learning_rate": 0.0001465204,
      "loss": 0.2766,
      "step": 133700
    },
    {
      "epoch": 0.0276,
      "grad_norm": 0.2499471753835678,
      "learning_rate": 0.0001464804,
      "loss": 0.2761,
      "step": 133800
    },
    {
      "epoch": 0.0278,
      "grad_norm": 0.1883542537689209,
      "learning_rate": 0.00014644040000000002,
      "loss": 0.2739,
      "step": 133900
    },
    {
      "epoch": 0.028,
      "grad_norm": 0.16889332234859467,
      "learning_rate": 0.0001464004,
      "loss": 0.284,
      "step": 134000
    },
    {
      "epoch": 0.0282,
      "grad_norm": 0.23524640500545502,
      "learning_rate": 0.0001463604,
      "loss": 0.265,
      "step": 134100
    },
    {
      "epoch": 0.0284,
      "grad_norm": 0.18847911059856415,
      "learning_rate": 0.0001463204,
      "loss": 0.2718,
      "step": 134200
    },
    {
      "epoch": 0.0286,
      "grad_norm": 0.1972835212945938,
      "learning_rate": 0.00014628040000000002,
      "loss": 0.2858,
      "step": 134300
    },
    {
      "epoch": 0.0288,
      "grad_norm": 0.2064981758594513,
      "learning_rate": 0.0001462404,
      "loss": 0.2715,
      "step": 134400
    },
    {
      "epoch": 0.029,
      "grad_norm": 0.39654669165611267,
      "learning_rate": 0.0001462004,
      "loss": 0.2831,
      "step": 134500
    },
    {
      "epoch": 0.0292,
      "grad_norm": 0.7286176085472107,
      "learning_rate": 0.0001461604,
      "loss": 0.2753,
      "step": 134600
    },
    {
      "epoch": 0.0294,
      "grad_norm": 0.2998080849647522,
      "learning_rate": 0.0001461204,
      "loss": 0.2737,
      "step": 134700
    },
    {
      "epoch": 0.0296,
      "grad_norm": 0.22182480990886688,
      "learning_rate": 0.0001460804,
      "loss": 0.2809,
      "step": 134800
    },
    {
      "epoch": 0.0298,
      "grad_norm": 0.3183176517486572,
      "learning_rate": 0.0001460404,
      "loss": 0.278,
      "step": 134900
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.24670830368995667,
      "learning_rate": 0.00014600040000000003,
      "loss": 0.284,
      "step": 135000
    },
    {
      "epoch": 0.0302,
      "grad_norm": 0.22563083469867706,
      "learning_rate": 0.0001459604,
      "loss": 0.2843,
      "step": 135100
    },
    {
      "epoch": 0.0304,
      "grad_norm": 0.2058597058057785,
      "learning_rate": 0.0001459204,
      "loss": 0.2767,
      "step": 135200
    },
    {
      "epoch": 0.0306,
      "grad_norm": 0.21196039021015167,
      "learning_rate": 0.0001458804,
      "loss": 0.2765,
      "step": 135300
    },
    {
      "epoch": 0.0308,
      "grad_norm": 0.24366965889930725,
      "learning_rate": 0.0001458404,
      "loss": 0.2783,
      "step": 135400
    },
    {
      "epoch": 0.031,
      "grad_norm": 0.24236196279525757,
      "learning_rate": 0.00014580040000000002,
      "loss": 0.2873,
      "step": 135500
    },
    {
      "epoch": 0.0312,
      "grad_norm": 0.2742379903793335,
      "learning_rate": 0.0001457604,
      "loss": 0.2895,
      "step": 135600
    },
    {
      "epoch": 0.0314,
      "grad_norm": 0.40432196855545044,
      "learning_rate": 0.0001457204,
      "loss": 0.2802,
      "step": 135700
    },
    {
      "epoch": 0.0316,
      "grad_norm": 0.21296831965446472,
      "learning_rate": 0.0001456804,
      "loss": 0.2787,
      "step": 135800
    },
    {
      "epoch": 0.0318,
      "grad_norm": 0.38033196330070496,
      "learning_rate": 0.00014564040000000002,
      "loss": 0.2832,
      "step": 135900
    },
    {
      "epoch": 0.032,
      "grad_norm": 0.28289541602134705,
      "learning_rate": 0.00014560040000000002,
      "loss": 0.2853,
      "step": 136000
    },
    {
      "epoch": 0.0322,
      "grad_norm": 0.21910668909549713,
      "learning_rate": 0.0001455604,
      "loss": 0.2872,
      "step": 136100
    },
    {
      "epoch": 0.0324,
      "grad_norm": 0.23765510320663452,
      "learning_rate": 0.0001455204,
      "loss": 0.2826,
      "step": 136200
    },
    {
      "epoch": 0.0326,
      "grad_norm": 0.27492231130599976,
      "learning_rate": 0.0001454804,
      "loss": 0.2825,
      "step": 136300
    },
    {
      "epoch": 0.0328,
      "grad_norm": 0.16266077756881714,
      "learning_rate": 0.00014544040000000002,
      "loss": 0.2949,
      "step": 136400
    },
    {
      "epoch": 0.033,
      "grad_norm": 0.21813906729221344,
      "learning_rate": 0.0001454004,
      "loss": 0.2898,
      "step": 136500
    },
    {
      "epoch": 0.0332,
      "grad_norm": 0.6559190154075623,
      "learning_rate": 0.0001453604,
      "loss": 0.2917,
      "step": 136600
    },
    {
      "epoch": 0.0334,
      "grad_norm": 0.21168918907642365,
      "learning_rate": 0.0001453204,
      "loss": 0.3655,
      "step": 136700
    },
    {
      "epoch": 0.0336,
      "grad_norm": 0.2966395616531372,
      "learning_rate": 0.0001452804,
      "loss": 0.2859,
      "step": 136800
    },
    {
      "epoch": 0.0338,
      "grad_norm": 0.20525045692920685,
      "learning_rate": 0.0001452404,
      "loss": 0.293,
      "step": 136900
    },
    {
      "epoch": 0.034,
      "grad_norm": 0.17695209383964539,
      "learning_rate": 0.0001452004,
      "loss": 0.2869,
      "step": 137000
    },
    {
      "epoch": 0.0342,
      "grad_norm": 0.182274729013443,
      "learning_rate": 0.0001451604,
      "loss": 0.2843,
      "step": 137100
    },
    {
      "epoch": 0.0344,
      "grad_norm": 0.8087812662124634,
      "learning_rate": 0.0001451204,
      "loss": 0.28,
      "step": 137200
    },
    {
      "epoch": 0.0346,
      "grad_norm": 0.33212119340896606,
      "learning_rate": 0.00014508040000000001,
      "loss": 0.2787,
      "step": 137300
    },
    {
      "epoch": 0.0348,
      "grad_norm": 0.18305575847625732,
      "learning_rate": 0.0001450404,
      "loss": 0.2812,
      "step": 137400
    },
    {
      "epoch": 0.035,
      "grad_norm": 0.17764481902122498,
      "learning_rate": 0.0001450004,
      "loss": 0.2773,
      "step": 137500
    },
    {
      "epoch": 0.0352,
      "grad_norm": 0.33276671171188354,
      "learning_rate": 0.0001449604,
      "loss": 0.2829,
      "step": 137600
    },
    {
      "epoch": 0.0354,
      "grad_norm": 0.2309039831161499,
      "learning_rate": 0.0001449204,
      "loss": 0.2744,
      "step": 137700
    },
    {
      "epoch": 0.0356,
      "grad_norm": 0.21504448354244232,
      "learning_rate": 0.0001448804,
      "loss": 0.2975,
      "step": 137800
    },
    {
      "epoch": 0.0358,
      "grad_norm": 0.23908798396587372,
      "learning_rate": 0.0001448404,
      "loss": 0.2891,
      "step": 137900
    },
    {
      "epoch": 0.036,
      "grad_norm": 0.18185609579086304,
      "learning_rate": 0.00014480040000000002,
      "loss": 0.2821,
      "step": 138000
    },
    {
      "epoch": 0.0362,
      "grad_norm": 0.3852616548538208,
      "learning_rate": 0.0001447604,
      "loss": 0.2854,
      "step": 138100
    },
    {
      "epoch": 0.0364,
      "grad_norm": 0.2340957075357437,
      "learning_rate": 0.0001447204,
      "loss": 0.2819,
      "step": 138200
    },
    {
      "epoch": 0.0366,
      "grad_norm": 0.23798440396785736,
      "learning_rate": 0.0001446804,
      "loss": 0.2838,
      "step": 138300
    },
    {
      "epoch": 0.0368,
      "grad_norm": 0.2655554413795471,
      "learning_rate": 0.0001446404,
      "loss": 0.2847,
      "step": 138400
    },
    {
      "epoch": 0.037,
      "grad_norm": 0.20889754593372345,
      "learning_rate": 0.00014460040000000002,
      "loss": 0.2906,
      "step": 138500
    },
    {
      "epoch": 0.0372,
      "grad_norm": 0.2111879587173462,
      "learning_rate": 0.00014456039999999999,
      "loss": 0.2766,
      "step": 138600
    },
    {
      "epoch": 0.0374,
      "grad_norm": 1.0969158411026,
      "learning_rate": 0.0001445204,
      "loss": 0.2879,
      "step": 138700
    },
    {
      "epoch": 0.0376,
      "grad_norm": 0.20623628795146942,
      "learning_rate": 0.0001444804,
      "loss": 0.288,
      "step": 138800
    },
    {
      "epoch": 0.0378,
      "grad_norm": 0.23224857449531555,
      "learning_rate": 0.00014444040000000002,
      "loss": 0.2794,
      "step": 138900
    },
    {
      "epoch": 0.038,
      "grad_norm": 0.20052120089530945,
      "learning_rate": 0.00014440040000000001,
      "loss": 0.2779,
      "step": 139000
    },
    {
      "epoch": 0.0382,
      "grad_norm": 0.21525472402572632,
      "learning_rate": 0.0001443604,
      "loss": 0.2861,
      "step": 139100
    },
    {
      "epoch": 0.0384,
      "grad_norm": 0.155807763338089,
      "learning_rate": 0.0001443204,
      "loss": 0.2868,
      "step": 139200
    },
    {
      "epoch": 0.0386,
      "grad_norm": 0.3168600797653198,
      "learning_rate": 0.0001442804,
      "loss": 0.2789,
      "step": 139300
    },
    {
      "epoch": 0.0388,
      "grad_norm": 0.23451729118824005,
      "learning_rate": 0.00014424040000000002,
      "loss": 0.2891,
      "step": 139400
    },
    {
      "epoch": 0.039,
      "grad_norm": 0.29926997423171997,
      "learning_rate": 0.0001442004,
      "loss": 0.2977,
      "step": 139500
    },
    {
      "epoch": 0.0392,
      "grad_norm": 0.27708712220191956,
      "learning_rate": 0.00014416040000000003,
      "loss": 0.2897,
      "step": 139600
    },
    {
      "epoch": 0.0394,
      "grad_norm": 0.17544563114643097,
      "learning_rate": 0.0001441204,
      "loss": 0.282,
      "step": 139700
    },
    {
      "epoch": 0.0396,
      "grad_norm": 0.24782776832580566,
      "learning_rate": 0.00014408040000000002,
      "loss": 0.2789,
      "step": 139800
    },
    {
      "epoch": 0.0398,
      "grad_norm": 0.23590870201587677,
      "learning_rate": 0.0001440404,
      "loss": 0.2794,
      "step": 139900
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.17239037156105042,
      "learning_rate": 0.0001440004,
      "loss": 0.2843,
      "step": 140000
    },
    {
      "epoch": 0.0402,
      "grad_norm": 0.22361674904823303,
      "learning_rate": 0.00014396040000000003,
      "loss": 0.2794,
      "step": 140100
    },
    {
      "epoch": 0.0404,
      "grad_norm": 0.3022303581237793,
      "learning_rate": 0.0001439204,
      "loss": 0.3019,
      "step": 140200
    },
    {
      "epoch": 0.0406,
      "grad_norm": 0.2557013928890228,
      "learning_rate": 0.0001438804,
      "loss": 0.2996,
      "step": 140300
    },
    {
      "epoch": 0.0408,
      "grad_norm": 0.278455913066864,
      "learning_rate": 0.0001438404,
      "loss": 0.2935,
      "step": 140400
    },
    {
      "epoch": 0.041,
      "grad_norm": 0.37940022349357605,
      "learning_rate": 0.0001438004,
      "loss": 0.2886,
      "step": 140500
    },
    {
      "epoch": 0.0412,
      "grad_norm": 0.34311047196388245,
      "learning_rate": 0.00014376040000000002,
      "loss": 0.2892,
      "step": 140600
    },
    {
      "epoch": 0.0414,
      "grad_norm": 0.3010990619659424,
      "learning_rate": 0.0001437204,
      "loss": 0.2841,
      "step": 140700
    },
    {
      "epoch": 0.0416,
      "grad_norm": 0.19272881746292114,
      "learning_rate": 0.0001436804,
      "loss": 0.3048,
      "step": 140800
    },
    {
      "epoch": 0.0418,
      "grad_norm": 0.20627625286579132,
      "learning_rate": 0.0001436404,
      "loss": 0.295,
      "step": 140900
    },
    {
      "epoch": 0.042,
      "grad_norm": 0.4096256494522095,
      "learning_rate": 0.00014360040000000002,
      "loss": 0.296,
      "step": 141000
    },
    {
      "epoch": 0.0422,
      "grad_norm": 0.20318618416786194,
      "learning_rate": 0.00014356040000000002,
      "loss": 0.2917,
      "step": 141100
    },
    {
      "epoch": 0.0424,
      "grad_norm": 0.5557262897491455,
      "learning_rate": 0.0001435204,
      "loss": 0.2922,
      "step": 141200
    },
    {
      "epoch": 0.0426,
      "grad_norm": 0.2203870415687561,
      "learning_rate": 0.0001434804,
      "loss": 0.2947,
      "step": 141300
    },
    {
      "epoch": 0.0428,
      "grad_norm": 0.23622627556324005,
      "learning_rate": 0.0001434404,
      "loss": 0.2935,
      "step": 141400
    },
    {
      "epoch": 0.043,
      "grad_norm": 0.24292029440402985,
      "learning_rate": 0.00014340040000000002,
      "loss": 0.2956,
      "step": 141500
    },
    {
      "epoch": 0.0432,
      "grad_norm": 0.18479755520820618,
      "learning_rate": 0.0001433604,
      "loss": 0.2997,
      "step": 141600
    },
    {
      "epoch": 0.0434,
      "grad_norm": 0.2119598090648651,
      "learning_rate": 0.0001433204,
      "loss": 0.2957,
      "step": 141700
    },
    {
      "epoch": 0.0436,
      "grad_norm": 0.2696900963783264,
      "learning_rate": 0.0001432804,
      "loss": 0.293,
      "step": 141800
    },
    {
      "epoch": 0.0438,
      "grad_norm": 0.17380273342132568,
      "learning_rate": 0.00014324040000000002,
      "loss": 0.2979,
      "step": 141900
    },
    {
      "epoch": 0.044,
      "grad_norm": 0.20531192421913147,
      "learning_rate": 0.0001432004,
      "loss": 0.2969,
      "step": 142000
    },
    {
      "epoch": 0.0442,
      "grad_norm": 0.2422923445701599,
      "learning_rate": 0.0001431604,
      "loss": 0.3016,
      "step": 142100
    },
    {
      "epoch": 0.0444,
      "grad_norm": 0.16883300244808197,
      "learning_rate": 0.0001431204,
      "loss": 0.2839,
      "step": 142200
    },
    {
      "epoch": 0.0446,
      "grad_norm": 0.1872025579214096,
      "learning_rate": 0.0001430804,
      "loss": 0.3367,
      "step": 142300
    },
    {
      "epoch": 0.0448,
      "grad_norm": 0.18559499084949493,
      "learning_rate": 0.00014304040000000001,
      "loss": 0.2872,
      "step": 142400
    },
    {
      "epoch": 0.045,
      "grad_norm": 0.22933410108089447,
      "learning_rate": 0.0001430004,
      "loss": 0.2856,
      "step": 142500
    },
    {
      "epoch": 0.0452,
      "grad_norm": 0.1616288274526596,
      "learning_rate": 0.00014296040000000003,
      "loss": 0.2832,
      "step": 142600
    },
    {
      "epoch": 0.0454,
      "grad_norm": 0.190198615193367,
      "learning_rate": 0.0001429204,
      "loss": 0.2811,
      "step": 142700
    },
    {
      "epoch": 0.0456,
      "grad_norm": 0.2478422075510025,
      "learning_rate": 0.00014288040000000002,
      "loss": 0.2808,
      "step": 142800
    },
    {
      "epoch": 0.0458,
      "grad_norm": 0.3362259566783905,
      "learning_rate": 0.0001428404,
      "loss": 0.285,
      "step": 142900
    },
    {
      "epoch": 0.046,
      "grad_norm": 0.1876300573348999,
      "learning_rate": 0.0001428004,
      "loss": 0.2844,
      "step": 143000
    },
    {
      "epoch": 0.0462,
      "grad_norm": 0.267817884683609,
      "learning_rate": 0.00014276040000000002,
      "loss": 0.2967,
      "step": 143100
    },
    {
      "epoch": 0.0464,
      "grad_norm": 0.2802794277667999,
      "learning_rate": 0.0001427204,
      "loss": 0.3118,
      "step": 143200
    },
    {
      "epoch": 0.0466,
      "grad_norm": 0.27438464760780334,
      "learning_rate": 0.0001426804,
      "loss": 0.2868,
      "step": 143300
    },
    {
      "epoch": 0.0468,
      "grad_norm": 0.22925788164138794,
      "learning_rate": 0.0001426404,
      "loss": 0.2945,
      "step": 143400
    },
    {
      "epoch": 0.047,
      "grad_norm": 0.17650224268436432,
      "learning_rate": 0.00014260040000000002,
      "loss": 0.2908,
      "step": 143500
    },
    {
      "epoch": 0.0472,
      "grad_norm": 0.1927453577518463,
      "learning_rate": 0.00014256040000000002,
      "loss": 0.2882,
      "step": 143600
    },
    {
      "epoch": 0.0474,
      "grad_norm": 0.2470424622297287,
      "learning_rate": 0.00014252039999999998,
      "loss": 0.2918,
      "step": 143700
    },
    {
      "epoch": 0.0476,
      "grad_norm": 0.2224511355161667,
      "learning_rate": 0.0001424804,
      "loss": 0.2924,
      "step": 143800
    },
    {
      "epoch": 0.0478,
      "grad_norm": 0.25327613949775696,
      "learning_rate": 0.0001424404,
      "loss": 0.2966,
      "step": 143900
    },
    {
      "epoch": 0.048,
      "grad_norm": 0.2762177288532257,
      "learning_rate": 0.00014240040000000002,
      "loss": 0.2976,
      "step": 144000
    },
    {
      "epoch": 0.0482,
      "grad_norm": 0.32673609256744385,
      "learning_rate": 0.0001423604,
      "loss": 0.2974,
      "step": 144100
    },
    {
      "epoch": 0.0484,
      "grad_norm": 0.4225119352340698,
      "learning_rate": 0.0001423204,
      "loss": 0.2931,
      "step": 144200
    },
    {
      "epoch": 0.0486,
      "grad_norm": 0.20474287867546082,
      "learning_rate": 0.0001422804,
      "loss": 0.2992,
      "step": 144300
    },
    {
      "epoch": 0.0488,
      "grad_norm": 0.210563525557518,
      "learning_rate": 0.0001422404,
      "loss": 0.2848,
      "step": 144400
    },
    {
      "epoch": 0.049,
      "grad_norm": 0.21854105591773987,
      "learning_rate": 0.00014220040000000001,
      "loss": 0.2965,
      "step": 144500
    },
    {
      "epoch": 0.0492,
      "grad_norm": 0.2694493532180786,
      "learning_rate": 0.0001421604,
      "loss": 0.2847,
      "step": 144600
    },
    {
      "epoch": 0.0494,
      "grad_norm": 0.33580654859542847,
      "learning_rate": 0.0001421204,
      "loss": 0.2971,
      "step": 144700
    },
    {
      "epoch": 0.0496,
      "grad_norm": 0.22318777441978455,
      "learning_rate": 0.0001420804,
      "loss": 0.2951,
      "step": 144800
    },
    {
      "epoch": 0.0498,
      "grad_norm": 0.21913912892341614,
      "learning_rate": 0.00014204040000000002,
      "loss": 0.2943,
      "step": 144900
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.1860969066619873,
      "learning_rate": 0.0001420004,
      "loss": 0.2937,
      "step": 145000
    },
    {
      "epoch": 0.0502,
      "grad_norm": 0.18611373007297516,
      "learning_rate": 0.0001419604,
      "loss": 0.2915,
      "step": 145100
    },
    {
      "epoch": 0.0504,
      "grad_norm": 0.3601659834384918,
      "learning_rate": 0.0001419204,
      "loss": 0.3132,
      "step": 145200
    },
    {
      "epoch": 0.0506,
      "grad_norm": 0.22689419984817505,
      "learning_rate": 0.0001418804,
      "loss": 0.2991,
      "step": 145300
    },
    {
      "epoch": 0.0508,
      "grad_norm": 0.22691553831100464,
      "learning_rate": 0.0001418404,
      "loss": 0.2964,
      "step": 145400
    },
    {
      "epoch": 0.051,
      "grad_norm": 0.18553423881530762,
      "learning_rate": 0.0001418004,
      "loss": 0.2975,
      "step": 145500
    },
    {
      "epoch": 0.0512,
      "grad_norm": 0.21025227010250092,
      "learning_rate": 0.00014176040000000003,
      "loss": 0.2923,
      "step": 145600
    },
    {
      "epoch": 0.0514,
      "grad_norm": 0.17990227043628693,
      "learning_rate": 0.0001417204,
      "loss": 0.3035,
      "step": 145700
    },
    {
      "epoch": 0.0516,
      "grad_norm": 0.18237918615341187,
      "learning_rate": 0.0001416804,
      "loss": 0.3022,
      "step": 145800
    },
    {
      "epoch": 0.0518,
      "grad_norm": 0.20058906078338623,
      "learning_rate": 0.0001416404,
      "loss": 0.2902,
      "step": 145900
    },
    {
      "epoch": 0.052,
      "grad_norm": 0.40441182255744934,
      "learning_rate": 0.0001416004,
      "loss": 0.3083,
      "step": 146000
    },
    {
      "epoch": 0.0522,
      "grad_norm": 0.2583122253417969,
      "learning_rate": 0.00014156040000000002,
      "loss": 0.285,
      "step": 146100
    },
    {
      "epoch": 0.0524,
      "grad_norm": 0.18074502050876617,
      "learning_rate": 0.0001415204,
      "loss": 0.2932,
      "step": 146200
    },
    {
      "epoch": 0.0526,
      "grad_norm": 0.3700363337993622,
      "learning_rate": 0.0001414804,
      "loss": 0.314,
      "step": 146300
    },
    {
      "epoch": 0.0528,
      "grad_norm": 0.20429983735084534,
      "learning_rate": 0.0001414404,
      "loss": 0.2994,
      "step": 146400
    },
    {
      "epoch": 0.053,
      "grad_norm": 0.4449596405029297,
      "learning_rate": 0.00014140040000000002,
      "loss": 0.2922,
      "step": 146500
    },
    {
      "epoch": 0.0532,
      "grad_norm": 0.27035078406333923,
      "learning_rate": 0.00014136040000000002,
      "loss": 0.3025,
      "step": 146600
    },
    {
      "epoch": 0.0534,
      "grad_norm": 0.1620853841304779,
      "learning_rate": 0.0001413204,
      "loss": 0.2883,
      "step": 146700
    },
    {
      "epoch": 0.0536,
      "grad_norm": 0.20872916281223297,
      "learning_rate": 0.0001412804,
      "loss": 0.2899,
      "step": 146800
    },
    {
      "epoch": 0.0538,
      "grad_norm": 0.171489417552948,
      "learning_rate": 0.0001412404,
      "loss": 0.293,
      "step": 146900
    },
    {
      "epoch": 0.054,
      "grad_norm": 0.19745928049087524,
      "learning_rate": 0.00014120040000000002,
      "loss": 0.2898,
      "step": 147000
    },
    {
      "epoch": 0.0542,
      "grad_norm": 0.18130256235599518,
      "learning_rate": 0.0001411604,
      "loss": 0.2923,
      "step": 147100
    },
    {
      "epoch": 0.0544,
      "grad_norm": 0.29075926542282104,
      "learning_rate": 0.0001411204,
      "loss": 0.3032,
      "step": 147200
    },
    {
      "epoch": 0.0546,
      "grad_norm": 0.21553675830364227,
      "learning_rate": 0.0001410804,
      "loss": 0.3012,
      "step": 147300
    },
    {
      "epoch": 0.0548,
      "grad_norm": 0.23423010110855103,
      "learning_rate": 0.00014104040000000002,
      "loss": 0.2911,
      "step": 147400
    },
    {
      "epoch": 0.055,
      "grad_norm": 0.2579517364501953,
      "learning_rate": 0.0001410004,
      "loss": 0.292,
      "step": 147500
    },
    {
      "epoch": 0.0552,
      "grad_norm": 0.23888039588928223,
      "learning_rate": 0.0001409604,
      "loss": 0.3001,
      "step": 147600
    },
    {
      "epoch": 0.0554,
      "grad_norm": 0.32866355776786804,
      "learning_rate": 0.0001409204,
      "loss": 0.318,
      "step": 147700
    },
    {
      "epoch": 0.0556,
      "grad_norm": 0.30911415815353394,
      "learning_rate": 0.0001408804,
      "loss": 0.3005,
      "step": 147800
    },
    {
      "epoch": 0.0558,
      "grad_norm": 0.2983124852180481,
      "learning_rate": 0.00014084040000000001,
      "loss": 0.2895,
      "step": 147900
    },
    {
      "epoch": 0.056,
      "grad_norm": 0.18074147403240204,
      "learning_rate": 0.0001408004,
      "loss": 0.3108,
      "step": 148000
    },
    {
      "epoch": 0.0562,
      "grad_norm": 0.17594757676124573,
      "learning_rate": 0.0001407604,
      "loss": 0.2967,
      "step": 148100
    },
    {
      "epoch": 0.0564,
      "grad_norm": 0.21774758398532867,
      "learning_rate": 0.0001407204,
      "loss": 0.299,
      "step": 148200
    },
    {
      "epoch": 0.0566,
      "grad_norm": 0.3890017569065094,
      "learning_rate": 0.0001406804,
      "loss": 0.2822,
      "step": 148300
    },
    {
      "epoch": 0.0568,
      "grad_norm": 0.23971693217754364,
      "learning_rate": 0.0001406404,
      "loss": 0.3042,
      "step": 148400
    },
    {
      "epoch": 0.057,
      "grad_norm": 0.14672812819480896,
      "learning_rate": 0.0001406004,
      "loss": 0.3023,
      "step": 148500
    },
    {
      "epoch": 0.0572,
      "grad_norm": 0.22197787463665009,
      "learning_rate": 0.00014056040000000002,
      "loss": 0.2946,
      "step": 148600
    },
    {
      "epoch": 0.0574,
      "grad_norm": 0.34892770648002625,
      "learning_rate": 0.0001405204,
      "loss": 0.3027,
      "step": 148700
    },
    {
      "epoch": 0.0576,
      "grad_norm": 0.22468386590480804,
      "learning_rate": 0.0001404804,
      "loss": 0.3036,
      "step": 148800
    },
    {
      "epoch": 0.0578,
      "grad_norm": 0.3529873192310333,
      "learning_rate": 0.0001404404,
      "loss": 0.3103,
      "step": 148900
    },
    {
      "epoch": 0.058,
      "grad_norm": 0.22541065514087677,
      "learning_rate": 0.0001404004,
      "loss": 0.3031,
      "step": 149000
    },
    {
      "epoch": 0.0582,
      "grad_norm": 0.19883927702903748,
      "learning_rate": 0.00014036040000000002,
      "loss": 0.2875,
      "step": 149100
    },
    {
      "epoch": 0.0584,
      "grad_norm": 0.16821545362472534,
      "learning_rate": 0.00014032039999999999,
      "loss": 0.2906,
      "step": 149200
    },
    {
      "epoch": 0.0586,
      "grad_norm": 0.2122853845357895,
      "learning_rate": 0.0001402804,
      "loss": 0.294,
      "step": 149300
    },
    {
      "epoch": 0.0588,
      "grad_norm": 0.21868352591991425,
      "learning_rate": 0.0001402404,
      "loss": 0.3014,
      "step": 149400
    },
    {
      "epoch": 0.059,
      "grad_norm": 0.26323187351226807,
      "learning_rate": 0.00014020040000000002,
      "loss": 0.3239,
      "step": 149500
    },
    {
      "epoch": 0.0592,
      "grad_norm": 0.28496459126472473,
      "learning_rate": 0.00014016040000000001,
      "loss": 0.3198,
      "step": 149600
    },
    {
      "epoch": 0.0594,
      "grad_norm": 0.21418221294879913,
      "learning_rate": 0.0001401204,
      "loss": 0.3037,
      "step": 149700
    },
    {
      "epoch": 0.0596,
      "grad_norm": 0.18050341308116913,
      "learning_rate": 0.0001400804,
      "loss": 0.2864,
      "step": 149800
    },
    {
      "epoch": 0.0598,
      "grad_norm": 0.18909691274166107,
      "learning_rate": 0.0001400404,
      "loss": 0.2765,
      "step": 149900
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.1939447522163391,
      "learning_rate": 0.00014000040000000002,
      "loss": 0.2759,
      "step": 150000
    },
    {
      "epoch": 0.0602,
      "grad_norm": 0.2437305450439453,
      "learning_rate": 0.0001399604,
      "loss": 0.2684,
      "step": 150100
    },
    {
      "epoch": 0.0604,
      "grad_norm": 0.18825781345367432,
      "learning_rate": 0.0001399204,
      "loss": 0.2675,
      "step": 150200
    },
    {
      "epoch": 0.0606,
      "grad_norm": 0.17045961320400238,
      "learning_rate": 0.0001398804,
      "loss": 0.2668,
      "step": 150300
    },
    {
      "epoch": 0.0608,
      "grad_norm": 0.16906027495861053,
      "learning_rate": 0.00013984040000000002,
      "loss": 0.2641,
      "step": 150400
    },
    {
      "epoch": 0.061,
      "grad_norm": 0.19279079139232635,
      "learning_rate": 0.0001398004,
      "loss": 0.2656,
      "step": 150500
    },
    {
      "epoch": 0.0612,
      "grad_norm": 0.16379393637180328,
      "learning_rate": 0.0001397604,
      "loss": 0.2643,
      "step": 150600
    },
    {
      "epoch": 0.0614,
      "grad_norm": 0.20136988162994385,
      "learning_rate": 0.0001397204,
      "loss": 0.2644,
      "step": 150700
    },
    {
      "epoch": 0.0616,
      "grad_norm": 0.16851016879081726,
      "learning_rate": 0.0001396804,
      "loss": 0.2688,
      "step": 150800
    },
    {
      "epoch": 0.0618,
      "grad_norm": 0.3253178596496582,
      "learning_rate": 0.0001396404,
      "loss": 0.2745,
      "step": 150900
    },
    {
      "epoch": 0.062,
      "grad_norm": 0.22693248093128204,
      "learning_rate": 0.0001396004,
      "loss": 0.264,
      "step": 151000
    },
    {
      "epoch": 0.0622,
      "grad_norm": 0.16674602031707764,
      "learning_rate": 0.00013956040000000003,
      "loss": 0.2702,
      "step": 151100
    },
    {
      "epoch": 0.0624,
      "grad_norm": 0.17631779611110687,
      "learning_rate": 0.00013952040000000002,
      "loss": 0.2707,
      "step": 151200
    },
    {
      "epoch": 0.0626,
      "grad_norm": 0.18675994873046875,
      "learning_rate": 0.0001394804,
      "loss": 0.2697,
      "step": 151300
    },
    {
      "epoch": 0.0628,
      "grad_norm": 0.19959695637226105,
      "learning_rate": 0.0001394404,
      "loss": 0.2674,
      "step": 151400
    },
    {
      "epoch": 0.063,
      "grad_norm": 0.1849915087223053,
      "learning_rate": 0.0001394004,
      "loss": 0.2662,
      "step": 151500
    },
    {
      "epoch": 0.0632,
      "grad_norm": 0.9266983866691589,
      "learning_rate": 0.00013936040000000002,
      "loss": 0.2693,
      "step": 151600
    },
    {
      "epoch": 0.0634,
      "grad_norm": 0.1627340465784073,
      "learning_rate": 0.00013932040000000002,
      "loss": 0.2692,
      "step": 151700
    },
    {
      "epoch": 0.0636,
      "grad_norm": 0.20373468101024628,
      "learning_rate": 0.0001392804,
      "loss": 0.263,
      "step": 151800
    },
    {
      "epoch": 0.0638,
      "grad_norm": 0.18756113946437836,
      "learning_rate": 0.0001392404,
      "loss": 0.2629,
      "step": 151900
    },
    {
      "epoch": 0.064,
      "grad_norm": 0.21463698148727417,
      "learning_rate": 0.0001392004,
      "loss": 0.2727,
      "step": 152000
    },
    {
      "epoch": 0.0642,
      "grad_norm": 0.18175888061523438,
      "learning_rate": 0.00013916040000000002,
      "loss": 0.2732,
      "step": 152100
    },
    {
      "epoch": 0.0644,
      "grad_norm": 0.19156472384929657,
      "learning_rate": 0.0001391204,
      "loss": 0.2681,
      "step": 152200
    },
    {
      "epoch": 0.0646,
      "grad_norm": 0.1933964043855667,
      "learning_rate": 0.0001390804,
      "loss": 0.2669,
      "step": 152300
    },
    {
      "epoch": 0.0648,
      "grad_norm": 0.16370782256126404,
      "learning_rate": 0.0001390404,
      "loss": 0.2637,
      "step": 152400
    },
    {
      "epoch": 0.065,
      "grad_norm": 0.1704418957233429,
      "learning_rate": 0.00013900040000000002,
      "loss": 0.2651,
      "step": 152500
    },
    {
      "epoch": 0.0652,
      "grad_norm": 0.16683322191238403,
      "learning_rate": 0.0001389604,
      "loss": 0.2704,
      "step": 152600
    },
    {
      "epoch": 0.0654,
      "grad_norm": 0.2169724553823471,
      "learning_rate": 0.0001389204,
      "loss": 0.2701,
      "step": 152700
    },
    {
      "epoch": 0.0656,
      "grad_norm": 0.16984739899635315,
      "learning_rate": 0.0001388804,
      "loss": 0.267,
      "step": 152800
    },
    {
      "epoch": 0.0658,
      "grad_norm": 0.1774764508008957,
      "learning_rate": 0.0001388404,
      "loss": 0.2694,
      "step": 152900
    },
    {
      "epoch": 0.066,
      "grad_norm": 0.25803592801094055,
      "learning_rate": 0.00013880040000000001,
      "loss": 0.2692,
      "step": 153000
    },
    {
      "epoch": 0.0662,
      "grad_norm": 0.16305796802043915,
      "learning_rate": 0.0001387604,
      "loss": 0.2666,
      "step": 153100
    },
    {
      "epoch": 0.0664,
      "grad_norm": 0.28455424308776855,
      "learning_rate": 0.00013872040000000003,
      "loss": 0.2692,
      "step": 153200
    },
    {
      "epoch": 0.0666,
      "grad_norm": 0.527635931968689,
      "learning_rate": 0.0001386804,
      "loss": 0.2709,
      "step": 153300
    },
    {
      "epoch": 0.0668,
      "grad_norm": 0.15370622277259827,
      "learning_rate": 0.00013864040000000002,
      "loss": 0.2709,
      "step": 153400
    },
    {
      "epoch": 0.067,
      "grad_norm": 0.17640765011310577,
      "learning_rate": 0.0001386004,
      "loss": 0.271,
      "step": 153500
    },
    {
      "epoch": 0.0672,
      "grad_norm": 0.19771219789981842,
      "learning_rate": 0.0001385604,
      "loss": 0.2709,
      "step": 153600
    },
    {
      "epoch": 0.0674,
      "grad_norm": 0.1551404446363449,
      "learning_rate": 0.00013852040000000002,
      "loss": 0.2661,
      "step": 153700
    },
    {
      "epoch": 0.0676,
      "grad_norm": 0.17845940589904785,
      "learning_rate": 0.0001384804,
      "loss": 0.2696,
      "step": 153800
    },
    {
      "epoch": 0.0678,
      "grad_norm": 0.1720665991306305,
      "learning_rate": 0.0001384404,
      "loss": 0.2648,
      "step": 153900
    },
    {
      "epoch": 0.068,
      "grad_norm": 0.1719100922346115,
      "learning_rate": 0.0001384004,
      "loss": 0.2688,
      "step": 154000
    },
    {
      "epoch": 0.0682,
      "grad_norm": 0.16675078868865967,
      "learning_rate": 0.00013836040000000002,
      "loss": 0.274,
      "step": 154100
    },
    {
      "epoch": 0.0684,
      "grad_norm": 0.21262894570827484,
      "learning_rate": 0.00013832040000000002,
      "loss": 0.2699,
      "step": 154200
    },
    {
      "epoch": 0.0686,
      "grad_norm": 0.21415004134178162,
      "learning_rate": 0.0001382804,
      "loss": 0.2634,
      "step": 154300
    },
    {
      "epoch": 0.0688,
      "grad_norm": 0.16432279348373413,
      "learning_rate": 0.0001382404,
      "loss": 0.2671,
      "step": 154400
    },
    {
      "epoch": 0.069,
      "grad_norm": 0.2584374248981476,
      "learning_rate": 0.0001382004,
      "loss": 0.2731,
      "step": 154500
    },
    {
      "epoch": 0.0692,
      "grad_norm": 0.1635134071111679,
      "learning_rate": 0.00013816040000000002,
      "loss": 0.2686,
      "step": 154600
    },
    {
      "epoch": 0.0694,
      "grad_norm": 0.23434379696846008,
      "learning_rate": 0.0001381204,
      "loss": 0.2749,
      "step": 154700
    },
    {
      "epoch": 0.0696,
      "grad_norm": 0.16734665632247925,
      "learning_rate": 0.0001380804,
      "loss": 0.2663,
      "step": 154800
    },
    {
      "epoch": 0.0698,
      "grad_norm": 0.17818975448608398,
      "learning_rate": 0.0001380404,
      "loss": 0.2702,
      "step": 154900
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.17465141415596008,
      "learning_rate": 0.0001380004,
      "loss": 0.2728,
      "step": 155000
    },
    {
      "epoch": 0.0702,
      "grad_norm": 0.24960975348949432,
      "learning_rate": 0.00013796040000000001,
      "loss": 0.2663,
      "step": 155100
    },
    {
      "epoch": 0.0704,
      "grad_norm": 0.22631630301475525,
      "learning_rate": 0.0001379204,
      "loss": 0.2728,
      "step": 155200
    },
    {
      "epoch": 0.0706,
      "grad_norm": 0.35504263639450073,
      "learning_rate": 0.0001378804,
      "loss": 0.2704,
      "step": 155300
    },
    {
      "epoch": 0.0708,
      "grad_norm": 0.19621668756008148,
      "learning_rate": 0.0001378404,
      "loss": 0.2723,
      "step": 155400
    },
    {
      "epoch": 0.071,
      "grad_norm": 0.14076755940914154,
      "learning_rate": 0.00013780040000000002,
      "loss": 0.2676,
      "step": 155500
    },
    {
      "epoch": 0.0712,
      "grad_norm": 0.19740091264247894,
      "learning_rate": 0.0001377604,
      "loss": 0.2654,
      "step": 155600
    },
    {
      "epoch": 0.0714,
      "grad_norm": 0.21776890754699707,
      "learning_rate": 0.0001377204,
      "loss": 0.2648,
      "step": 155700
    },
    {
      "epoch": 0.0716,
      "grad_norm": 0.19757398962974548,
      "learning_rate": 0.0001376804,
      "loss": 0.2706,
      "step": 155800
    },
    {
      "epoch": 0.0718,
      "grad_norm": 0.17804448306560516,
      "learning_rate": 0.0001376404,
      "loss": 0.2692,
      "step": 155900
    },
    {
      "epoch": 0.072,
      "grad_norm": 0.28381094336509705,
      "learning_rate": 0.0001376004,
      "loss": 0.2658,
      "step": 156000
    },
    {
      "epoch": 0.0722,
      "grad_norm": 0.19585195183753967,
      "learning_rate": 0.0001375604,
      "loss": 0.27,
      "step": 156100
    },
    {
      "epoch": 0.0724,
      "grad_norm": 0.1824631541967392,
      "learning_rate": 0.00013752040000000003,
      "loss": 0.2726,
      "step": 156200
    },
    {
      "epoch": 0.0726,
      "grad_norm": 0.16430000960826874,
      "learning_rate": 0.0001374804,
      "loss": 0.2702,
      "step": 156300
    },
    {
      "epoch": 0.0728,
      "grad_norm": 0.14714659750461578,
      "learning_rate": 0.0001374404,
      "loss": 0.2704,
      "step": 156400
    },
    {
      "epoch": 0.073,
      "grad_norm": 0.2480650097131729,
      "learning_rate": 0.0001374004,
      "loss": 0.2651,
      "step": 156500
    },
    {
      "epoch": 0.0732,
      "grad_norm": 0.1619512140750885,
      "learning_rate": 0.0001373604,
      "loss": 0.2661,
      "step": 156600
    },
    {
      "epoch": 0.0734,
      "grad_norm": 0.16349893808364868,
      "learning_rate": 0.00013732040000000002,
      "loss": 0.2684,
      "step": 156700
    },
    {
      "epoch": 0.0736,
      "grad_norm": 0.31547239422798157,
      "learning_rate": 0.0001372804,
      "loss": 0.267,
      "step": 156800
    },
    {
      "epoch": 0.0738,
      "grad_norm": 0.17574888467788696,
      "learning_rate": 0.0001372404,
      "loss": 0.2712,
      "step": 156900
    },
    {
      "epoch": 0.074,
      "grad_norm": 0.21195268630981445,
      "learning_rate": 0.0001372004,
      "loss": 0.2689,
      "step": 157000
    },
    {
      "epoch": 0.0742,
      "grad_norm": 0.16163906455039978,
      "learning_rate": 0.00013716040000000002,
      "loss": 0.2696,
      "step": 157100
    },
    {
      "epoch": 0.0744,
      "grad_norm": 0.22510552406311035,
      "learning_rate": 0.00013712040000000002,
      "loss": 0.2691,
      "step": 157200
    },
    {
      "epoch": 0.0746,
      "grad_norm": 0.27251046895980835,
      "learning_rate": 0.0001370804,
      "loss": 0.2678,
      "step": 157300
    },
    {
      "epoch": 0.0748,
      "grad_norm": 0.24684669077396393,
      "learning_rate": 0.0001370404,
      "loss": 0.267,
      "step": 157400
    },
    {
      "epoch": 0.075,
      "grad_norm": 0.2159154713153839,
      "learning_rate": 0.0001370004,
      "loss": 0.2792,
      "step": 157500
    },
    {
      "epoch": 0.0752,
      "grad_norm": 0.16219884157180786,
      "learning_rate": 0.00013696040000000002,
      "loss": 0.2714,
      "step": 157600
    },
    {
      "epoch": 0.0754,
      "grad_norm": 0.28837162256240845,
      "learning_rate": 0.0001369204,
      "loss": 0.2668,
      "step": 157700
    },
    {
      "epoch": 0.0756,
      "grad_norm": 0.23692043125629425,
      "learning_rate": 0.0001368804,
      "loss": 0.2673,
      "step": 157800
    },
    {
      "epoch": 0.0758,
      "grad_norm": 0.1836722046136856,
      "learning_rate": 0.0001368404,
      "loss": 0.2685,
      "step": 157900
    },
    {
      "epoch": 0.076,
      "grad_norm": 0.18108327686786652,
      "learning_rate": 0.00013680040000000002,
      "loss": 0.2725,
      "step": 158000
    },
    {
      "epoch": 0.0762,
      "grad_norm": 0.14798210561275482,
      "learning_rate": 0.0001367604,
      "loss": 0.2682,
      "step": 158100
    },
    {
      "epoch": 0.0764,
      "grad_norm": 0.3181903064250946,
      "learning_rate": 0.0001367204,
      "loss": 0.2608,
      "step": 158200
    },
    {
      "epoch": 0.0766,
      "grad_norm": 0.15039204061031342,
      "learning_rate": 0.0001366804,
      "loss": 0.2676,
      "step": 158300
    },
    {
      "epoch": 0.0768,
      "grad_norm": 0.18739774823188782,
      "learning_rate": 0.0001366404,
      "loss": 0.269,
      "step": 158400
    },
    {
      "epoch": 0.077,
      "grad_norm": 0.28927820920944214,
      "learning_rate": 0.00013660040000000001,
      "loss": 0.2656,
      "step": 158500
    },
    {
      "epoch": 0.0772,
      "grad_norm": 0.174117773771286,
      "learning_rate": 0.0001365604,
      "loss": 0.2639,
      "step": 158600
    },
    {
      "epoch": 0.0774,
      "grad_norm": 0.20544083416461945,
      "learning_rate": 0.0001365204,
      "loss": 0.2686,
      "step": 158700
    },
    {
      "epoch": 0.0776,
      "grad_norm": 0.219607412815094,
      "learning_rate": 0.0001364804,
      "loss": 0.2665,
      "step": 158800
    },
    {
      "epoch": 0.0778,
      "grad_norm": 0.16388162970542908,
      "learning_rate": 0.0001364404,
      "loss": 0.2676,
      "step": 158900
    },
    {
      "epoch": 0.078,
      "grad_norm": 0.2054087072610855,
      "learning_rate": 0.0001364004,
      "loss": 0.2767,
      "step": 159000
    },
    {
      "epoch": 0.0782,
      "grad_norm": 0.15745876729488373,
      "learning_rate": 0.0001363604,
      "loss": 0.2713,
      "step": 159100
    },
    {
      "epoch": 0.0784,
      "grad_norm": 0.17080196738243103,
      "learning_rate": 0.00013632040000000002,
      "loss": 0.2622,
      "step": 159200
    },
    {
      "epoch": 0.0786,
      "grad_norm": 0.2554245591163635,
      "learning_rate": 0.0001362804,
      "loss": 0.2658,
      "step": 159300
    },
    {
      "epoch": 0.0788,
      "grad_norm": 0.25696536898612976,
      "learning_rate": 0.0001362404,
      "loss": 0.2684,
      "step": 159400
    },
    {
      "epoch": 0.079,
      "grad_norm": 0.15896977484226227,
      "learning_rate": 0.0001362004,
      "loss": 0.2698,
      "step": 159500
    },
    {
      "epoch": 0.0792,
      "grad_norm": 0.17716248333454132,
      "learning_rate": 0.0001361604,
      "loss": 0.2713,
      "step": 159600
    },
    {
      "epoch": 0.0794,
      "grad_norm": 0.2042681723833084,
      "learning_rate": 0.00013612040000000002,
      "loss": 0.272,
      "step": 159700
    },
    {
      "epoch": 0.0796,
      "grad_norm": 0.21027176082134247,
      "learning_rate": 0.00013608039999999999,
      "loss": 0.2701,
      "step": 159800
    },
    {
      "epoch": 0.0798,
      "grad_norm": 0.17798981070518494,
      "learning_rate": 0.0001360404,
      "loss": 0.2722,
      "step": 159900
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.21452707052230835,
      "learning_rate": 0.0001360004,
      "loss": 0.2704,
      "step": 160000
    },
    {
      "epoch": 0.0802,
      "grad_norm": 0.19280506670475006,
      "learning_rate": 0.00013596040000000002,
      "loss": 0.2714,
      "step": 160100
    },
    {
      "epoch": 0.0804,
      "grad_norm": 0.1743934005498886,
      "learning_rate": 0.00013592040000000001,
      "loss": 0.2681,
      "step": 160200
    },
    {
      "epoch": 0.0806,
      "grad_norm": 0.20632795989513397,
      "learning_rate": 0.0001358804,
      "loss": 0.2767,
      "step": 160300
    },
    {
      "epoch": 0.0808,
      "grad_norm": 0.18461468815803528,
      "learning_rate": 0.0001358404,
      "loss": 0.2716,
      "step": 160400
    },
    {
      "epoch": 0.081,
      "grad_norm": 0.18317373096942902,
      "learning_rate": 0.0001358004,
      "loss": 0.2742,
      "step": 160500
    },
    {
      "epoch": 0.0812,
      "grad_norm": 0.18154622614383698,
      "learning_rate": 0.00013576040000000002,
      "loss": 0.2647,
      "step": 160600
    },
    {
      "epoch": 0.0814,
      "grad_norm": 0.20829536020755768,
      "learning_rate": 0.0001357204,
      "loss": 0.2679,
      "step": 160700
    },
    {
      "epoch": 0.0816,
      "grad_norm": 0.19264093041419983,
      "learning_rate": 0.0001356804,
      "loss": 0.2655,
      "step": 160800
    },
    {
      "epoch": 0.0818,
      "grad_norm": 0.22138293087482452,
      "learning_rate": 0.0001356404,
      "loss": 0.272,
      "step": 160900
    },
    {
      "epoch": 0.082,
      "grad_norm": 0.18807452917099,
      "learning_rate": 0.00013560040000000002,
      "loss": 0.265,
      "step": 161000
    },
    {
      "epoch": 0.0822,
      "grad_norm": 0.1932212859392166,
      "learning_rate": 0.0001355604,
      "loss": 0.259,
      "step": 161100
    },
    {
      "epoch": 0.0824,
      "grad_norm": 0.1871035099029541,
      "learning_rate": 0.0001355204,
      "loss": 0.2631,
      "step": 161200
    },
    {
      "epoch": 0.0826,
      "grad_norm": 0.16003726422786713,
      "learning_rate": 0.0001354804,
      "loss": 0.2654,
      "step": 161300
    },
    {
      "epoch": 0.0828,
      "grad_norm": 0.1761871874332428,
      "learning_rate": 0.0001354404,
      "loss": 0.2669,
      "step": 161400
    },
    {
      "epoch": 0.083,
      "grad_norm": 0.27383115887641907,
      "learning_rate": 0.0001354004,
      "loss": 0.2621,
      "step": 161500
    },
    {
      "epoch": 0.0832,
      "grad_norm": 0.1668674498796463,
      "learning_rate": 0.0001353604,
      "loss": 0.2665,
      "step": 161600
    },
    {
      "epoch": 0.0834,
      "grad_norm": 0.19219066202640533,
      "learning_rate": 0.00013532040000000003,
      "loss": 0.2672,
      "step": 161700
    },
    {
      "epoch": 0.0836,
      "grad_norm": 0.2823545038700104,
      "learning_rate": 0.0001352804,
      "loss": 0.2655,
      "step": 161800
    },
    {
      "epoch": 0.0838,
      "grad_norm": 0.18881294131278992,
      "learning_rate": 0.00013524040000000001,
      "loss": 0.266,
      "step": 161900
    },
    {
      "epoch": 0.084,
      "grad_norm": 0.1976308971643448,
      "learning_rate": 0.0001352004,
      "loss": 0.2691,
      "step": 162000
    },
    {
      "epoch": 0.0842,
      "grad_norm": 0.22563625872135162,
      "learning_rate": 0.0001351604,
      "loss": 0.2655,
      "step": 162100
    },
    {
      "epoch": 0.0844,
      "grad_norm": 0.18993031978607178,
      "learning_rate": 0.00013512040000000002,
      "loss": 0.269,
      "step": 162200
    },
    {
      "epoch": 0.0846,
      "grad_norm": 0.2097342163324356,
      "learning_rate": 0.0001350804,
      "loss": 0.2652,
      "step": 162300
    },
    {
      "epoch": 0.0848,
      "grad_norm": 0.1664929836988449,
      "learning_rate": 0.0001350404,
      "loss": 0.2654,
      "step": 162400
    },
    {
      "epoch": 0.085,
      "grad_norm": 0.1798952966928482,
      "learning_rate": 0.0001350004,
      "loss": 0.2661,
      "step": 162500
    },
    {
      "epoch": 0.0852,
      "grad_norm": 0.17582497000694275,
      "learning_rate": 0.0001349604,
      "loss": 0.2651,
      "step": 162600
    },
    {
      "epoch": 0.0854,
      "grad_norm": 0.28709277510643005,
      "learning_rate": 0.00013492040000000002,
      "loss": 0.2669,
      "step": 162700
    },
    {
      "epoch": 0.0856,
      "grad_norm": 0.1520322561264038,
      "learning_rate": 0.00013488039999999998,
      "loss": 0.2652,
      "step": 162800
    },
    {
      "epoch": 0.0858,
      "grad_norm": 0.1492086499929428,
      "learning_rate": 0.0001348404,
      "loss": 0.2667,
      "step": 162900
    },
    {
      "epoch": 0.086,
      "grad_norm": 0.25710374116897583,
      "learning_rate": 0.0001348004,
      "loss": 0.2697,
      "step": 163000
    },
    {
      "epoch": 0.0862,
      "grad_norm": 0.20499125123023987,
      "learning_rate": 0.00013476040000000002,
      "loss": 0.264,
      "step": 163100
    },
    {
      "epoch": 0.0864,
      "grad_norm": 0.19236227869987488,
      "learning_rate": 0.0001347204,
      "loss": 0.2692,
      "step": 163200
    },
    {
      "epoch": 0.0866,
      "grad_norm": 0.1704840511083603,
      "learning_rate": 0.0001346804,
      "loss": 0.2857,
      "step": 163300
    },
    {
      "epoch": 0.0868,
      "grad_norm": 0.3140028417110443,
      "learning_rate": 0.0001346404,
      "loss": 0.2689,
      "step": 163400
    },
    {
      "epoch": 0.087,
      "grad_norm": 0.16201360523700714,
      "learning_rate": 0.0001346004,
      "loss": 0.2646,
      "step": 163500
    },
    {
      "epoch": 0.0872,
      "grad_norm": 0.22757022082805634,
      "learning_rate": 0.00013456040000000001,
      "loss": 0.266,
      "step": 163600
    },
    {
      "epoch": 0.0874,
      "grad_norm": 0.15154384076595306,
      "learning_rate": 0.0001345204,
      "loss": 0.2688,
      "step": 163700
    },
    {
      "epoch": 0.0876,
      "grad_norm": 0.2439052164554596,
      "learning_rate": 0.00013448040000000003,
      "loss": 0.2656,
      "step": 163800
    },
    {
      "epoch": 0.0878,
      "grad_norm": 0.1832723617553711,
      "learning_rate": 0.0001344404,
      "loss": 0.2657,
      "step": 163900
    },
    {
      "epoch": 0.088,
      "grad_norm": 0.2097846120595932,
      "learning_rate": 0.00013440040000000002,
      "loss": 0.2633,
      "step": 164000
    },
    {
      "epoch": 0.0882,
      "grad_norm": 0.2136882096529007,
      "learning_rate": 0.0001343604,
      "loss": 0.2756,
      "step": 164100
    },
    {
      "epoch": 0.0884,
      "grad_norm": 0.20696696639060974,
      "learning_rate": 0.0001343204,
      "loss": 0.2675,
      "step": 164200
    },
    {
      "epoch": 0.0886,
      "grad_norm": 0.2125898003578186,
      "learning_rate": 0.00013428040000000002,
      "loss": 0.2728,
      "step": 164300
    },
    {
      "epoch": 0.0888,
      "grad_norm": 0.16671323776245117,
      "learning_rate": 0.0001342404,
      "loss": 0.2667,
      "step": 164400
    },
    {
      "epoch": 0.089,
      "grad_norm": 0.18992967903614044,
      "learning_rate": 0.0001342004,
      "loss": 0.2666,
      "step": 164500
    },
    {
      "epoch": 0.0892,
      "grad_norm": 0.167401522397995,
      "learning_rate": 0.0001341604,
      "loss": 0.2649,
      "step": 164600
    },
    {
      "epoch": 0.0894,
      "grad_norm": 0.18029069900512695,
      "learning_rate": 0.00013412040000000002,
      "loss": 0.2683,
      "step": 164700
    },
    {
      "epoch": 0.0896,
      "grad_norm": 0.17287403345108032,
      "learning_rate": 0.00013408040000000002,
      "loss": 0.2697,
      "step": 164800
    },
    {
      "epoch": 0.0898,
      "grad_norm": 0.15097787976264954,
      "learning_rate": 0.0001340404,
      "loss": 0.265,
      "step": 164900
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.1807047575712204,
      "learning_rate": 0.0001340004,
      "loss": 0.2669,
      "step": 165000
    },
    {
      "epoch": 0.0902,
      "grad_norm": 0.20332248508930206,
      "learning_rate": 0.0001339604,
      "loss": 0.2699,
      "step": 165100
    },
    {
      "epoch": 0.0904,
      "grad_norm": 0.24715851247310638,
      "learning_rate": 0.00013392040000000002,
      "loss": 0.2661,
      "step": 165200
    },
    {
      "epoch": 0.0906,
      "grad_norm": 0.1838693916797638,
      "learning_rate": 0.0001338804,
      "loss": 0.2645,
      "step": 165300
    },
    {
      "epoch": 0.0908,
      "grad_norm": 0.29177427291870117,
      "learning_rate": 0.0001338404,
      "loss": 0.2686,
      "step": 165400
    },
    {
      "epoch": 0.091,
      "grad_norm": 0.16363517940044403,
      "learning_rate": 0.0001338004,
      "loss": 0.2663,
      "step": 165500
    },
    {
      "epoch": 0.0912,
      "grad_norm": 0.17923811078071594,
      "learning_rate": 0.00013376040000000002,
      "loss": 0.2637,
      "step": 165600
    },
    {
      "epoch": 0.0914,
      "grad_norm": 0.14745573699474335,
      "learning_rate": 0.00013372040000000001,
      "loss": 0.2644,
      "step": 165700
    },
    {
      "epoch": 0.0916,
      "grad_norm": 0.17744286358356476,
      "learning_rate": 0.0001336804,
      "loss": 0.2697,
      "step": 165800
    },
    {
      "epoch": 0.0918,
      "grad_norm": 0.2632181942462921,
      "learning_rate": 0.0001336404,
      "loss": 0.2622,
      "step": 165900
    },
    {
      "epoch": 0.092,
      "grad_norm": 0.22580741345882416,
      "learning_rate": 0.0001336004,
      "loss": 0.2623,
      "step": 166000
    },
    {
      "epoch": 0.0922,
      "grad_norm": 0.2017957717180252,
      "learning_rate": 0.00013356040000000002,
      "loss": 0.2723,
      "step": 166100
    },
    {
      "epoch": 0.0924,
      "grad_norm": 0.20650072395801544,
      "learning_rate": 0.0001335204,
      "loss": 0.2646,
      "step": 166200
    },
    {
      "epoch": 0.0926,
      "grad_norm": 0.19651727378368378,
      "learning_rate": 0.0001334804,
      "loss": 0.2638,
      "step": 166300
    },
    {
      "epoch": 0.0928,
      "grad_norm": 0.16937686502933502,
      "learning_rate": 0.0001334404,
      "loss": 0.2627,
      "step": 166400
    },
    {
      "epoch": 0.093,
      "grad_norm": 0.16137665510177612,
      "learning_rate": 0.0001334004,
      "loss": 0.2643,
      "step": 166500
    },
    {
      "epoch": 0.0932,
      "grad_norm": 0.18096907436847687,
      "learning_rate": 0.0001333604,
      "loss": 0.2623,
      "step": 166600
    },
    {
      "epoch": 0.0934,
      "grad_norm": 0.1687513142824173,
      "learning_rate": 0.0001333204,
      "loss": 0.2679,
      "step": 166700
    },
    {
      "epoch": 0.0936,
      "grad_norm": 0.18491438031196594,
      "learning_rate": 0.00013328040000000003,
      "loss": 0.2663,
      "step": 166800
    },
    {
      "epoch": 0.0938,
      "grad_norm": 0.21767160296440125,
      "learning_rate": 0.0001332404,
      "loss": 0.2614,
      "step": 166900
    },
    {
      "epoch": 0.094,
      "grad_norm": 0.18175816535949707,
      "learning_rate": 0.0001332004,
      "loss": 0.2638,
      "step": 167000
    },
    {
      "epoch": 0.0942,
      "grad_norm": 0.17034652829170227,
      "learning_rate": 0.0001331604,
      "loss": 0.2656,
      "step": 167100
    },
    {
      "epoch": 0.0944,
      "grad_norm": 0.1829775869846344,
      "learning_rate": 0.0001331204,
      "loss": 0.2723,
      "step": 167200
    },
    {
      "epoch": 0.0946,
      "grad_norm": 0.16462214291095734,
      "learning_rate": 0.00013308040000000002,
      "loss": 0.2678,
      "step": 167300
    },
    {
      "epoch": 0.0948,
      "grad_norm": 0.18933749198913574,
      "learning_rate": 0.0001330404,
      "loss": 0.2598,
      "step": 167400
    },
    {
      "epoch": 0.095,
      "grad_norm": 0.2435947209596634,
      "learning_rate": 0.0001330004,
      "loss": 0.2648,
      "step": 167500
    },
    {
      "epoch": 0.0952,
      "grad_norm": 0.20233914256095886,
      "learning_rate": 0.0001329604,
      "loss": 0.2659,
      "step": 167600
    },
    {
      "epoch": 0.0954,
      "grad_norm": 0.17714782059192657,
      "learning_rate": 0.00013292040000000002,
      "loss": 0.2644,
      "step": 167700
    },
    {
      "epoch": 0.0956,
      "grad_norm": 0.14922687411308289,
      "learning_rate": 0.00013288040000000002,
      "loss": 0.2695,
      "step": 167800
    },
    {
      "epoch": 0.0958,
      "grad_norm": 0.1976463347673416,
      "learning_rate": 0.0001328404,
      "loss": 0.2628,
      "step": 167900
    },
    {
      "epoch": 0.096,
      "grad_norm": 0.18765133619308472,
      "learning_rate": 0.0001328004,
      "loss": 0.2612,
      "step": 168000
    },
    {
      "epoch": 0.0962,
      "grad_norm": 0.2899118959903717,
      "learning_rate": 0.0001327604,
      "loss": 0.2677,
      "step": 168100
    },
    {
      "epoch": 0.0964,
      "grad_norm": 0.20916017889976501,
      "learning_rate": 0.00013272040000000002,
      "loss": 0.2664,
      "step": 168200
    },
    {
      "epoch": 0.0966,
      "grad_norm": 0.20737506449222565,
      "learning_rate": 0.0001326804,
      "loss": 0.2585,
      "step": 168300
    },
    {
      "epoch": 0.0968,
      "grad_norm": 0.16495130956172943,
      "learning_rate": 0.0001326404,
      "loss": 0.2618,
      "step": 168400
    },
    {
      "epoch": 0.097,
      "grad_norm": 0.3457525074481964,
      "learning_rate": 0.0001326004,
      "loss": 0.2603,
      "step": 168500
    },
    {
      "epoch": 0.0972,
      "grad_norm": 0.16713711619377136,
      "learning_rate": 0.00013256040000000002,
      "loss": 0.2631,
      "step": 168600
    },
    {
      "epoch": 0.0974,
      "grad_norm": 0.4487501382827759,
      "learning_rate": 0.0001325204,
      "loss": 0.2655,
      "step": 168700
    },
    {
      "epoch": 0.0976,
      "grad_norm": 0.16808322072029114,
      "learning_rate": 0.0001324804,
      "loss": 0.265,
      "step": 168800
    },
    {
      "epoch": 0.0978,
      "grad_norm": 0.1778227984905243,
      "learning_rate": 0.0001324404,
      "loss": 0.2629,
      "step": 168900
    },
    {
      "epoch": 0.098,
      "grad_norm": 0.30036282539367676,
      "learning_rate": 0.0001324004,
      "loss": 0.269,
      "step": 169000
    },
    {
      "epoch": 0.0982,
      "grad_norm": 0.5122041702270508,
      "learning_rate": 0.00013236040000000001,
      "loss": 0.2575,
      "step": 169100
    },
    {
      "epoch": 0.0984,
      "grad_norm": 0.14579719305038452,
      "learning_rate": 0.0001323204,
      "loss": 0.2615,
      "step": 169200
    },
    {
      "epoch": 0.0986,
      "grad_norm": 0.2066614180803299,
      "learning_rate": 0.00013228040000000003,
      "loss": 0.2618,
      "step": 169300
    },
    {
      "epoch": 0.0988,
      "grad_norm": 0.20652034878730774,
      "learning_rate": 0.0001322404,
      "loss": 0.2491,
      "step": 169400
    },
    {
      "epoch": 0.099,
      "grad_norm": 0.23203743994235992,
      "learning_rate": 0.0001322004,
      "loss": 0.2618,
      "step": 169500
    },
    {
      "epoch": 0.0992,
      "grad_norm": 0.2068849503993988,
      "learning_rate": 0.0001321604,
      "loss": 0.2606,
      "step": 169600
    },
    {
      "epoch": 0.0994,
      "grad_norm": 0.19553183019161224,
      "learning_rate": 0.0001321204,
      "loss": 0.2661,
      "step": 169700
    },
    {
      "epoch": 0.0996,
      "grad_norm": 0.17411251366138458,
      "learning_rate": 0.00013208040000000002,
      "loss": 0.2633,
      "step": 169800
    },
    {
      "epoch": 0.0998,
      "grad_norm": 0.23143243789672852,
      "learning_rate": 0.0001320404,
      "loss": 0.2562,
      "step": 169900
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.18598131835460663,
      "learning_rate": 0.0001320004,
      "loss": 0.2585,
      "step": 170000
    },
    {
      "epoch": 0.1002,
      "grad_norm": 0.2250274121761322,
      "learning_rate": 0.0001319604,
      "loss": 0.2631,
      "step": 170100
    },
    {
      "epoch": 0.1004,
      "grad_norm": 0.2246299684047699,
      "learning_rate": 0.0001319204,
      "loss": 0.2663,
      "step": 170200
    },
    {
      "epoch": 0.1006,
      "grad_norm": 0.18941205739974976,
      "learning_rate": 0.00013188040000000002,
      "loss": 0.2609,
      "step": 170300
    },
    {
      "epoch": 0.1008,
      "grad_norm": 0.16915076971054077,
      "learning_rate": 0.00013184039999999999,
      "loss": 0.2615,
      "step": 170400
    },
    {
      "epoch": 0.101,
      "grad_norm": 0.18507012724876404,
      "learning_rate": 0.0001318004,
      "loss": 0.2668,
      "step": 170500
    },
    {
      "epoch": 0.1012,
      "grad_norm": 0.19404035806655884,
      "learning_rate": 0.0001317604,
      "loss": 0.2604,
      "step": 170600
    },
    {
      "epoch": 0.1014,
      "grad_norm": 0.20818071067333221,
      "learning_rate": 0.00013172040000000002,
      "loss": 0.2635,
      "step": 170700
    },
    {
      "epoch": 0.1016,
      "grad_norm": 0.20802633464336395,
      "learning_rate": 0.00013168040000000001,
      "loss": 0.2608,
      "step": 170800
    },
    {
      "epoch": 0.1018,
      "grad_norm": 0.25072330236434937,
      "learning_rate": 0.0001316404,
      "loss": 0.2589,
      "step": 170900
    },
    {
      "epoch": 0.102,
      "grad_norm": 0.1490134745836258,
      "learning_rate": 0.0001316004,
      "loss": 0.2642,
      "step": 171000
    },
    {
      "epoch": 0.1022,
      "grad_norm": 0.15050648152828217,
      "learning_rate": 0.0001315604,
      "loss": 0.2625,
      "step": 171100
    },
    {
      "epoch": 0.1024,
      "grad_norm": 0.16798128187656403,
      "learning_rate": 0.00013152040000000002,
      "loss": 0.2601,
      "step": 171200
    },
    {
      "epoch": 0.1026,
      "grad_norm": 0.1618351936340332,
      "learning_rate": 0.0001314804,
      "loss": 0.2618,
      "step": 171300
    },
    {
      "epoch": 0.1028,
      "grad_norm": 0.19324688613414764,
      "learning_rate": 0.0001314404,
      "loss": 0.2617,
      "step": 171400
    },
    {
      "epoch": 0.103,
      "grad_norm": 0.16977021098136902,
      "learning_rate": 0.0001314004,
      "loss": 0.2583,
      "step": 171500
    },
    {
      "epoch": 0.1032,
      "grad_norm": 0.1766977161169052,
      "learning_rate": 0.00013136040000000002,
      "loss": 0.2617,
      "step": 171600
    },
    {
      "epoch": 0.1034,
      "grad_norm": 0.16178765892982483,
      "learning_rate": 0.0001313204,
      "loss": 0.2735,
      "step": 171700
    },
    {
      "epoch": 0.1036,
      "grad_norm": 0.2680456340312958,
      "learning_rate": 0.0001312804,
      "loss": 0.261,
      "step": 171800
    },
    {
      "epoch": 0.1038,
      "grad_norm": 0.1756458580493927,
      "learning_rate": 0.0001312404,
      "loss": 0.2736,
      "step": 171900
    },
    {
      "epoch": 0.104,
      "grad_norm": 0.14756278693675995,
      "learning_rate": 0.0001312004,
      "loss": 0.2646,
      "step": 172000
    },
    {
      "epoch": 0.1042,
      "grad_norm": 0.4272075593471527,
      "learning_rate": 0.0001311604,
      "loss": 0.2837,
      "step": 172100
    },
    {
      "epoch": 0.1044,
      "grad_norm": 0.2615519165992737,
      "learning_rate": 0.0001311204,
      "loss": 0.2742,
      "step": 172200
    },
    {
      "epoch": 0.1046,
      "grad_norm": 0.15839534997940063,
      "learning_rate": 0.00013108040000000003,
      "loss": 0.2642,
      "step": 172300
    },
    {
      "epoch": 0.1048,
      "grad_norm": 0.20636843144893646,
      "learning_rate": 0.0001310404,
      "loss": 0.2663,
      "step": 172400
    },
    {
      "epoch": 0.105,
      "grad_norm": 0.16299419105052948,
      "learning_rate": 0.00013100040000000001,
      "loss": 0.2599,
      "step": 172500
    },
    {
      "epoch": 0.1052,
      "grad_norm": 0.19725507497787476,
      "learning_rate": 0.0001309604,
      "loss": 0.2705,
      "step": 172600
    },
    {
      "epoch": 0.1054,
      "grad_norm": 0.20907683670520782,
      "learning_rate": 0.0001309204,
      "loss": 0.2637,
      "step": 172700
    },
    {
      "epoch": 0.1056,
      "grad_norm": 0.1777603030204773,
      "learning_rate": 0.00013088040000000002,
      "loss": 0.2696,
      "step": 172800
    },
    {
      "epoch": 0.1058,
      "grad_norm": 0.2649913430213928,
      "learning_rate": 0.0001308404,
      "loss": 0.2602,
      "step": 172900
    },
    {
      "epoch": 0.106,
      "grad_norm": 0.21683943271636963,
      "learning_rate": 0.0001308004,
      "loss": 0.268,
      "step": 173000
    },
    {
      "epoch": 0.1062,
      "grad_norm": 0.1863240897655487,
      "learning_rate": 0.0001307604,
      "loss": 0.2699,
      "step": 173100
    },
    {
      "epoch": 0.1064,
      "grad_norm": 0.38173478841781616,
      "learning_rate": 0.0001307204,
      "loss": 0.2741,
      "step": 173200
    },
    {
      "epoch": 0.1066,
      "grad_norm": 0.19350466132164001,
      "learning_rate": 0.00013068040000000002,
      "loss": 0.264,
      "step": 173300
    },
    {
      "epoch": 0.1068,
      "grad_norm": 0.21752768754959106,
      "learning_rate": 0.00013064039999999998,
      "loss": 0.2638,
      "step": 173400
    },
    {
      "epoch": 0.107,
      "grad_norm": 0.22547826170921326,
      "learning_rate": 0.0001306004,
      "loss": 0.2669,
      "step": 173500
    },
    {
      "epoch": 0.1072,
      "grad_norm": 0.17793551087379456,
      "learning_rate": 0.0001305604,
      "loss": 0.2668,
      "step": 173600
    },
    {
      "epoch": 0.1074,
      "grad_norm": 0.18250034749507904,
      "learning_rate": 0.00013052040000000002,
      "loss": 0.2634,
      "step": 173700
    },
    {
      "epoch": 0.1076,
      "grad_norm": 0.17843498289585114,
      "learning_rate": 0.0001304804,
      "loss": 0.2632,
      "step": 173800
    },
    {
      "epoch": 0.1078,
      "grad_norm": 0.2799086570739746,
      "learning_rate": 0.0001304404,
      "loss": 0.2619,
      "step": 173900
    },
    {
      "epoch": 0.108,
      "grad_norm": 0.2233753800392151,
      "learning_rate": 0.0001304004,
      "loss": 0.2625,
      "step": 174000
    },
    {
      "epoch": 0.1082,
      "grad_norm": 0.18286100029945374,
      "learning_rate": 0.0001303604,
      "loss": 0.2634,
      "step": 174100
    },
    {
      "epoch": 0.1084,
      "grad_norm": 0.1985454261302948,
      "learning_rate": 0.0001303204,
      "loss": 0.2626,
      "step": 174200
    },
    {
      "epoch": 0.1086,
      "grad_norm": 0.191930890083313,
      "learning_rate": 0.0001302804,
      "loss": 0.261,
      "step": 174300
    },
    {
      "epoch": 0.1088,
      "grad_norm": 0.17980682849884033,
      "learning_rate": 0.0001302404,
      "loss": 0.264,
      "step": 174400
    },
    {
      "epoch": 0.109,
      "grad_norm": 0.16460399329662323,
      "learning_rate": 0.0001302004,
      "loss": 0.2657,
      "step": 174500
    },
    {
      "epoch": 0.1092,
      "grad_norm": 0.22608372569084167,
      "learning_rate": 0.00013016040000000002,
      "loss": 0.2671,
      "step": 174600
    },
    {
      "epoch": 0.1094,
      "grad_norm": 0.22313453257083893,
      "learning_rate": 0.0001301204,
      "loss": 0.2633,
      "step": 174700
    },
    {
      "epoch": 0.1096,
      "grad_norm": 0.20686650276184082,
      "learning_rate": 0.0001300804,
      "loss": 0.2655,
      "step": 174800
    },
    {
      "epoch": 0.1098,
      "grad_norm": 0.23883718252182007,
      "learning_rate": 0.0001300404,
      "loss": 0.2633,
      "step": 174900
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.19800545275211334,
      "learning_rate": 0.0001300004,
      "loss": 0.268,
      "step": 175000
    },
    {
      "epoch": 0.1102,
      "grad_norm": 0.18685345351696014,
      "learning_rate": 0.0001299604,
      "loss": 0.2678,
      "step": 175100
    },
    {
      "epoch": 0.1104,
      "grad_norm": 0.22192412614822388,
      "learning_rate": 0.0001299204,
      "loss": 0.2689,
      "step": 175200
    },
    {
      "epoch": 0.1106,
      "grad_norm": 0.20726250112056732,
      "learning_rate": 0.00012988040000000002,
      "loss": 0.2613,
      "step": 175300
    },
    {
      "epoch": 0.1108,
      "grad_norm": 0.2257356494665146,
      "learning_rate": 0.0001298404,
      "loss": 0.264,
      "step": 175400
    },
    {
      "epoch": 0.111,
      "grad_norm": 0.21950094401836395,
      "learning_rate": 0.0001298004,
      "loss": 0.2664,
      "step": 175500
    },
    {
      "epoch": 0.1112,
      "grad_norm": 0.18882116675376892,
      "learning_rate": 0.0001297604,
      "loss": 0.2719,
      "step": 175600
    },
    {
      "epoch": 0.1114,
      "grad_norm": 0.2492358237504959,
      "learning_rate": 0.0001297204,
      "loss": 0.2588,
      "step": 175700
    },
    {
      "epoch": 0.1116,
      "grad_norm": 0.15519103407859802,
      "learning_rate": 0.00012968040000000002,
      "loss": 0.2629,
      "step": 175800
    },
    {
      "epoch": 0.1118,
      "grad_norm": 0.16526469588279724,
      "learning_rate": 0.00012964039999999999,
      "loss": 0.2594,
      "step": 175900
    },
    {
      "epoch": 0.112,
      "grad_norm": 0.1781369298696518,
      "learning_rate": 0.0001296004,
      "loss": 0.2638,
      "step": 176000
    },
    {
      "epoch": 0.1122,
      "grad_norm": 0.19480067491531372,
      "learning_rate": 0.0001295604,
      "loss": 0.263,
      "step": 176100
    },
    {
      "epoch": 0.1124,
      "grad_norm": 0.20331743359565735,
      "learning_rate": 0.00012952040000000002,
      "loss": 0.2595,
      "step": 176200
    },
    {
      "epoch": 0.1126,
      "grad_norm": 0.404156357049942,
      "learning_rate": 0.00012948040000000001,
      "loss": 0.264,
      "step": 176300
    },
    {
      "epoch": 0.1128,
      "grad_norm": 0.17432494461536407,
      "learning_rate": 0.0001294404,
      "loss": 0.2616,
      "step": 176400
    },
    {
      "epoch": 0.113,
      "grad_norm": 0.17324478924274445,
      "learning_rate": 0.0001294004,
      "loss": 0.2566,
      "step": 176500
    },
    {
      "epoch": 0.1132,
      "grad_norm": 0.20663540065288544,
      "learning_rate": 0.0001293604,
      "loss": 0.2659,
      "step": 176600
    },
    {
      "epoch": 0.1134,
      "grad_norm": 0.16922618448734283,
      "learning_rate": 0.00012932040000000002,
      "loss": 0.2621,
      "step": 176700
    },
    {
      "epoch": 0.1136,
      "grad_norm": 0.1731468141078949,
      "learning_rate": 0.0001292804,
      "loss": 0.2583,
      "step": 176800
    },
    {
      "epoch": 0.1138,
      "grad_norm": 0.3375247120857239,
      "learning_rate": 0.0001292404,
      "loss": 0.2624,
      "step": 176900
    },
    {
      "epoch": 0.114,
      "grad_norm": 0.14527542889118195,
      "learning_rate": 0.0001292004,
      "loss": 0.2598,
      "step": 177000
    },
    {
      "epoch": 0.1142,
      "grad_norm": 0.18705546855926514,
      "learning_rate": 0.0001291604,
      "loss": 0.2602,
      "step": 177100
    },
    {
      "epoch": 0.1144,
      "grad_norm": 0.23862403631210327,
      "learning_rate": 0.0001291204,
      "loss": 0.2568,
      "step": 177200
    },
    {
      "epoch": 0.1146,
      "grad_norm": 0.21297907829284668,
      "learning_rate": 0.0001290804,
      "loss": 0.2687,
      "step": 177300
    },
    {
      "epoch": 0.1148,
      "grad_norm": 0.24559397995471954,
      "learning_rate": 0.00012904040000000003,
      "loss": 0.2638,
      "step": 177400
    },
    {
      "epoch": 0.115,
      "grad_norm": 0.1739484816789627,
      "learning_rate": 0.0001290004,
      "loss": 0.26,
      "step": 177500
    },
    {
      "epoch": 0.1152,
      "grad_norm": 0.1745755672454834,
      "learning_rate": 0.0001289604,
      "loss": 0.2781,
      "step": 177600
    },
    {
      "epoch": 0.1154,
      "grad_norm": 0.4448339343070984,
      "learning_rate": 0.0001289204,
      "loss": 0.2606,
      "step": 177700
    },
    {
      "epoch": 0.1156,
      "grad_norm": 0.18532177805900574,
      "learning_rate": 0.0001288804,
      "loss": 0.2695,
      "step": 177800
    },
    {
      "epoch": 0.1158,
      "grad_norm": 0.20550072193145752,
      "learning_rate": 0.00012884040000000002,
      "loss": 0.2586,
      "step": 177900
    },
    {
      "epoch": 0.116,
      "grad_norm": 0.18768703937530518,
      "learning_rate": 0.0001288004,
      "loss": 0.2581,
      "step": 178000
    },
    {
      "epoch": 0.1162,
      "grad_norm": 0.21779786050319672,
      "learning_rate": 0.0001287604,
      "loss": 0.2598,
      "step": 178100
    },
    {
      "epoch": 0.1164,
      "grad_norm": 0.16268861293792725,
      "learning_rate": 0.0001287204,
      "loss": 0.2639,
      "step": 178200
    },
    {
      "epoch": 0.1166,
      "grad_norm": 0.18373139202594757,
      "learning_rate": 0.00012868040000000002,
      "loss": 0.2573,
      "step": 178300
    },
    {
      "epoch": 0.1168,
      "grad_norm": 0.17486025393009186,
      "learning_rate": 0.00012864040000000002,
      "loss": 0.2628,
      "step": 178400
    },
    {
      "epoch": 0.117,
      "grad_norm": 0.17271597683429718,
      "learning_rate": 0.0001286004,
      "loss": 0.2603,
      "step": 178500
    },
    {
      "epoch": 0.1172,
      "grad_norm": 0.15884092450141907,
      "learning_rate": 0.0001285604,
      "loss": 0.2616,
      "step": 178600
    },
    {
      "epoch": 0.1174,
      "grad_norm": 0.22521570324897766,
      "learning_rate": 0.0001285204,
      "loss": 0.2596,
      "step": 178700
    },
    {
      "epoch": 0.1176,
      "grad_norm": 0.1907728910446167,
      "learning_rate": 0.00012848040000000002,
      "loss": 0.2585,
      "step": 178800
    },
    {
      "epoch": 0.1178,
      "grad_norm": 0.24702465534210205,
      "learning_rate": 0.0001284404,
      "loss": 0.268,
      "step": 178900
    },
    {
      "epoch": 0.118,
      "grad_norm": 0.15964145958423615,
      "learning_rate": 0.0001284004,
      "loss": 0.2633,
      "step": 179000
    },
    {
      "epoch": 0.1182,
      "grad_norm": 0.17996646463871002,
      "learning_rate": 0.0001283604,
      "loss": 0.2581,
      "step": 179100
    },
    {
      "epoch": 0.1184,
      "grad_norm": 0.1755184382200241,
      "learning_rate": 0.00012832040000000002,
      "loss": 0.26,
      "step": 179200
    },
    {
      "epoch": 0.1186,
      "grad_norm": 0.16567598283290863,
      "learning_rate": 0.0001282804,
      "loss": 0.2568,
      "step": 179300
    },
    {
      "epoch": 0.1188,
      "grad_norm": 0.16007539629936218,
      "learning_rate": 0.0001282404,
      "loss": 0.262,
      "step": 179400
    },
    {
      "epoch": 0.119,
      "grad_norm": 0.28034693002700806,
      "learning_rate": 0.0001282004,
      "loss": 0.2603,
      "step": 179500
    },
    {
      "epoch": 0.1192,
      "grad_norm": 0.28950920701026917,
      "learning_rate": 0.0001281604,
      "loss": 0.2611,
      "step": 179600
    },
    {
      "epoch": 0.1194,
      "grad_norm": 0.18529054522514343,
      "learning_rate": 0.00012812040000000001,
      "loss": 0.2574,
      "step": 179700
    },
    {
      "epoch": 0.1196,
      "grad_norm": 0.18269339203834534,
      "learning_rate": 0.0001280804,
      "loss": 0.2644,
      "step": 179800
    },
    {
      "epoch": 0.1198,
      "grad_norm": 0.27880364656448364,
      "learning_rate": 0.00012804040000000003,
      "loss": 0.2618,
      "step": 179900
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.1725231260061264,
      "learning_rate": 0.0001280004,
      "loss": 0.2599,
      "step": 180000
    },
    {
      "epoch": 0.1202,
      "grad_norm": 0.2305331975221634,
      "learning_rate": 0.00012796040000000002,
      "loss": 0.2625,
      "step": 180100
    },
    {
      "epoch": 0.1204,
      "grad_norm": 0.14379462599754333,
      "learning_rate": 0.0001279204,
      "loss": 0.2627,
      "step": 180200
    },
    {
      "epoch": 0.1206,
      "grad_norm": 0.16537553071975708,
      "learning_rate": 0.0001278804,
      "loss": 0.2721,
      "step": 180300
    },
    {
      "epoch": 0.1208,
      "grad_norm": 0.21422922611236572,
      "learning_rate": 0.00012784040000000002,
      "loss": 0.2627,
      "step": 180400
    },
    {
      "epoch": 0.121,
      "grad_norm": 0.1767064481973648,
      "learning_rate": 0.0001278004,
      "loss": 0.2594,
      "step": 180500
    },
    {
      "epoch": 0.1212,
      "grad_norm": 0.23920567333698273,
      "learning_rate": 0.0001277604,
      "loss": 0.2595,
      "step": 180600
    },
    {
      "epoch": 0.1214,
      "grad_norm": 0.16521887481212616,
      "learning_rate": 0.0001277204,
      "loss": 0.2612,
      "step": 180700
    },
    {
      "epoch": 0.1216,
      "grad_norm": 0.1697589010000229,
      "learning_rate": 0.0001276804,
      "loss": 0.2619,
      "step": 180800
    },
    {
      "epoch": 0.1218,
      "grad_norm": 0.20626257359981537,
      "learning_rate": 0.00012764040000000002,
      "loss": 0.2642,
      "step": 180900
    },
    {
      "epoch": 0.122,
      "grad_norm": 0.16152922809123993,
      "learning_rate": 0.00012760039999999999,
      "loss": 0.2637,
      "step": 181000
    },
    {
      "epoch": 0.1222,
      "grad_norm": 0.15324938297271729,
      "learning_rate": 0.0001275604,
      "loss": 0.2557,
      "step": 181100
    },
    {
      "epoch": 0.1224,
      "grad_norm": 0.18400631844997406,
      "learning_rate": 0.0001275204,
      "loss": 0.2643,
      "step": 181200
    },
    {
      "epoch": 0.1226,
      "grad_norm": 0.1637372076511383,
      "learning_rate": 0.00012748040000000002,
      "loss": 0.2673,
      "step": 181300
    },
    {
      "epoch": 0.1228,
      "grad_norm": 0.23258568346500397,
      "learning_rate": 0.00012744040000000001,
      "loss": 0.2619,
      "step": 181400
    },
    {
      "epoch": 0.123,
      "grad_norm": 0.16462406516075134,
      "learning_rate": 0.0001274004,
      "loss": 0.2589,
      "step": 181500
    },
    {
      "epoch": 0.1232,
      "grad_norm": 0.19165945053100586,
      "learning_rate": 0.0001273604,
      "loss": 0.2649,
      "step": 181600
    },
    {
      "epoch": 0.1234,
      "grad_norm": 0.23923306167125702,
      "learning_rate": 0.0001273204,
      "loss": 0.2711,
      "step": 181700
    },
    {
      "epoch": 0.1236,
      "grad_norm": 0.23697294294834137,
      "learning_rate": 0.00012728040000000002,
      "loss": 0.2658,
      "step": 181800
    },
    {
      "epoch": 0.1238,
      "grad_norm": 0.24161309003829956,
      "learning_rate": 0.0001272404,
      "loss": 0.258,
      "step": 181900
    },
    {
      "epoch": 0.124,
      "grad_norm": 0.1704692542552948,
      "learning_rate": 0.0001272004,
      "loss": 0.2615,
      "step": 182000
    },
    {
      "epoch": 0.1242,
      "grad_norm": 0.16872115433216095,
      "learning_rate": 0.0001271604,
      "loss": 0.2674,
      "step": 182100
    },
    {
      "epoch": 0.1244,
      "grad_norm": 0.20729516446590424,
      "learning_rate": 0.00012712040000000002,
      "loss": 0.2585,
      "step": 182200
    },
    {
      "epoch": 0.1246,
      "grad_norm": 0.20898325741291046,
      "learning_rate": 0.0001270804,
      "loss": 0.2627,
      "step": 182300
    },
    {
      "epoch": 0.1248,
      "grad_norm": 0.1441413015127182,
      "learning_rate": 0.0001270404,
      "loss": 0.2603,
      "step": 182400
    },
    {
      "epoch": 0.125,
      "grad_norm": 0.18905945122241974,
      "learning_rate": 0.0001270004,
      "loss": 0.254,
      "step": 182500
    },
    {
      "epoch": 0.1252,
      "grad_norm": 0.17578963935375214,
      "learning_rate": 0.0001269604,
      "loss": 0.2608,
      "step": 182600
    },
    {
      "epoch": 0.1254,
      "grad_norm": 0.23273304104804993,
      "learning_rate": 0.0001269204,
      "loss": 0.2625,
      "step": 182700
    },
    {
      "epoch": 0.1256,
      "grad_norm": 0.228327676653862,
      "learning_rate": 0.0001268804,
      "loss": 0.2598,
      "step": 182800
    },
    {
      "epoch": 0.1258,
      "grad_norm": 0.1630987524986267,
      "learning_rate": 0.00012684040000000003,
      "loss": 0.2645,
      "step": 182900
    },
    {
      "epoch": 0.126,
      "grad_norm": 0.18932506442070007,
      "learning_rate": 0.0001268004,
      "loss": 0.2633,
      "step": 183000
    },
    {
      "epoch": 0.1262,
      "grad_norm": 0.18664585053920746,
      "learning_rate": 0.00012676040000000001,
      "loss": 0.2627,
      "step": 183100
    },
    {
      "epoch": 0.1264,
      "grad_norm": 0.2066875547170639,
      "learning_rate": 0.0001267204,
      "loss": 0.2598,
      "step": 183200
    },
    {
      "epoch": 0.1266,
      "grad_norm": 0.20927661657333374,
      "learning_rate": 0.0001266804,
      "loss": 0.2659,
      "step": 183300
    },
    {
      "epoch": 0.1268,
      "grad_norm": 0.204673171043396,
      "learning_rate": 0.00012664040000000002,
      "loss": 0.2605,
      "step": 183400
    },
    {
      "epoch": 0.127,
      "grad_norm": 0.23662003874778748,
      "learning_rate": 0.0001266004,
      "loss": 0.2638,
      "step": 183500
    },
    {
      "epoch": 0.1272,
      "grad_norm": 0.30359625816345215,
      "learning_rate": 0.0001265604,
      "loss": 0.2597,
      "step": 183600
    },
    {
      "epoch": 0.1274,
      "grad_norm": 0.2385796308517456,
      "learning_rate": 0.0001265204,
      "loss": 0.2633,
      "step": 183700
    },
    {
      "epoch": 0.1276,
      "grad_norm": 0.3185318112373352,
      "learning_rate": 0.00012648040000000002,
      "loss": 0.2609,
      "step": 183800
    },
    {
      "epoch": 0.1278,
      "grad_norm": 0.17162935435771942,
      "learning_rate": 0.00012644040000000002,
      "loss": 0.2624,
      "step": 183900
    },
    {
      "epoch": 0.128,
      "grad_norm": 0.16286444664001465,
      "learning_rate": 0.00012640039999999998,
      "loss": 0.2602,
      "step": 184000
    },
    {
      "epoch": 0.1282,
      "grad_norm": 0.15972454845905304,
      "learning_rate": 0.0001263604,
      "loss": 0.2578,
      "step": 184100
    },
    {
      "epoch": 0.1284,
      "grad_norm": 0.1734430193901062,
      "learning_rate": 0.0001263204,
      "loss": 0.2639,
      "step": 184200
    },
    {
      "epoch": 0.1286,
      "grad_norm": 0.16511160135269165,
      "learning_rate": 0.00012628040000000002,
      "loss": 0.2606,
      "step": 184300
    },
    {
      "epoch": 0.1288,
      "grad_norm": 0.22905713319778442,
      "learning_rate": 0.0001262404,
      "loss": 0.258,
      "step": 184400
    },
    {
      "epoch": 0.129,
      "grad_norm": 0.16669723391532898,
      "learning_rate": 0.0001262004,
      "loss": 0.2582,
      "step": 184500
    },
    {
      "epoch": 0.1292,
      "grad_norm": 0.18425410985946655,
      "learning_rate": 0.0001261604,
      "loss": 0.2649,
      "step": 184600
    },
    {
      "epoch": 0.1294,
      "grad_norm": 0.18333709239959717,
      "learning_rate": 0.0001261204,
      "loss": 0.2636,
      "step": 184700
    },
    {
      "epoch": 0.1296,
      "grad_norm": 0.19844196736812592,
      "learning_rate": 0.0001260804,
      "loss": 0.263,
      "step": 184800
    },
    {
      "epoch": 0.1298,
      "grad_norm": 0.21080031991004944,
      "learning_rate": 0.0001260404,
      "loss": 0.2617,
      "step": 184900
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.16929513216018677,
      "learning_rate": 0.0001260004,
      "loss": 0.2637,
      "step": 185000
    },
    {
      "epoch": 0.1302,
      "grad_norm": 0.17420923709869385,
      "learning_rate": 0.0001259604,
      "loss": 0.2655,
      "step": 185100
    },
    {
      "epoch": 0.1304,
      "grad_norm": 0.15296240150928497,
      "learning_rate": 0.00012592040000000001,
      "loss": 0.2654,
      "step": 185200
    },
    {
      "epoch": 0.1306,
      "grad_norm": 0.18833494186401367,
      "learning_rate": 0.0001258804,
      "loss": 0.2657,
      "step": 185300
    },
    {
      "epoch": 0.1308,
      "grad_norm": 0.1517093926668167,
      "learning_rate": 0.0001258404,
      "loss": 0.2694,
      "step": 185400
    },
    {
      "epoch": 0.131,
      "grad_norm": 0.22326596081256866,
      "learning_rate": 0.0001258004,
      "loss": 0.2616,
      "step": 185500
    },
    {
      "epoch": 0.1312,
      "grad_norm": 0.17264465987682343,
      "learning_rate": 0.0001257604,
      "loss": 0.2618,
      "step": 185600
    },
    {
      "epoch": 0.1314,
      "grad_norm": 0.28218379616737366,
      "learning_rate": 0.0001257204,
      "loss": 0.2641,
      "step": 185700
    },
    {
      "epoch": 0.1316,
      "grad_norm": 0.19472543895244598,
      "learning_rate": 0.0001256804,
      "loss": 0.2638,
      "step": 185800
    },
    {
      "epoch": 0.1318,
      "grad_norm": 0.17265281081199646,
      "learning_rate": 0.00012564040000000002,
      "loss": 0.2624,
      "step": 185900
    },
    {
      "epoch": 0.132,
      "grad_norm": 0.16616639494895935,
      "learning_rate": 0.0001256004,
      "loss": 0.2569,
      "step": 186000
    },
    {
      "epoch": 0.1322,
      "grad_norm": 0.2015598714351654,
      "learning_rate": 0.0001255604,
      "loss": 0.2739,
      "step": 186100
    },
    {
      "epoch": 0.1324,
      "grad_norm": 0.16873787343502045,
      "learning_rate": 0.0001255204,
      "loss": 0.2643,
      "step": 186200
    },
    {
      "epoch": 0.1326,
      "grad_norm": 0.3382190763950348,
      "learning_rate": 0.0001254804,
      "loss": 0.2612,
      "step": 186300
    },
    {
      "epoch": 0.1328,
      "grad_norm": 0.1990320235490799,
      "learning_rate": 0.00012544040000000002,
      "loss": 0.2641,
      "step": 186400
    },
    {
      "epoch": 0.133,
      "grad_norm": 0.17715241014957428,
      "learning_rate": 0.00012540039999999999,
      "loss": 0.2619,
      "step": 186500
    },
    {
      "epoch": 0.1332,
      "grad_norm": 0.18182724714279175,
      "learning_rate": 0.0001253604,
      "loss": 0.2632,
      "step": 186600
    },
    {
      "epoch": 0.1334,
      "grad_norm": 0.24784061312675476,
      "learning_rate": 0.0001253204,
      "loss": 0.2643,
      "step": 186700
    },
    {
      "epoch": 0.1336,
      "grad_norm": 0.1730160117149353,
      "learning_rate": 0.00012528040000000002,
      "loss": 0.2599,
      "step": 186800
    },
    {
      "epoch": 0.1338,
      "grad_norm": 0.25901031494140625,
      "learning_rate": 0.00012524040000000001,
      "loss": 0.2627,
      "step": 186900
    },
    {
      "epoch": 0.134,
      "grad_norm": 0.2513553500175476,
      "learning_rate": 0.0001252004,
      "loss": 0.2601,
      "step": 187000
    },
    {
      "epoch": 0.1342,
      "grad_norm": 0.18127425014972687,
      "learning_rate": 0.0001251604,
      "loss": 0.2688,
      "step": 187100
    },
    {
      "epoch": 0.1344,
      "grad_norm": 0.16403953731060028,
      "learning_rate": 0.0001251204,
      "loss": 0.2604,
      "step": 187200
    },
    {
      "epoch": 0.1346,
      "grad_norm": 0.19153177738189697,
      "learning_rate": 0.00012508040000000002,
      "loss": 0.2607,
      "step": 187300
    },
    {
      "epoch": 0.1348,
      "grad_norm": 0.17384052276611328,
      "learning_rate": 0.0001250404,
      "loss": 0.2579,
      "step": 187400
    },
    {
      "epoch": 0.135,
      "grad_norm": 0.17870062589645386,
      "learning_rate": 0.0001250004,
      "loss": 0.2605,
      "step": 187500
    },
    {
      "epoch": 0.1352,
      "grad_norm": 0.16759884357452393,
      "learning_rate": 0.0001249604,
      "loss": 0.2579,
      "step": 187600
    },
    {
      "epoch": 0.1354,
      "grad_norm": 0.19198445975780487,
      "learning_rate": 0.0001249204,
      "loss": 0.2583,
      "step": 187700
    },
    {
      "epoch": 0.1356,
      "grad_norm": 0.17366041243076324,
      "learning_rate": 0.0001248804,
      "loss": 0.2566,
      "step": 187800
    },
    {
      "epoch": 0.1358,
      "grad_norm": 0.3349681794643402,
      "learning_rate": 0.0001248404,
      "loss": 0.264,
      "step": 187900
    },
    {
      "epoch": 0.136,
      "grad_norm": 0.21217289566993713,
      "learning_rate": 0.0001248004,
      "loss": 0.2657,
      "step": 188000
    },
    {
      "epoch": 0.1362,
      "grad_norm": 0.17328070104122162,
      "learning_rate": 0.0001247604,
      "loss": 0.2552,
      "step": 188100
    },
    {
      "epoch": 0.1364,
      "grad_norm": 0.3321928381919861,
      "learning_rate": 0.0001247204,
      "loss": 0.2617,
      "step": 188200
    },
    {
      "epoch": 0.1366,
      "grad_norm": 0.16838262975215912,
      "learning_rate": 0.0001246804,
      "loss": 0.2674,
      "step": 188300
    },
    {
      "epoch": 0.1368,
      "grad_norm": 0.19490031898021698,
      "learning_rate": 0.0001246404,
      "loss": 0.2651,
      "step": 188400
    },
    {
      "epoch": 0.137,
      "grad_norm": 0.2795969247817993,
      "learning_rate": 0.0001246004,
      "loss": 0.2598,
      "step": 188500
    },
    {
      "epoch": 0.1372,
      "grad_norm": 0.18094158172607422,
      "learning_rate": 0.0001245604,
      "loss": 0.2631,
      "step": 188600
    },
    {
      "epoch": 0.1374,
      "grad_norm": 0.17936332523822784,
      "learning_rate": 0.0001245204,
      "loss": 0.2599,
      "step": 188700
    },
    {
      "epoch": 0.1376,
      "grad_norm": 0.4011989235877991,
      "learning_rate": 0.0001244804,
      "loss": 0.2577,
      "step": 188800
    },
    {
      "epoch": 0.1378,
      "grad_norm": 0.1783379763364792,
      "learning_rate": 0.00012444040000000002,
      "loss": 0.2562,
      "step": 188900
    },
    {
      "epoch": 0.138,
      "grad_norm": 0.18040119111537933,
      "learning_rate": 0.00012440040000000002,
      "loss": 0.2579,
      "step": 189000
    },
    {
      "epoch": 0.1382,
      "grad_norm": 0.181350976228714,
      "learning_rate": 0.0001243604,
      "loss": 0.2662,
      "step": 189100
    },
    {
      "epoch": 0.1384,
      "grad_norm": 0.18030929565429688,
      "learning_rate": 0.0001243204,
      "loss": 0.2588,
      "step": 189200
    },
    {
      "epoch": 0.1386,
      "grad_norm": 0.19084039330482483,
      "learning_rate": 0.0001242804,
      "loss": 0.2654,
      "step": 189300
    },
    {
      "epoch": 0.1388,
      "grad_norm": 0.23569531738758087,
      "learning_rate": 0.00012424040000000002,
      "loss": 0.259,
      "step": 189400
    },
    {
      "epoch": 0.139,
      "grad_norm": 0.2429647147655487,
      "learning_rate": 0.0001242004,
      "loss": 0.2581,
      "step": 189500
    },
    {
      "epoch": 0.1392,
      "grad_norm": 0.20208866894245148,
      "learning_rate": 0.0001241604,
      "loss": 0.2563,
      "step": 189600
    },
    {
      "epoch": 0.1394,
      "grad_norm": 0.1706075370311737,
      "learning_rate": 0.0001241204,
      "loss": 0.2634,
      "step": 189700
    },
    {
      "epoch": 0.1396,
      "grad_norm": 0.1640571802854538,
      "learning_rate": 0.00012408040000000002,
      "loss": 0.2582,
      "step": 189800
    },
    {
      "epoch": 0.1398,
      "grad_norm": 0.22723062336444855,
      "learning_rate": 0.0001240404,
      "loss": 0.2603,
      "step": 189900
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.16876673698425293,
      "learning_rate": 0.0001240004,
      "loss": 0.2636,
      "step": 190000
    },
    {
      "epoch": 0.1402,
      "grad_norm": 0.2153484970331192,
      "learning_rate": 0.0001239604,
      "loss": 0.2586,
      "step": 190100
    },
    {
      "epoch": 0.1404,
      "grad_norm": 0.21755315363407135,
      "learning_rate": 0.0001239204,
      "loss": 0.2565,
      "step": 190200
    },
    {
      "epoch": 0.1406,
      "grad_norm": 0.18556717038154602,
      "learning_rate": 0.00012388040000000001,
      "loss": 0.2631,
      "step": 190300
    },
    {
      "epoch": 0.1408,
      "grad_norm": 0.19718804955482483,
      "learning_rate": 0.0001238404,
      "loss": 0.2593,
      "step": 190400
    },
    {
      "epoch": 0.141,
      "grad_norm": 0.2352330982685089,
      "learning_rate": 0.00012380040000000003,
      "loss": 0.2606,
      "step": 190500
    },
    {
      "epoch": 0.1412,
      "grad_norm": 0.22349053621292114,
      "learning_rate": 0.0001237604,
      "loss": 0.2614,
      "step": 190600
    },
    {
      "epoch": 0.1414,
      "grad_norm": 0.27996236085891724,
      "learning_rate": 0.00012372040000000002,
      "loss": 0.2651,
      "step": 190700
    },
    {
      "epoch": 0.1416,
      "grad_norm": 0.21903091669082642,
      "learning_rate": 0.0001236804,
      "loss": 0.2577,
      "step": 190800
    },
    {
      "epoch": 0.1418,
      "grad_norm": 0.1741315722465515,
      "learning_rate": 0.0001236404,
      "loss": 0.2526,
      "step": 190900
    },
    {
      "epoch": 0.142,
      "grad_norm": 0.2520033121109009,
      "learning_rate": 0.00012360040000000002,
      "loss": 0.259,
      "step": 191000
    },
    {
      "epoch": 0.1422,
      "grad_norm": 0.23828014731407166,
      "learning_rate": 0.0001235604,
      "loss": 0.2594,
      "step": 191100
    },
    {
      "epoch": 0.1424,
      "grad_norm": 0.16912546753883362,
      "learning_rate": 0.0001235204,
      "loss": 0.2616,
      "step": 191200
    },
    {
      "epoch": 0.1426,
      "grad_norm": 0.17198620736598969,
      "learning_rate": 0.0001234804,
      "loss": 0.256,
      "step": 191300
    },
    {
      "epoch": 0.1428,
      "grad_norm": 0.17525877058506012,
      "learning_rate": 0.0001234404,
      "loss": 0.2601,
      "step": 191400
    },
    {
      "epoch": 0.143,
      "grad_norm": 0.16216328740119934,
      "learning_rate": 0.00012340040000000002,
      "loss": 0.2641,
      "step": 191500
    },
    {
      "epoch": 0.1432,
      "grad_norm": 0.1828595995903015,
      "learning_rate": 0.00012336039999999999,
      "loss": 0.2582,
      "step": 191600
    },
    {
      "epoch": 0.1434,
      "grad_norm": 0.16869176924228668,
      "learning_rate": 0.0001233204,
      "loss": 0.2554,
      "step": 191700
    },
    {
      "epoch": 0.1436,
      "grad_norm": 0.19655516743659973,
      "learning_rate": 0.0001232804,
      "loss": 0.2663,
      "step": 191800
    },
    {
      "epoch": 0.1438,
      "grad_norm": 0.20977061986923218,
      "learning_rate": 0.00012324040000000002,
      "loss": 0.2565,
      "step": 191900
    },
    {
      "epoch": 0.144,
      "grad_norm": 0.22906339168548584,
      "learning_rate": 0.00012320040000000001,
      "loss": 0.2553,
      "step": 192000
    },
    {
      "epoch": 0.1442,
      "grad_norm": 0.1683686375617981,
      "learning_rate": 0.0001231604,
      "loss": 0.2581,
      "step": 192100
    },
    {
      "epoch": 0.1444,
      "grad_norm": 0.20973190665245056,
      "learning_rate": 0.0001231204,
      "loss": 0.2642,
      "step": 192200
    },
    {
      "epoch": 0.1446,
      "grad_norm": 0.14577773213386536,
      "learning_rate": 0.0001230804,
      "loss": 0.26,
      "step": 192300
    },
    {
      "epoch": 0.1448,
      "grad_norm": 0.19426365196704865,
      "learning_rate": 0.00012304040000000002,
      "loss": 0.2622,
      "step": 192400
    },
    {
      "epoch": 0.145,
      "grad_norm": 0.17555978894233704,
      "learning_rate": 0.0001230004,
      "loss": 0.2598,
      "step": 192500
    },
    {
      "epoch": 0.1452,
      "grad_norm": 0.2241448312997818,
      "learning_rate": 0.0001229604,
      "loss": 0.2549,
      "step": 192600
    },
    {
      "epoch": 0.1454,
      "grad_norm": 0.24879616498947144,
      "learning_rate": 0.0001229204,
      "loss": 0.2525,
      "step": 192700
    },
    {
      "epoch": 0.1456,
      "grad_norm": 0.18659988045692444,
      "learning_rate": 0.00012288040000000002,
      "loss": 0.2635,
      "step": 192800
    },
    {
      "epoch": 0.1458,
      "grad_norm": 0.19412420690059662,
      "learning_rate": 0.0001228404,
      "loss": 0.2553,
      "step": 192900
    },
    {
      "epoch": 0.146,
      "grad_norm": 0.16501417756080627,
      "learning_rate": 0.0001228004,
      "loss": 0.2621,
      "step": 193000
    },
    {
      "epoch": 0.1462,
      "grad_norm": 0.18333154916763306,
      "learning_rate": 0.0001227604,
      "loss": 0.2545,
      "step": 193100
    },
    {
      "epoch": 0.1464,
      "grad_norm": 0.25169798731803894,
      "learning_rate": 0.0001227204,
      "loss": 0.26,
      "step": 193200
    },
    {
      "epoch": 0.1466,
      "grad_norm": 0.18788355588912964,
      "learning_rate": 0.0001226804,
      "loss": 0.2575,
      "step": 193300
    },
    {
      "epoch": 0.1468,
      "grad_norm": 0.23078390955924988,
      "learning_rate": 0.0001226404,
      "loss": 0.2717,
      "step": 193400
    },
    {
      "epoch": 0.147,
      "grad_norm": 0.18893170356750488,
      "learning_rate": 0.00012260040000000003,
      "loss": 0.2572,
      "step": 193500
    },
    {
      "epoch": 0.1472,
      "grad_norm": 0.20808455348014832,
      "learning_rate": 0.0001225604,
      "loss": 0.2576,
      "step": 193600
    },
    {
      "epoch": 0.1474,
      "grad_norm": 0.1902913898229599,
      "learning_rate": 0.00012252040000000001,
      "loss": 0.2568,
      "step": 193700
    },
    {
      "epoch": 0.1476,
      "grad_norm": 0.17236173152923584,
      "learning_rate": 0.0001224804,
      "loss": 0.2541,
      "step": 193800
    },
    {
      "epoch": 0.1478,
      "grad_norm": 0.22261063754558563,
      "learning_rate": 0.0001224404,
      "loss": 0.2572,
      "step": 193900
    },
    {
      "epoch": 0.148,
      "grad_norm": 0.21105404198169708,
      "learning_rate": 0.00012240040000000002,
      "loss": 0.2595,
      "step": 194000
    },
    {
      "epoch": 0.1482,
      "grad_norm": 0.1775384098291397,
      "learning_rate": 0.0001223604,
      "loss": 0.2566,
      "step": 194100
    },
    {
      "epoch": 0.1484,
      "grad_norm": 0.16679984331130981,
      "learning_rate": 0.0001223204,
      "loss": 0.2567,
      "step": 194200
    },
    {
      "epoch": 0.1486,
      "grad_norm": 0.15811361372470856,
      "learning_rate": 0.0001222804,
      "loss": 0.2627,
      "step": 194300
    },
    {
      "epoch": 0.1488,
      "grad_norm": 0.1975756138563156,
      "learning_rate": 0.00012224040000000002,
      "loss": 0.2629,
      "step": 194400
    },
    {
      "epoch": 0.149,
      "grad_norm": 0.2405877411365509,
      "learning_rate": 0.00012220040000000002,
      "loss": 0.2591,
      "step": 194500
    },
    {
      "epoch": 0.1492,
      "grad_norm": 0.1887446641921997,
      "learning_rate": 0.0001221604,
      "loss": 0.26,
      "step": 194600
    },
    {
      "epoch": 0.1494,
      "grad_norm": 0.23812128603458405,
      "learning_rate": 0.0001221204,
      "loss": 0.2602,
      "step": 194700
    },
    {
      "epoch": 0.1496,
      "grad_norm": 0.2019837647676468,
      "learning_rate": 0.0001220804,
      "loss": 0.2614,
      "step": 194800
    },
    {
      "epoch": 0.1498,
      "grad_norm": 0.18969768285751343,
      "learning_rate": 0.0001220404,
      "loss": 0.2597,
      "step": 194900
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.2558167278766632,
      "learning_rate": 0.00012200040000000001,
      "loss": 0.2601,
      "step": 195000
    },
    {
      "epoch": 0.1502,
      "grad_norm": 0.16301903128623962,
      "learning_rate": 0.00012196039999999999,
      "loss": 0.2613,
      "step": 195100
    },
    {
      "epoch": 0.1504,
      "grad_norm": 0.19650568068027496,
      "learning_rate": 0.0001219204,
      "loss": 0.2643,
      "step": 195200
    },
    {
      "epoch": 0.1506,
      "grad_norm": 0.19182758033275604,
      "learning_rate": 0.0001218804,
      "loss": 0.2622,
      "step": 195300
    },
    {
      "epoch": 0.1508,
      "grad_norm": 0.16109509766101837,
      "learning_rate": 0.00012184040000000001,
      "loss": 0.2567,
      "step": 195400
    },
    {
      "epoch": 0.151,
      "grad_norm": 0.20243091881275177,
      "learning_rate": 0.00012180040000000002,
      "loss": 0.2561,
      "step": 195500
    },
    {
      "epoch": 0.1512,
      "grad_norm": 0.18910594284534454,
      "learning_rate": 0.0001217604,
      "loss": 0.2589,
      "step": 195600
    },
    {
      "epoch": 0.1514,
      "grad_norm": 0.17565961182117462,
      "learning_rate": 0.00012172040000000001,
      "loss": 0.2575,
      "step": 195700
    },
    {
      "epoch": 0.1516,
      "grad_norm": 0.18943890929222107,
      "learning_rate": 0.0001216804,
      "loss": 0.262,
      "step": 195800
    },
    {
      "epoch": 0.1518,
      "grad_norm": 0.1772325187921524,
      "learning_rate": 0.00012164040000000001,
      "loss": 0.2659,
      "step": 195900
    },
    {
      "epoch": 0.152,
      "grad_norm": 0.17579589784145355,
      "learning_rate": 0.00012160040000000002,
      "loss": 0.2611,
      "step": 196000
    },
    {
      "epoch": 0.1522,
      "grad_norm": 0.1820620447397232,
      "learning_rate": 0.0001215604,
      "loss": 0.2621,
      "step": 196100
    },
    {
      "epoch": 0.1524,
      "grad_norm": 0.20846767723560333,
      "learning_rate": 0.0001215204,
      "loss": 0.2607,
      "step": 196200
    },
    {
      "epoch": 0.1526,
      "grad_norm": 0.20666386187076569,
      "learning_rate": 0.00012148040000000001,
      "loss": 0.2587,
      "step": 196300
    },
    {
      "epoch": 0.1528,
      "grad_norm": 0.3281834125518799,
      "learning_rate": 0.0001214404,
      "loss": 0.2539,
      "step": 196400
    },
    {
      "epoch": 0.153,
      "grad_norm": 0.20208649337291718,
      "learning_rate": 0.00012140040000000001,
      "loss": 0.2578,
      "step": 196500
    },
    {
      "epoch": 0.1532,
      "grad_norm": 0.28220322728157043,
      "learning_rate": 0.00012136039999999999,
      "loss": 0.2558,
      "step": 196600
    },
    {
      "epoch": 0.1534,
      "grad_norm": 0.33225101232528687,
      "learning_rate": 0.0001213204,
      "loss": 0.2661,
      "step": 196700
    },
    {
      "epoch": 0.1536,
      "grad_norm": 0.24750415980815887,
      "learning_rate": 0.0001212804,
      "loss": 0.2624,
      "step": 196800
    },
    {
      "epoch": 0.1538,
      "grad_norm": 0.1799037754535675,
      "learning_rate": 0.00012124040000000001,
      "loss": 0.2584,
      "step": 196900
    },
    {
      "epoch": 0.154,
      "grad_norm": 0.20016971230506897,
      "learning_rate": 0.00012120040000000002,
      "loss": 0.2599,
      "step": 197000
    },
    {
      "epoch": 0.1542,
      "grad_norm": 0.18166713416576385,
      "learning_rate": 0.0001211604,
      "loss": 0.2567,
      "step": 197100
    },
    {
      "epoch": 0.1544,
      "grad_norm": 0.1939893662929535,
      "learning_rate": 0.0001211204,
      "loss": 0.2646,
      "step": 197200
    },
    {
      "epoch": 0.1546,
      "grad_norm": 0.18318818509578705,
      "learning_rate": 0.0001210804,
      "loss": 0.2552,
      "step": 197300
    },
    {
      "epoch": 0.1548,
      "grad_norm": 0.27068910002708435,
      "learning_rate": 0.00012104040000000001,
      "loss": 0.2594,
      "step": 197400
    },
    {
      "epoch": 0.155,
      "grad_norm": 0.20049890875816345,
      "learning_rate": 0.00012100040000000001,
      "loss": 0.2569,
      "step": 197500
    },
    {
      "epoch": 0.1552,
      "grad_norm": 0.1976662427186966,
      "learning_rate": 0.0001209604,
      "loss": 0.2544,
      "step": 197600
    },
    {
      "epoch": 0.1554,
      "grad_norm": 0.17978644371032715,
      "learning_rate": 0.0001209204,
      "loss": 0.2626,
      "step": 197700
    },
    {
      "epoch": 0.1556,
      "grad_norm": 0.18449051678180695,
      "learning_rate": 0.00012088040000000001,
      "loss": 0.2609,
      "step": 197800
    },
    {
      "epoch": 0.1558,
      "grad_norm": 0.21777702867984772,
      "learning_rate": 0.0001208404,
      "loss": 0.2557,
      "step": 197900
    },
    {
      "epoch": 0.156,
      "grad_norm": 0.34685173630714417,
      "learning_rate": 0.00012080040000000001,
      "loss": 0.2598,
      "step": 198000
    },
    {
      "epoch": 0.1562,
      "grad_norm": 0.21102797985076904,
      "learning_rate": 0.00012076039999999999,
      "loss": 0.258,
      "step": 198100
    },
    {
      "epoch": 0.1564,
      "grad_norm": 0.1801084280014038,
      "learning_rate": 0.0001207204,
      "loss": 0.2545,
      "step": 198200
    },
    {
      "epoch": 0.1566,
      "grad_norm": 0.24270586669445038,
      "learning_rate": 0.0001206804,
      "loss": 0.2594,
      "step": 198300
    },
    {
      "epoch": 0.1568,
      "grad_norm": 0.1699536144733429,
      "learning_rate": 0.00012064040000000001,
      "loss": 0.2665,
      "step": 198400
    },
    {
      "epoch": 0.157,
      "grad_norm": 0.15559914708137512,
      "learning_rate": 0.00012060040000000002,
      "loss": 0.2577,
      "step": 198500
    },
    {
      "epoch": 0.1572,
      "grad_norm": 0.1730017513036728,
      "learning_rate": 0.0001205604,
      "loss": 0.2578,
      "step": 198600
    },
    {
      "epoch": 0.1574,
      "grad_norm": 0.18403738737106323,
      "learning_rate": 0.0001205204,
      "loss": 0.257,
      "step": 198700
    },
    {
      "epoch": 0.1576,
      "grad_norm": 0.17950795590877533,
      "learning_rate": 0.0001204804,
      "loss": 0.2581,
      "step": 198800
    },
    {
      "epoch": 0.1578,
      "grad_norm": 0.18229971826076508,
      "learning_rate": 0.0001204404,
      "loss": 0.2611,
      "step": 198900
    },
    {
      "epoch": 0.158,
      "grad_norm": 0.1665981113910675,
      "learning_rate": 0.00012040040000000001,
      "loss": 0.2562,
      "step": 199000
    },
    {
      "epoch": 0.1582,
      "grad_norm": 0.1657390296459198,
      "learning_rate": 0.0001203604,
      "loss": 0.2629,
      "step": 199100
    },
    {
      "epoch": 0.1584,
      "grad_norm": 0.18978402018547058,
      "learning_rate": 0.0001203204,
      "loss": 0.2613,
      "step": 199200
    },
    {
      "epoch": 0.1586,
      "grad_norm": 0.18768629431724548,
      "learning_rate": 0.00012028040000000001,
      "loss": 0.2564,
      "step": 199300
    },
    {
      "epoch": 0.1588,
      "grad_norm": 0.3750452697277069,
      "learning_rate": 0.00012024040000000002,
      "loss": 0.2577,
      "step": 199400
    },
    {
      "epoch": 0.159,
      "grad_norm": 0.25771307945251465,
      "learning_rate": 0.00012020040000000001,
      "loss": 0.2642,
      "step": 199500
    },
    {
      "epoch": 0.1592,
      "grad_norm": 0.27435046434402466,
      "learning_rate": 0.00012016039999999999,
      "loss": 0.263,
      "step": 199600
    },
    {
      "epoch": 0.1594,
      "grad_norm": 0.1829531192779541,
      "learning_rate": 0.0001201204,
      "loss": 0.2606,
      "step": 199700
    },
    {
      "epoch": 0.1596,
      "grad_norm": 0.14673271775245667,
      "learning_rate": 0.0001200804,
      "loss": 0.2575,
      "step": 199800
    },
    {
      "epoch": 0.1598,
      "grad_norm": 0.19495797157287598,
      "learning_rate": 0.00012004040000000001,
      "loss": 0.2545,
      "step": 199900
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.20886088907718658,
      "learning_rate": 0.00012000040000000002,
      "loss": 0.2584,
      "step": 200000
    },
    {
      "epoch": 0.1602,
      "grad_norm": 0.18301694095134735,
      "learning_rate": 0.0001199604,
      "loss": 0.2581,
      "step": 200100
    },
    {
      "epoch": 0.1604,
      "grad_norm": 0.1781015843153,
      "learning_rate": 0.0001199204,
      "loss": 0.2583,
      "step": 200200
    },
    {
      "epoch": 0.1606,
      "grad_norm": 0.15711505711078644,
      "learning_rate": 0.0001198804,
      "loss": 0.2594,
      "step": 200300
    },
    {
      "epoch": 0.1608,
      "grad_norm": 0.21567769348621368,
      "learning_rate": 0.0001198404,
      "loss": 0.2553,
      "step": 200400
    },
    {
      "epoch": 0.161,
      "grad_norm": 0.18019573390483856,
      "learning_rate": 0.00011980040000000001,
      "loss": 0.2653,
      "step": 200500
    },
    {
      "epoch": 0.1612,
      "grad_norm": 0.1789156198501587,
      "learning_rate": 0.00011976039999999999,
      "loss": 0.2604,
      "step": 200600
    },
    {
      "epoch": 0.1614,
      "grad_norm": 0.26856061816215515,
      "learning_rate": 0.0001197204,
      "loss": 0.2625,
      "step": 200700
    },
    {
      "epoch": 0.1616,
      "grad_norm": 0.16721145808696747,
      "learning_rate": 0.00011968040000000001,
      "loss": 0.2628,
      "step": 200800
    },
    {
      "epoch": 0.1618,
      "grad_norm": 0.1759210228919983,
      "learning_rate": 0.00011964040000000001,
      "loss": 0.2534,
      "step": 200900
    },
    {
      "epoch": 0.162,
      "grad_norm": 0.15787266194820404,
      "learning_rate": 0.00011960040000000001,
      "loss": 0.2586,
      "step": 201000
    },
    {
      "epoch": 0.1622,
      "grad_norm": 0.1680240035057068,
      "learning_rate": 0.0001195604,
      "loss": 0.2583,
      "step": 201100
    },
    {
      "epoch": 0.1624,
      "grad_norm": 0.19923582673072815,
      "learning_rate": 0.0001195204,
      "loss": 0.2637,
      "step": 201200
    },
    {
      "epoch": 0.1626,
      "grad_norm": 0.33889421820640564,
      "learning_rate": 0.0001194804,
      "loss": 0.261,
      "step": 201300
    },
    {
      "epoch": 0.1628,
      "grad_norm": 0.22548066079616547,
      "learning_rate": 0.00011944040000000001,
      "loss": 0.2557,
      "step": 201400
    },
    {
      "epoch": 0.163,
      "grad_norm": 0.2182902842760086,
      "learning_rate": 0.00011940040000000002,
      "loss": 0.2593,
      "step": 201500
    },
    {
      "epoch": 0.1632,
      "grad_norm": 0.27033495903015137,
      "learning_rate": 0.00011936040000000001,
      "loss": 0.2588,
      "step": 201600
    },
    {
      "epoch": 0.1634,
      "grad_norm": 0.3093793988227844,
      "learning_rate": 0.0001193204,
      "loss": 0.2561,
      "step": 201700
    },
    {
      "epoch": 0.1636,
      "grad_norm": 0.20741653442382812,
      "learning_rate": 0.0001192804,
      "loss": 0.2623,
      "step": 201800
    },
    {
      "epoch": 0.1638,
      "grad_norm": 0.2333420366048813,
      "learning_rate": 0.0001192404,
      "loss": 0.2574,
      "step": 201900
    },
    {
      "epoch": 0.164,
      "grad_norm": 0.20007722079753876,
      "learning_rate": 0.00011920040000000001,
      "loss": 0.2618,
      "step": 202000
    },
    {
      "epoch": 0.1642,
      "grad_norm": 0.18958450853824615,
      "learning_rate": 0.00011916040000000002,
      "loss": 0.2545,
      "step": 202100
    },
    {
      "epoch": 0.1644,
      "grad_norm": 0.22179192304611206,
      "learning_rate": 0.0001191204,
      "loss": 0.2637,
      "step": 202200
    },
    {
      "epoch": 0.1646,
      "grad_norm": 0.1823706179857254,
      "learning_rate": 0.0001190804,
      "loss": 0.2567,
      "step": 202300
    },
    {
      "epoch": 0.1648,
      "grad_norm": 0.19839805364608765,
      "learning_rate": 0.00011904040000000001,
      "loss": 0.2582,
      "step": 202400
    },
    {
      "epoch": 0.165,
      "grad_norm": 0.1794578582048416,
      "learning_rate": 0.0001190004,
      "loss": 0.2563,
      "step": 202500
    },
    {
      "epoch": 0.1652,
      "grad_norm": 0.1632799506187439,
      "learning_rate": 0.00011896040000000001,
      "loss": 0.2544,
      "step": 202600
    },
    {
      "epoch": 0.1654,
      "grad_norm": 0.14698299765586853,
      "learning_rate": 0.0001189204,
      "loss": 0.2583,
      "step": 202700
    },
    {
      "epoch": 0.1656,
      "grad_norm": 0.1564948856830597,
      "learning_rate": 0.0001188804,
      "loss": 0.2544,
      "step": 202800
    },
    {
      "epoch": 0.1658,
      "grad_norm": 0.27698883414268494,
      "learning_rate": 0.00011884040000000001,
      "loss": 0.2603,
      "step": 202900
    },
    {
      "epoch": 0.166,
      "grad_norm": 0.17376329004764557,
      "learning_rate": 0.00011880040000000002,
      "loss": 0.2569,
      "step": 203000
    },
    {
      "epoch": 0.1662,
      "grad_norm": 0.21339082717895508,
      "learning_rate": 0.00011876040000000002,
      "loss": 0.2554,
      "step": 203100
    },
    {
      "epoch": 0.1664,
      "grad_norm": 0.19938743114471436,
      "learning_rate": 0.0001187204,
      "loss": 0.2703,
      "step": 203200
    },
    {
      "epoch": 0.1666,
      "grad_norm": 0.1865181177854538,
      "learning_rate": 0.0001186804,
      "loss": 0.2559,
      "step": 203300
    },
    {
      "epoch": 0.1668,
      "grad_norm": 0.18764150142669678,
      "learning_rate": 0.0001186404,
      "loss": 0.2565,
      "step": 203400
    },
    {
      "epoch": 0.167,
      "grad_norm": 0.16691163182258606,
      "learning_rate": 0.00011860040000000001,
      "loss": 0.2616,
      "step": 203500
    },
    {
      "epoch": 0.1672,
      "grad_norm": 0.20289692282676697,
      "learning_rate": 0.00011856040000000002,
      "loss": 0.2585,
      "step": 203600
    },
    {
      "epoch": 0.1674,
      "grad_norm": 0.2008928805589676,
      "learning_rate": 0.0001185204,
      "loss": 0.2654,
      "step": 203700
    },
    {
      "epoch": 0.1676,
      "grad_norm": 0.29621559381484985,
      "learning_rate": 0.0001184804,
      "loss": 0.2649,
      "step": 203800
    },
    {
      "epoch": 0.1678,
      "grad_norm": 0.23792728781700134,
      "learning_rate": 0.00011844040000000001,
      "loss": 0.2606,
      "step": 203900
    },
    {
      "epoch": 0.168,
      "grad_norm": 0.1951097548007965,
      "learning_rate": 0.0001184004,
      "loss": 0.2603,
      "step": 204000
    },
    {
      "epoch": 0.1682,
      "grad_norm": 0.18603862822055817,
      "learning_rate": 0.00011836040000000001,
      "loss": 0.2567,
      "step": 204100
    },
    {
      "epoch": 0.1684,
      "grad_norm": 0.248612180352211,
      "learning_rate": 0.00011832039999999999,
      "loss": 0.2588,
      "step": 204200
    },
    {
      "epoch": 0.1686,
      "grad_norm": 0.3616507351398468,
      "learning_rate": 0.0001182804,
      "loss": 0.2568,
      "step": 204300
    },
    {
      "epoch": 0.1688,
      "grad_norm": 0.22224590182304382,
      "learning_rate": 0.00011824040000000001,
      "loss": 0.2589,
      "step": 204400
    },
    {
      "epoch": 0.169,
      "grad_norm": 0.22203755378723145,
      "learning_rate": 0.00011820040000000001,
      "loss": 0.2572,
      "step": 204500
    },
    {
      "epoch": 0.1692,
      "grad_norm": 0.16411060094833374,
      "learning_rate": 0.00011816040000000002,
      "loss": 0.2578,
      "step": 204600
    },
    {
      "epoch": 0.1694,
      "grad_norm": 0.17382003366947174,
      "learning_rate": 0.0001181204,
      "loss": 0.2546,
      "step": 204700
    },
    {
      "epoch": 0.1696,
      "grad_norm": 0.21096429228782654,
      "learning_rate": 0.00011808040000000001,
      "loss": 0.2619,
      "step": 204800
    },
    {
      "epoch": 0.1698,
      "grad_norm": 0.2046499401330948,
      "learning_rate": 0.0001180404,
      "loss": 0.2538,
      "step": 204900
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.1903340220451355,
      "learning_rate": 0.00011800040000000001,
      "loss": 0.2695,
      "step": 205000
    },
    {
      "epoch": 0.1702,
      "grad_norm": 0.19105155766010284,
      "learning_rate": 0.00011796040000000002,
      "loss": 0.2592,
      "step": 205100
    },
    {
      "epoch": 0.1704,
      "grad_norm": 0.18848542869091034,
      "learning_rate": 0.0001179204,
      "loss": 0.2607,
      "step": 205200
    },
    {
      "epoch": 0.1706,
      "grad_norm": 0.19775767624378204,
      "learning_rate": 0.0001178804,
      "loss": 0.2514,
      "step": 205300
    },
    {
      "epoch": 0.1708,
      "grad_norm": 0.2048797607421875,
      "learning_rate": 0.00011784040000000001,
      "loss": 0.2584,
      "step": 205400
    },
    {
      "epoch": 0.171,
      "grad_norm": 0.17378486692905426,
      "learning_rate": 0.0001178004,
      "loss": 0.2611,
      "step": 205500
    },
    {
      "epoch": 0.1712,
      "grad_norm": 0.19337409734725952,
      "learning_rate": 0.00011776040000000001,
      "loss": 0.2552,
      "step": 205600
    },
    {
      "epoch": 0.1714,
      "grad_norm": 0.1871696412563324,
      "learning_rate": 0.00011772039999999999,
      "loss": 0.2555,
      "step": 205700
    },
    {
      "epoch": 0.1716,
      "grad_norm": 0.15516479313373566,
      "learning_rate": 0.0001176804,
      "loss": 0.255,
      "step": 205800
    },
    {
      "epoch": 0.1718,
      "grad_norm": 0.18520411849021912,
      "learning_rate": 0.0001176404,
      "loss": 0.2607,
      "step": 205900
    },
    {
      "epoch": 0.172,
      "grad_norm": 0.1725122481584549,
      "learning_rate": 0.00011760040000000001,
      "loss": 0.2599,
      "step": 206000
    },
    {
      "epoch": 0.1722,
      "grad_norm": 0.24798348546028137,
      "learning_rate": 0.00011756040000000002,
      "loss": 0.2654,
      "step": 206100
    },
    {
      "epoch": 0.1724,
      "grad_norm": 0.15769024193286896,
      "learning_rate": 0.0001175204,
      "loss": 0.2753,
      "step": 206200
    },
    {
      "epoch": 0.1726,
      "grad_norm": 0.2662038207054138,
      "learning_rate": 0.00011748040000000001,
      "loss": 0.2642,
      "step": 206300
    },
    {
      "epoch": 0.1728,
      "grad_norm": 0.22056128084659576,
      "learning_rate": 0.0001174404,
      "loss": 0.2573,
      "step": 206400
    },
    {
      "epoch": 0.173,
      "grad_norm": 0.24194368720054626,
      "learning_rate": 0.00011740040000000001,
      "loss": 0.2605,
      "step": 206500
    },
    {
      "epoch": 0.1732,
      "grad_norm": 0.2201591581106186,
      "learning_rate": 0.00011736040000000002,
      "loss": 0.26,
      "step": 206600
    },
    {
      "epoch": 0.1734,
      "grad_norm": 0.17984557151794434,
      "learning_rate": 0.0001173204,
      "loss": 0.2588,
      "step": 206700
    },
    {
      "epoch": 0.1736,
      "grad_norm": 0.2862537205219269,
      "learning_rate": 0.0001172804,
      "loss": 0.2656,
      "step": 206800
    },
    {
      "epoch": 0.1738,
      "grad_norm": 0.20301680266857147,
      "learning_rate": 0.00011724040000000001,
      "loss": 0.2596,
      "step": 206900
    },
    {
      "epoch": 0.174,
      "grad_norm": 0.16194941103458405,
      "learning_rate": 0.0001172004,
      "loss": 0.2604,
      "step": 207000
    },
    {
      "epoch": 0.1742,
      "grad_norm": 0.20305371284484863,
      "learning_rate": 0.00011716040000000001,
      "loss": 0.2605,
      "step": 207100
    },
    {
      "epoch": 0.1744,
      "grad_norm": 0.17961016297340393,
      "learning_rate": 0.00011712039999999999,
      "loss": 0.2553,
      "step": 207200
    },
    {
      "epoch": 0.1746,
      "grad_norm": 0.15147832036018372,
      "learning_rate": 0.0001170804,
      "loss": 0.2608,
      "step": 207300
    },
    {
      "epoch": 0.1748,
      "grad_norm": 0.17695245146751404,
      "learning_rate": 0.0001170404,
      "loss": 0.2577,
      "step": 207400
    },
    {
      "epoch": 0.175,
      "grad_norm": 0.16684728860855103,
      "learning_rate": 0.00011700040000000001,
      "loss": 0.2602,
      "step": 207500
    },
    {
      "epoch": 0.1752,
      "grad_norm": 0.1773560494184494,
      "learning_rate": 0.00011696040000000002,
      "loss": 0.2592,
      "step": 207600
    },
    {
      "epoch": 0.1754,
      "grad_norm": 0.1829720139503479,
      "learning_rate": 0.0001169204,
      "loss": 0.2533,
      "step": 207700
    },
    {
      "epoch": 0.1756,
      "grad_norm": 0.14604096114635468,
      "learning_rate": 0.0001168804,
      "loss": 0.2559,
      "step": 207800
    },
    {
      "epoch": 0.1758,
      "grad_norm": 0.16805389523506165,
      "learning_rate": 0.0001168404,
      "loss": 0.2565,
      "step": 207900
    },
    {
      "epoch": 0.176,
      "grad_norm": 0.2917552888393402,
      "learning_rate": 0.00011680040000000001,
      "loss": 0.2596,
      "step": 208000
    },
    {
      "epoch": 0.1762,
      "grad_norm": 0.1552281528711319,
      "learning_rate": 0.00011676040000000001,
      "loss": 0.2595,
      "step": 208100
    },
    {
      "epoch": 0.1764,
      "grad_norm": 0.21594230830669403,
      "learning_rate": 0.0001167204,
      "loss": 0.26,
      "step": 208200
    },
    {
      "epoch": 0.1766,
      "grad_norm": 0.2598268389701843,
      "learning_rate": 0.0001166804,
      "loss": 0.2634,
      "step": 208300
    },
    {
      "epoch": 0.1768,
      "grad_norm": 0.17285478115081787,
      "learning_rate": 0.00011664040000000001,
      "loss": 0.2663,
      "step": 208400
    },
    {
      "epoch": 0.177,
      "grad_norm": 0.2128245234489441,
      "learning_rate": 0.00011660040000000002,
      "loss": 0.2596,
      "step": 208500
    },
    {
      "epoch": 0.1772,
      "grad_norm": 0.17592047154903412,
      "learning_rate": 0.00011656040000000001,
      "loss": 0.259,
      "step": 208600
    },
    {
      "epoch": 0.1774,
      "grad_norm": 0.1863638013601303,
      "learning_rate": 0.00011652039999999999,
      "loss": 0.2574,
      "step": 208700
    },
    {
      "epoch": 0.1776,
      "grad_norm": 0.1951245218515396,
      "learning_rate": 0.0001164804,
      "loss": 0.2569,
      "step": 208800
    },
    {
      "epoch": 0.1778,
      "grad_norm": 0.20469717681407928,
      "learning_rate": 0.0001164404,
      "loss": 0.2548,
      "step": 208900
    },
    {
      "epoch": 0.178,
      "grad_norm": 0.18521174788475037,
      "learning_rate": 0.00011640040000000001,
      "loss": 0.2581,
      "step": 209000
    },
    {
      "epoch": 0.1782,
      "grad_norm": 0.16419340670108795,
      "learning_rate": 0.00011636040000000002,
      "loss": 0.2557,
      "step": 209100
    },
    {
      "epoch": 0.1784,
      "grad_norm": 0.2025509625673294,
      "learning_rate": 0.0001163204,
      "loss": 0.2553,
      "step": 209200
    },
    {
      "epoch": 0.1786,
      "grad_norm": 0.2792588174343109,
      "learning_rate": 0.0001162804,
      "loss": 0.2554,
      "step": 209300
    },
    {
      "epoch": 0.1788,
      "grad_norm": 0.18840503692626953,
      "learning_rate": 0.0001162404,
      "loss": 0.2612,
      "step": 209400
    },
    {
      "epoch": 0.179,
      "grad_norm": 0.3353501558303833,
      "learning_rate": 0.0001162004,
      "loss": 0.2575,
      "step": 209500
    },
    {
      "epoch": 0.1792,
      "grad_norm": 0.17007525265216827,
      "learning_rate": 0.00011616040000000001,
      "loss": 0.2534,
      "step": 209600
    },
    {
      "epoch": 0.1794,
      "grad_norm": 0.1712578982114792,
      "learning_rate": 0.0001161204,
      "loss": 0.2579,
      "step": 209700
    },
    {
      "epoch": 0.1796,
      "grad_norm": 0.16284547746181488,
      "learning_rate": 0.0001160804,
      "loss": 0.2578,
      "step": 209800
    },
    {
      "epoch": 0.1798,
      "grad_norm": 0.22904619574546814,
      "learning_rate": 0.00011604040000000001,
      "loss": 0.2554,
      "step": 209900
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.1613014191389084,
      "learning_rate": 0.00011600040000000002,
      "loss": 0.2569,
      "step": 210000
    },
    {
      "epoch": 0.1802,
      "grad_norm": 0.24920882284641266,
      "learning_rate": 0.00011596040000000001,
      "loss": 0.2523,
      "step": 210100
    },
    {
      "epoch": 0.1804,
      "grad_norm": 0.16219836473464966,
      "learning_rate": 0.0001159204,
      "loss": 0.257,
      "step": 210200
    },
    {
      "epoch": 0.1806,
      "grad_norm": 0.14738017320632935,
      "learning_rate": 0.0001158804,
      "loss": 0.2534,
      "step": 210300
    },
    {
      "epoch": 0.1808,
      "grad_norm": 0.16476276516914368,
      "learning_rate": 0.0001158404,
      "loss": 0.261,
      "step": 210400
    },
    {
      "epoch": 0.181,
      "grad_norm": 0.20898793637752533,
      "learning_rate": 0.00011580040000000001,
      "loss": 0.261,
      "step": 210500
    },
    {
      "epoch": 0.1812,
      "grad_norm": 0.16828548908233643,
      "learning_rate": 0.00011576040000000002,
      "loss": 0.2502,
      "step": 210600
    },
    {
      "epoch": 0.1814,
      "grad_norm": 0.17808620631694794,
      "learning_rate": 0.0001157204,
      "loss": 0.2445,
      "step": 210700
    },
    {
      "epoch": 0.1816,
      "grad_norm": 0.16437947750091553,
      "learning_rate": 0.0001156804,
      "loss": 0.2521,
      "step": 210800
    },
    {
      "epoch": 0.1818,
      "grad_norm": 0.2212170660495758,
      "learning_rate": 0.0001156404,
      "loss": 0.2532,
      "step": 210900
    },
    {
      "epoch": 0.182,
      "grad_norm": 0.22296178340911865,
      "learning_rate": 0.0001156004,
      "loss": 0.2594,
      "step": 211000
    },
    {
      "epoch": 0.1822,
      "grad_norm": 0.19012941420078278,
      "learning_rate": 0.00011556040000000001,
      "loss": 0.2601,
      "step": 211100
    },
    {
      "epoch": 0.1824,
      "grad_norm": 0.2223205268383026,
      "learning_rate": 0.00011552039999999999,
      "loss": 0.2571,
      "step": 211200
    },
    {
      "epoch": 0.1826,
      "grad_norm": 0.193226620554924,
      "learning_rate": 0.0001154804,
      "loss": 0.2635,
      "step": 211300
    },
    {
      "epoch": 0.1828,
      "grad_norm": 0.19742873311042786,
      "learning_rate": 0.00011544040000000001,
      "loss": 0.2635,
      "step": 211400
    },
    {
      "epoch": 0.183,
      "grad_norm": 0.2180163711309433,
      "learning_rate": 0.00011540040000000001,
      "loss": 0.2669,
      "step": 211500
    },
    {
      "epoch": 0.1832,
      "grad_norm": 0.20754875242710114,
      "learning_rate": 0.00011536040000000001,
      "loss": 0.252,
      "step": 211600
    },
    {
      "epoch": 0.1834,
      "grad_norm": 0.16631507873535156,
      "learning_rate": 0.0001153204,
      "loss": 0.256,
      "step": 211700
    },
    {
      "epoch": 0.1836,
      "grad_norm": 0.1716802716255188,
      "learning_rate": 0.0001152804,
      "loss": 0.2578,
      "step": 211800
    },
    {
      "epoch": 0.1838,
      "grad_norm": 0.22143660485744476,
      "learning_rate": 0.0001152404,
      "loss": 0.2566,
      "step": 211900
    },
    {
      "epoch": 0.184,
      "grad_norm": 0.2684420347213745,
      "learning_rate": 0.00011520040000000001,
      "loss": 0.2515,
      "step": 212000
    },
    {
      "epoch": 0.1842,
      "grad_norm": 0.22609145939350128,
      "learning_rate": 0.00011516040000000002,
      "loss": 0.2532,
      "step": 212100
    },
    {
      "epoch": 0.1844,
      "grad_norm": 0.1780252754688263,
      "learning_rate": 0.0001151204,
      "loss": 0.2579,
      "step": 212200
    },
    {
      "epoch": 0.1846,
      "grad_norm": 0.256816029548645,
      "learning_rate": 0.0001150804,
      "loss": 0.2628,
      "step": 212300
    },
    {
      "epoch": 0.1848,
      "grad_norm": 0.22887232899665833,
      "learning_rate": 0.0001150404,
      "loss": 0.254,
      "step": 212400
    },
    {
      "epoch": 0.185,
      "grad_norm": 0.17870065569877625,
      "learning_rate": 0.0001150004,
      "loss": 0.2525,
      "step": 212500
    },
    {
      "epoch": 0.1852,
      "grad_norm": 0.4473213255405426,
      "learning_rate": 0.00011496040000000001,
      "loss": 0.2565,
      "step": 212600
    },
    {
      "epoch": 0.1854,
      "grad_norm": 0.1810619980096817,
      "learning_rate": 0.00011492039999999999,
      "loss": 0.2565,
      "step": 212700
    },
    {
      "epoch": 0.1856,
      "grad_norm": 0.1680411398410797,
      "learning_rate": 0.0001148804,
      "loss": 0.2555,
      "step": 212800
    },
    {
      "epoch": 0.1858,
      "grad_norm": 0.22519907355308533,
      "learning_rate": 0.0001148404,
      "loss": 0.252,
      "step": 212900
    },
    {
      "epoch": 0.186,
      "grad_norm": 0.18328481912612915,
      "learning_rate": 0.00011480040000000001,
      "loss": 0.2567,
      "step": 213000
    },
    {
      "epoch": 0.1862,
      "grad_norm": 0.4796239733695984,
      "learning_rate": 0.0001147604,
      "loss": 0.2552,
      "step": 213100
    },
    {
      "epoch": 0.1864,
      "grad_norm": 0.17960181832313538,
      "learning_rate": 0.0001147204,
      "loss": 0.2552,
      "step": 213200
    },
    {
      "epoch": 0.1866,
      "grad_norm": 0.21092750132083893,
      "learning_rate": 0.0001146804,
      "loss": 0.2591,
      "step": 213300
    },
    {
      "epoch": 0.1868,
      "grad_norm": 0.1919945329427719,
      "learning_rate": 0.0001146404,
      "loss": 0.2524,
      "step": 213400
    },
    {
      "epoch": 0.187,
      "grad_norm": 0.1630232334136963,
      "learning_rate": 0.00011460040000000001,
      "loss": 0.2548,
      "step": 213500
    },
    {
      "epoch": 0.1872,
      "grad_norm": 0.15087801218032837,
      "learning_rate": 0.00011456040000000002,
      "loss": 0.2595,
      "step": 213600
    },
    {
      "epoch": 0.1874,
      "grad_norm": 0.31527626514434814,
      "learning_rate": 0.00011452040000000002,
      "loss": 0.2553,
      "step": 213700
    },
    {
      "epoch": 0.1876,
      "grad_norm": 0.27339211106300354,
      "learning_rate": 0.0001144804,
      "loss": 0.2617,
      "step": 213800
    },
    {
      "epoch": 0.1878,
      "grad_norm": 0.19266119599342346,
      "learning_rate": 0.00011444040000000001,
      "loss": 0.2628,
      "step": 213900
    },
    {
      "epoch": 0.188,
      "grad_norm": 0.26474958658218384,
      "learning_rate": 0.0001144004,
      "loss": 0.2533,
      "step": 214000
    },
    {
      "epoch": 0.1882,
      "grad_norm": 0.16048365831375122,
      "learning_rate": 0.00011436040000000001,
      "loss": 0.261,
      "step": 214100
    },
    {
      "epoch": 0.1884,
      "grad_norm": 0.20829501748085022,
      "learning_rate": 0.00011432040000000002,
      "loss": 0.2562,
      "step": 214200
    },
    {
      "epoch": 0.1886,
      "grad_norm": 0.2848805785179138,
      "learning_rate": 0.0001142804,
      "loss": 0.2586,
      "step": 214300
    },
    {
      "epoch": 0.1888,
      "grad_norm": 0.2071709930896759,
      "learning_rate": 0.0001142404,
      "loss": 0.2534,
      "step": 214400
    },
    {
      "epoch": 0.189,
      "grad_norm": 0.20445095002651215,
      "learning_rate": 0.00011420040000000001,
      "loss": 0.2553,
      "step": 214500
    },
    {
      "epoch": 0.1892,
      "grad_norm": 0.1799619346857071,
      "learning_rate": 0.0001141604,
      "loss": 0.2571,
      "step": 214600
    },
    {
      "epoch": 0.1894,
      "grad_norm": 0.23064669966697693,
      "learning_rate": 0.00011412040000000001,
      "loss": 0.2581,
      "step": 214700
    },
    {
      "epoch": 0.1896,
      "grad_norm": 0.20549724996089935,
      "learning_rate": 0.00011408039999999999,
      "loss": 0.2563,
      "step": 214800
    },
    {
      "epoch": 0.1898,
      "grad_norm": 0.3230568766593933,
      "learning_rate": 0.0001140404,
      "loss": 0.2549,
      "step": 214900
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.15154241025447845,
      "learning_rate": 0.00011400040000000001,
      "loss": 0.256,
      "step": 215000
    },
    {
      "epoch": 0.1902,
      "grad_norm": 0.18913942575454712,
      "learning_rate": 0.00011396040000000001,
      "loss": 0.2632,
      "step": 215100
    },
    {
      "epoch": 0.1904,
      "grad_norm": 0.19355298578739166,
      "learning_rate": 0.00011392040000000002,
      "loss": 0.2577,
      "step": 215200
    },
    {
      "epoch": 0.1906,
      "grad_norm": 0.16174174845218658,
      "learning_rate": 0.0001138804,
      "loss": 0.2504,
      "step": 215300
    },
    {
      "epoch": 0.1908,
      "grad_norm": 0.24760814011096954,
      "learning_rate": 0.00011384040000000001,
      "loss": 0.2541,
      "step": 215400
    },
    {
      "epoch": 0.191,
      "grad_norm": 0.19982276856899261,
      "learning_rate": 0.0001138004,
      "loss": 0.2553,
      "step": 215500
    },
    {
      "epoch": 0.1912,
      "grad_norm": 0.22562120854854584,
      "learning_rate": 0.00011376040000000001,
      "loss": 0.2564,
      "step": 215600
    },
    {
      "epoch": 0.1914,
      "grad_norm": 0.2680492401123047,
      "learning_rate": 0.00011372040000000002,
      "loss": 0.2582,
      "step": 215700
    },
    {
      "epoch": 0.1916,
      "grad_norm": 0.22115637362003326,
      "learning_rate": 0.0001136804,
      "loss": 0.2576,
      "step": 215800
    },
    {
      "epoch": 0.1918,
      "grad_norm": 0.19587989151477814,
      "learning_rate": 0.0001136404,
      "loss": 0.2566,
      "step": 215900
    },
    {
      "epoch": 0.192,
      "grad_norm": 0.25761643052101135,
      "learning_rate": 0.00011360040000000001,
      "loss": 0.2572,
      "step": 216000
    },
    {
      "epoch": 0.1922,
      "grad_norm": 0.1627892553806305,
      "learning_rate": 0.0001135604,
      "loss": 0.255,
      "step": 216100
    },
    {
      "epoch": 0.1924,
      "grad_norm": 0.18419532477855682,
      "learning_rate": 0.00011352040000000001,
      "loss": 0.2558,
      "step": 216200
    },
    {
      "epoch": 0.1926,
      "grad_norm": 0.18171828985214233,
      "learning_rate": 0.00011348039999999999,
      "loss": 0.2548,
      "step": 216300
    },
    {
      "epoch": 0.1928,
      "grad_norm": 0.16602705419063568,
      "learning_rate": 0.0001134404,
      "loss": 0.2542,
      "step": 216400
    },
    {
      "epoch": 0.193,
      "grad_norm": 0.17187577486038208,
      "learning_rate": 0.0001134004,
      "loss": 0.2559,
      "step": 216500
    },
    {
      "epoch": 0.1932,
      "grad_norm": 0.2111077755689621,
      "learning_rate": 0.00011336040000000001,
      "loss": 0.2578,
      "step": 216600
    },
    {
      "epoch": 0.1934,
      "grad_norm": 0.1766299456357956,
      "learning_rate": 0.00011332040000000002,
      "loss": 0.254,
      "step": 216700
    },
    {
      "epoch": 0.1936,
      "grad_norm": 0.2486806958913803,
      "learning_rate": 0.0001132804,
      "loss": 0.2642,
      "step": 216800
    },
    {
      "epoch": 0.1938,
      "grad_norm": 0.1899750679731369,
      "learning_rate": 0.00011324040000000001,
      "loss": 0.2542,
      "step": 216900
    },
    {
      "epoch": 0.194,
      "grad_norm": 0.18645033240318298,
      "learning_rate": 0.0001132004,
      "loss": 0.2589,
      "step": 217000
    },
    {
      "epoch": 0.1942,
      "grad_norm": 0.1879652589559555,
      "learning_rate": 0.00011316040000000001,
      "loss": 0.2623,
      "step": 217100
    },
    {
      "epoch": 0.1944,
      "grad_norm": 0.21320343017578125,
      "learning_rate": 0.00011312040000000002,
      "loss": 0.2572,
      "step": 217200
    },
    {
      "epoch": 0.1946,
      "grad_norm": 0.19344399869441986,
      "learning_rate": 0.0001130804,
      "loss": 0.2525,
      "step": 217300
    },
    {
      "epoch": 0.1948,
      "grad_norm": 0.16314131021499634,
      "learning_rate": 0.0001130404,
      "loss": 0.2532,
      "step": 217400
    },
    {
      "epoch": 0.195,
      "grad_norm": 0.153633713722229,
      "learning_rate": 0.00011300040000000001,
      "loss": 0.2613,
      "step": 217500
    },
    {
      "epoch": 0.1952,
      "grad_norm": 0.231405109167099,
      "learning_rate": 0.00011296040000000002,
      "loss": 0.2613,
      "step": 217600
    },
    {
      "epoch": 0.1954,
      "grad_norm": 0.17982491850852966,
      "learning_rate": 0.00011292040000000001,
      "loss": 0.2568,
      "step": 217700
    },
    {
      "epoch": 0.1956,
      "grad_norm": 0.21746788918972015,
      "learning_rate": 0.00011288039999999999,
      "loss": 0.2552,
      "step": 217800
    },
    {
      "epoch": 0.1958,
      "grad_norm": 0.1724511682987213,
      "learning_rate": 0.0001128404,
      "loss": 0.2593,
      "step": 217900
    },
    {
      "epoch": 0.196,
      "grad_norm": 0.1525942087173462,
      "learning_rate": 0.0001128004,
      "loss": 0.2647,
      "step": 218000
    },
    {
      "epoch": 0.1962,
      "grad_norm": 0.20681273937225342,
      "learning_rate": 0.00011276040000000001,
      "loss": 0.2569,
      "step": 218100
    },
    {
      "epoch": 0.1964,
      "grad_norm": 0.25222960114479065,
      "learning_rate": 0.00011272040000000002,
      "loss": 0.256,
      "step": 218200
    },
    {
      "epoch": 0.1966,
      "grad_norm": 0.19064846634864807,
      "learning_rate": 0.0001126804,
      "loss": 0.2507,
      "step": 218300
    },
    {
      "epoch": 0.1968,
      "grad_norm": 0.1997394561767578,
      "learning_rate": 0.0001126404,
      "loss": 0.2612,
      "step": 218400
    },
    {
      "epoch": 0.197,
      "grad_norm": 0.1820058673620224,
      "learning_rate": 0.0001126004,
      "loss": 0.2559,
      "step": 218500
    },
    {
      "epoch": 0.1972,
      "grad_norm": 0.23284319043159485,
      "learning_rate": 0.00011256040000000001,
      "loss": 0.2578,
      "step": 218600
    },
    {
      "epoch": 0.1974,
      "grad_norm": 0.3271949291229248,
      "learning_rate": 0.00011252040000000001,
      "loss": 0.2555,
      "step": 218700
    },
    {
      "epoch": 0.1976,
      "grad_norm": 0.18260078132152557,
      "learning_rate": 0.0001124804,
      "loss": 0.255,
      "step": 218800
    },
    {
      "epoch": 0.1978,
      "grad_norm": 0.1611817330121994,
      "learning_rate": 0.0001124404,
      "loss": 0.2601,
      "step": 218900
    },
    {
      "epoch": 0.198,
      "grad_norm": 0.22105377912521362,
      "learning_rate": 0.00011240040000000001,
      "loss": 0.2551,
      "step": 219000
    },
    {
      "epoch": 0.1982,
      "grad_norm": 0.22783438861370087,
      "learning_rate": 0.00011236040000000002,
      "loss": 0.2536,
      "step": 219100
    },
    {
      "epoch": 0.1984,
      "grad_norm": 0.14097750186920166,
      "learning_rate": 0.00011232040000000001,
      "loss": 0.251,
      "step": 219200
    },
    {
      "epoch": 0.1986,
      "grad_norm": 0.16217005252838135,
      "learning_rate": 0.0001122804,
      "loss": 0.2546,
      "step": 219300
    },
    {
      "epoch": 0.1988,
      "grad_norm": 0.28094884753227234,
      "learning_rate": 0.0001122404,
      "loss": 0.2597,
      "step": 219400
    },
    {
      "epoch": 0.199,
      "grad_norm": 0.1714087426662445,
      "learning_rate": 0.0001122004,
      "loss": 0.2532,
      "step": 219500
    },
    {
      "epoch": 0.1992,
      "grad_norm": 0.2341964989900589,
      "learning_rate": 0.00011216040000000001,
      "loss": 0.2566,
      "step": 219600
    },
    {
      "epoch": 0.1994,
      "grad_norm": 0.16600032150745392,
      "learning_rate": 0.00011212040000000002,
      "loss": 0.2502,
      "step": 219700
    },
    {
      "epoch": 0.1996,
      "grad_norm": 0.3427354693412781,
      "learning_rate": 0.0001120804,
      "loss": 0.2568,
      "step": 219800
    },
    {
      "epoch": 0.1998,
      "grad_norm": 0.2476433664560318,
      "learning_rate": 0.0001120404,
      "loss": 0.2579,
      "step": 219900
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.24420465528964996,
      "learning_rate": 0.0001120004,
      "loss": 0.2519,
      "step": 220000
    },
    {
      "epoch": 0.2002,
      "grad_norm": 0.4817318618297577,
      "learning_rate": 0.0001119604,
      "loss": 0.3021,
      "step": 220100
    },
    {
      "epoch": 0.2004,
      "grad_norm": 0.42361992597579956,
      "learning_rate": 0.00011192040000000001,
      "loss": 0.3307,
      "step": 220200
    },
    {
      "epoch": 0.2006,
      "grad_norm": 0.20908264815807343,
      "learning_rate": 0.0001118804,
      "loss": 0.316,
      "step": 220300
    },
    {
      "epoch": 0.2008,
      "grad_norm": 0.4234218895435333,
      "learning_rate": 0.0001118404,
      "loss": 0.3001,
      "step": 220400
    },
    {
      "epoch": 0.201,
      "grad_norm": 0.5726023316383362,
      "learning_rate": 0.00011180040000000001,
      "loss": 0.3016,
      "step": 220500
    },
    {
      "epoch": 0.2012,
      "grad_norm": 0.28417375683784485,
      "learning_rate": 0.00011176040000000002,
      "loss": 0.3389,
      "step": 220600
    },
    {
      "epoch": 0.2014,
      "grad_norm": 0.3290591537952423,
      "learning_rate": 0.00011172040000000001,
      "loss": 0.3019,
      "step": 220700
    },
    {
      "epoch": 0.2016,
      "grad_norm": 0.5273597836494446,
      "learning_rate": 0.0001116804,
      "loss": 0.3008,
      "step": 220800
    },
    {
      "epoch": 0.2018,
      "grad_norm": 0.19379092752933502,
      "learning_rate": 0.0001116404,
      "loss": 0.3282,
      "step": 220900
    },
    {
      "epoch": 0.202,
      "grad_norm": 0.19746637344360352,
      "learning_rate": 0.0001116004,
      "loss": 0.2928,
      "step": 221000
    },
    {
      "epoch": 0.2022,
      "grad_norm": 0.2760812044143677,
      "learning_rate": 0.00011156040000000001,
      "loss": 0.2961,
      "step": 221100
    },
    {
      "epoch": 0.2024,
      "grad_norm": 0.21008889377117157,
      "learning_rate": 0.00011152040000000002,
      "loss": 0.2986,
      "step": 221200
    },
    {
      "epoch": 0.2026,
      "grad_norm": 0.4908127784729004,
      "learning_rate": 0.0001114804,
      "loss": 0.3456,
      "step": 221300
    },
    {
      "epoch": 0.2028,
      "grad_norm": 0.1982090175151825,
      "learning_rate": 0.0001114404,
      "loss": 0.3186,
      "step": 221400
    },
    {
      "epoch": 0.203,
      "grad_norm": 0.37534865736961365,
      "learning_rate": 0.0001114004,
      "loss": 0.2995,
      "step": 221500
    },
    {
      "epoch": 0.2032,
      "grad_norm": 0.24973396956920624,
      "learning_rate": 0.0001113604,
      "loss": 0.2996,
      "step": 221600
    },
    {
      "epoch": 0.2034,
      "grad_norm": 0.21024543046951294,
      "learning_rate": 0.00011132040000000001,
      "loss": 0.2907,
      "step": 221700
    },
    {
      "epoch": 0.2036,
      "grad_norm": 0.3225703537464142,
      "learning_rate": 0.00011128039999999999,
      "loss": 0.2911,
      "step": 221800
    },
    {
      "epoch": 0.2038,
      "grad_norm": 0.4581950306892395,
      "learning_rate": 0.0001112404,
      "loss": 0.3637,
      "step": 221900
    },
    {
      "epoch": 0.204,
      "grad_norm": 0.5433754920959473,
      "learning_rate": 0.00011120040000000001,
      "loss": 0.3487,
      "step": 222000
    },
    {
      "epoch": 0.2042,
      "grad_norm": 0.22788307070732117,
      "learning_rate": 0.00011116040000000001,
      "loss": 0.3004,
      "step": 222100
    },
    {
      "epoch": 0.2044,
      "grad_norm": 0.16002947092056274,
      "learning_rate": 0.00011112040000000001,
      "loss": 0.2993,
      "step": 222200
    },
    {
      "epoch": 0.2046,
      "grad_norm": 0.28923583030700684,
      "learning_rate": 0.0001110804,
      "loss": 0.3036,
      "step": 222300
    },
    {
      "epoch": 0.2048,
      "grad_norm": 0.3555202782154083,
      "learning_rate": 0.0001110404,
      "loss": 0.338,
      "step": 222400
    },
    {
      "epoch": 0.205,
      "grad_norm": 0.1331017166376114,
      "learning_rate": 0.0001110004,
      "loss": 0.2904,
      "step": 222500
    },
    {
      "epoch": 0.2052,
      "grad_norm": 0.269748717546463,
      "learning_rate": 0.00011096040000000001,
      "loss": 0.3133,
      "step": 222600
    },
    {
      "epoch": 0.2054,
      "grad_norm": 0.919607937335968,
      "learning_rate": 0.00011092040000000002,
      "loss": 0.3133,
      "step": 222700
    },
    {
      "epoch": 0.2056,
      "grad_norm": 0.27156761288642883,
      "learning_rate": 0.0001108804,
      "loss": 0.3065,
      "step": 222800
    },
    {
      "epoch": 0.2058,
      "grad_norm": 0.48692193627357483,
      "learning_rate": 0.0001108404,
      "loss": 0.3025,
      "step": 222900
    },
    {
      "epoch": 0.206,
      "grad_norm": 0.8623552322387695,
      "learning_rate": 0.00011080040000000001,
      "loss": 0.296,
      "step": 223000
    },
    {
      "epoch": 0.2062,
      "grad_norm": 0.18269303441047668,
      "learning_rate": 0.0001107604,
      "loss": 0.277,
      "step": 223100
    },
    {
      "epoch": 0.2064,
      "grad_norm": 0.37067535519599915,
      "learning_rate": 0.00011072040000000001,
      "loss": 0.3043,
      "step": 223200
    },
    {
      "epoch": 0.2066,
      "grad_norm": 0.25582921504974365,
      "learning_rate": 0.00011068039999999999,
      "loss": 0.2949,
      "step": 223300
    },
    {
      "epoch": 0.2068,
      "grad_norm": 0.19148407876491547,
      "learning_rate": 0.0001106404,
      "loss": 0.2955,
      "step": 223400
    },
    {
      "epoch": 0.207,
      "grad_norm": 0.24919286370277405,
      "learning_rate": 0.0001106004,
      "loss": 0.3014,
      "step": 223500
    },
    {
      "epoch": 0.2072,
      "grad_norm": 0.24463938176631927,
      "learning_rate": 0.00011056040000000001,
      "loss": 0.3057,
      "step": 223600
    },
    {
      "epoch": 0.2074,
      "grad_norm": 0.35586199164390564,
      "learning_rate": 0.0001105204,
      "loss": 0.29,
      "step": 223700
    },
    {
      "epoch": 0.2076,
      "grad_norm": 0.319210946559906,
      "learning_rate": 0.0001104804,
      "loss": 0.368,
      "step": 223800
    },
    {
      "epoch": 0.2078,
      "grad_norm": 0.282400906085968,
      "learning_rate": 0.0001104404,
      "loss": 0.292,
      "step": 223900
    },
    {
      "epoch": 0.208,
      "grad_norm": 0.17433995008468628,
      "learning_rate": 0.0001104004,
      "loss": 0.2892,
      "step": 224000
    },
    {
      "epoch": 0.2082,
      "grad_norm": 0.5151282548904419,
      "learning_rate": 0.00011036040000000001,
      "loss": 0.2962,
      "step": 224100
    },
    {
      "epoch": 0.2084,
      "grad_norm": 0.2073827087879181,
      "learning_rate": 0.00011032040000000002,
      "loss": 0.2928,
      "step": 224200
    },
    {
      "epoch": 0.2086,
      "grad_norm": 0.40052372217178345,
      "learning_rate": 0.0001102804,
      "loss": 0.297,
      "step": 224300
    },
    {
      "epoch": 0.2088,
      "grad_norm": 0.26308512687683105,
      "learning_rate": 0.0001102404,
      "loss": 0.3246,
      "step": 224400
    },
    {
      "epoch": 0.209,
      "grad_norm": 0.35187768936157227,
      "learning_rate": 0.00011020040000000001,
      "loss": 0.3031,
      "step": 224500
    },
    {
      "epoch": 0.2092,
      "grad_norm": 0.1798812747001648,
      "learning_rate": 0.0001101604,
      "loss": 0.2947,
      "step": 224600
    },
    {
      "epoch": 0.2094,
      "grad_norm": 0.2565075159072876,
      "learning_rate": 0.00011012040000000001,
      "loss": 0.3018,
      "step": 224700
    },
    {
      "epoch": 0.2096,
      "grad_norm": 0.2061193436384201,
      "learning_rate": 0.00011008039999999999,
      "loss": 0.2986,
      "step": 224800
    },
    {
      "epoch": 0.2098,
      "grad_norm": 0.15753225982189178,
      "learning_rate": 0.0001100404,
      "loss": 0.2993,
      "step": 224900
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.18000277876853943,
      "learning_rate": 0.0001100004,
      "loss": 0.2809,
      "step": 225000
    },
    {
      "epoch": 0.2102,
      "grad_norm": 0.17245367169380188,
      "learning_rate": 0.00010996040000000001,
      "loss": 0.3003,
      "step": 225100
    },
    {
      "epoch": 0.2104,
      "grad_norm": 0.27620208263397217,
      "learning_rate": 0.0001099204,
      "loss": 0.2863,
      "step": 225200
    },
    {
      "epoch": 0.2106,
      "grad_norm": 0.1691516935825348,
      "learning_rate": 0.0001098804,
      "loss": 0.3085,
      "step": 225300
    },
    {
      "epoch": 0.2108,
      "grad_norm": 0.25270703434944153,
      "learning_rate": 0.00010984039999999999,
      "loss": 0.3018,
      "step": 225400
    },
    {
      "epoch": 0.211,
      "grad_norm": 0.26797574758529663,
      "learning_rate": 0.0001098004,
      "loss": 0.2919,
      "step": 225500
    },
    {
      "epoch": 0.2112,
      "grad_norm": 0.1746404618024826,
      "learning_rate": 0.00010976040000000001,
      "loss": 0.282,
      "step": 225600
    },
    {
      "epoch": 0.2114,
      "grad_norm": 0.2220151722431183,
      "learning_rate": 0.00010972040000000001,
      "loss": 0.3083,
      "step": 225700
    },
    {
      "epoch": 0.2116,
      "grad_norm": 0.3341788947582245,
      "learning_rate": 0.0001096804,
      "loss": 0.2846,
      "step": 225800
    },
    {
      "epoch": 0.2118,
      "grad_norm": 0.2519269585609436,
      "learning_rate": 0.0001096404,
      "loss": 0.3022,
      "step": 225900
    },
    {
      "epoch": 0.212,
      "grad_norm": 0.4205581545829773,
      "learning_rate": 0.00010960040000000001,
      "loss": 0.3163,
      "step": 226000
    },
    {
      "epoch": 0.2122,
      "grad_norm": 0.2838189899921417,
      "learning_rate": 0.0001095604,
      "loss": 0.2892,
      "step": 226100
    },
    {
      "epoch": 0.2124,
      "grad_norm": 0.25370433926582336,
      "learning_rate": 0.00010952040000000001,
      "loss": 0.282,
      "step": 226200
    },
    {
      "epoch": 0.2126,
      "grad_norm": 0.26368477940559387,
      "learning_rate": 0.00010948040000000002,
      "loss": 0.2784,
      "step": 226300
    },
    {
      "epoch": 0.2128,
      "grad_norm": 0.46951645612716675,
      "learning_rate": 0.0001094404,
      "loss": 0.3009,
      "step": 226400
    },
    {
      "epoch": 0.213,
      "grad_norm": 0.4232272207736969,
      "learning_rate": 0.0001094004,
      "loss": 0.2887,
      "step": 226500
    },
    {
      "epoch": 0.2132,
      "grad_norm": 0.19953791797161102,
      "learning_rate": 0.00010936040000000001,
      "loss": 0.2851,
      "step": 226600
    },
    {
      "epoch": 0.2134,
      "grad_norm": 0.28893914818763733,
      "learning_rate": 0.00010932040000000002,
      "loss": 0.2904,
      "step": 226700
    },
    {
      "epoch": 0.2136,
      "grad_norm": 0.18542243540287018,
      "learning_rate": 0.00010928040000000001,
      "loss": 0.2925,
      "step": 226800
    },
    {
      "epoch": 0.2138,
      "grad_norm": 0.2543767988681793,
      "learning_rate": 0.00010924039999999999,
      "loss": 0.2859,
      "step": 226900
    },
    {
      "epoch": 0.214,
      "grad_norm": 0.380341500043869,
      "learning_rate": 0.0001092004,
      "loss": 0.2948,
      "step": 227000
    },
    {
      "epoch": 0.2142,
      "grad_norm": 0.6200868487358093,
      "learning_rate": 0.0001091604,
      "loss": 0.3034,
      "step": 227100
    },
    {
      "epoch": 0.2144,
      "grad_norm": 0.40341097116470337,
      "learning_rate": 0.00010912040000000001,
      "loss": 0.2803,
      "step": 227200
    },
    {
      "epoch": 0.2146,
      "grad_norm": 0.27767351269721985,
      "learning_rate": 0.00010908040000000002,
      "loss": 0.2922,
      "step": 227300
    },
    {
      "epoch": 0.2148,
      "grad_norm": 0.18957166373729706,
      "learning_rate": 0.0001090404,
      "loss": 0.3052,
      "step": 227400
    },
    {
      "epoch": 0.215,
      "grad_norm": 0.21950003504753113,
      "learning_rate": 0.00010900040000000001,
      "loss": 0.3,
      "step": 227500
    },
    {
      "epoch": 0.2152,
      "grad_norm": 0.36292436718940735,
      "learning_rate": 0.0001089604,
      "loss": 0.291,
      "step": 227600
    },
    {
      "epoch": 0.2154,
      "grad_norm": 0.4383956789970398,
      "learning_rate": 0.00010892040000000001,
      "loss": 0.302,
      "step": 227700
    },
    {
      "epoch": 0.2156,
      "grad_norm": 0.18310613930225372,
      "learning_rate": 0.00010888040000000002,
      "loss": 0.3081,
      "step": 227800
    },
    {
      "epoch": 0.2158,
      "grad_norm": 0.21227945387363434,
      "learning_rate": 0.0001088404,
      "loss": 0.2791,
      "step": 227900
    },
    {
      "epoch": 0.216,
      "grad_norm": 0.25162750482559204,
      "learning_rate": 0.0001088004,
      "loss": 0.2922,
      "step": 228000
    },
    {
      "epoch": 0.2162,
      "grad_norm": 0.4016803205013275,
      "learning_rate": 0.00010876040000000001,
      "loss": 0.325,
      "step": 228100
    },
    {
      "epoch": 0.2164,
      "grad_norm": 0.24426665902137756,
      "learning_rate": 0.00010872040000000002,
      "loss": 0.2946,
      "step": 228200
    },
    {
      "epoch": 0.2166,
      "grad_norm": 0.35735762119293213,
      "learning_rate": 0.00010868040000000001,
      "loss": 0.2953,
      "step": 228300
    },
    {
      "epoch": 0.2168,
      "grad_norm": 0.17666518688201904,
      "learning_rate": 0.0001086404,
      "loss": 0.2933,
      "step": 228400
    },
    {
      "epoch": 0.217,
      "grad_norm": 0.4070577323436737,
      "learning_rate": 0.0001086004,
      "loss": 0.2955,
      "step": 228500
    },
    {
      "epoch": 0.2172,
      "grad_norm": 0.24427427351474762,
      "learning_rate": 0.0001085604,
      "loss": 0.2858,
      "step": 228600
    },
    {
      "epoch": 0.2174,
      "grad_norm": 0.21364831924438477,
      "learning_rate": 0.00010852040000000001,
      "loss": 0.2919,
      "step": 228700
    },
    {
      "epoch": 0.2176,
      "grad_norm": 0.20407788455486298,
      "learning_rate": 0.00010848040000000002,
      "loss": 0.2898,
      "step": 228800
    },
    {
      "epoch": 0.2178,
      "grad_norm": 0.31540021300315857,
      "learning_rate": 0.0001084404,
      "loss": 0.2987,
      "step": 228900
    },
    {
      "epoch": 0.218,
      "grad_norm": 0.15270252525806427,
      "learning_rate": 0.0001084004,
      "loss": 0.2872,
      "step": 229000
    },
    {
      "epoch": 0.2182,
      "grad_norm": 0.23393623530864716,
      "learning_rate": 0.0001083604,
      "loss": 0.3058,
      "step": 229100
    },
    {
      "epoch": 0.2184,
      "grad_norm": 0.5334755778312683,
      "learning_rate": 0.00010832040000000001,
      "loss": 0.29,
      "step": 229200
    },
    {
      "epoch": 0.2186,
      "grad_norm": 0.1923484355211258,
      "learning_rate": 0.00010828040000000001,
      "loss": 0.2873,
      "step": 229300
    },
    {
      "epoch": 0.2188,
      "grad_norm": 0.23377379775047302,
      "learning_rate": 0.0001082404,
      "loss": 0.2823,
      "step": 229400
    },
    {
      "epoch": 0.219,
      "grad_norm": 0.22919674217700958,
      "learning_rate": 0.0001082004,
      "loss": 0.2836,
      "step": 229500
    },
    {
      "epoch": 0.2192,
      "grad_norm": 0.20516985654830933,
      "learning_rate": 0.00010816040000000001,
      "loss": 0.29,
      "step": 229600
    },
    {
      "epoch": 0.2194,
      "grad_norm": 0.19901040196418762,
      "learning_rate": 0.00010812040000000002,
      "loss": 0.2963,
      "step": 229700
    },
    {
      "epoch": 0.2196,
      "grad_norm": 0.24688464403152466,
      "learning_rate": 0.00010808040000000001,
      "loss": 0.2855,
      "step": 229800
    },
    {
      "epoch": 0.2198,
      "grad_norm": 0.23144333064556122,
      "learning_rate": 0.0001080404,
      "loss": 0.2964,
      "step": 229900
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.2202957421541214,
      "learning_rate": 0.0001080004,
      "loss": 0.2925,
      "step": 230000
    },
    {
      "epoch": 0.2202,
      "grad_norm": 0.23589707911014557,
      "learning_rate": 0.0001079604,
      "loss": 0.2759,
      "step": 230100
    },
    {
      "epoch": 0.2204,
      "grad_norm": 0.31147778034210205,
      "learning_rate": 0.00010792040000000001,
      "loss": 0.2668,
      "step": 230200
    },
    {
      "epoch": 0.2206,
      "grad_norm": 0.36505284905433655,
      "learning_rate": 0.00010788040000000002,
      "loss": 0.2807,
      "step": 230300
    },
    {
      "epoch": 0.2208,
      "grad_norm": 0.24104100465774536,
      "learning_rate": 0.0001078404,
      "loss": 0.2802,
      "step": 230400
    },
    {
      "epoch": 0.221,
      "grad_norm": 0.2541787922382355,
      "learning_rate": 0.0001078004,
      "loss": 0.2738,
      "step": 230500
    },
    {
      "epoch": 0.2212,
      "grad_norm": 0.18488843739032745,
      "learning_rate": 0.0001077604,
      "loss": 0.2781,
      "step": 230600
    },
    {
      "epoch": 0.2214,
      "grad_norm": 0.30733954906463623,
      "learning_rate": 0.0001077204,
      "loss": 0.2689,
      "step": 230700
    },
    {
      "epoch": 0.2216,
      "grad_norm": 0.4184741675853729,
      "learning_rate": 0.00010768040000000001,
      "loss": 0.2821,
      "step": 230800
    },
    {
      "epoch": 0.2218,
      "grad_norm": 0.234775573015213,
      "learning_rate": 0.0001076404,
      "loss": 0.2788,
      "step": 230900
    },
    {
      "epoch": 0.222,
      "grad_norm": 0.1866404116153717,
      "learning_rate": 0.0001076004,
      "loss": 0.2805,
      "step": 231000
    },
    {
      "epoch": 0.2222,
      "grad_norm": 0.20000393688678741,
      "learning_rate": 0.00010756040000000001,
      "loss": 0.2786,
      "step": 231100
    },
    {
      "epoch": 0.2224,
      "grad_norm": 0.2550050914287567,
      "learning_rate": 0.00010752040000000001,
      "loss": 0.2808,
      "step": 231200
    },
    {
      "epoch": 0.2226,
      "grad_norm": 0.19352710247039795,
      "learning_rate": 0.00010748040000000001,
      "loss": 0.2794,
      "step": 231300
    },
    {
      "epoch": 0.2228,
      "grad_norm": 0.33794355392456055,
      "learning_rate": 0.0001074404,
      "loss": 0.2654,
      "step": 231400
    },
    {
      "epoch": 0.223,
      "grad_norm": 0.1755092740058899,
      "learning_rate": 0.0001074004,
      "loss": 0.2777,
      "step": 231500
    },
    {
      "epoch": 0.2232,
      "grad_norm": 0.2609795331954956,
      "learning_rate": 0.0001073604,
      "loss": 0.2749,
      "step": 231600
    },
    {
      "epoch": 0.2234,
      "grad_norm": 0.23666484653949738,
      "learning_rate": 0.00010732040000000001,
      "loss": 0.2837,
      "step": 231700
    },
    {
      "epoch": 0.2236,
      "grad_norm": 0.28469526767730713,
      "learning_rate": 0.00010728040000000002,
      "loss": 0.283,
      "step": 231800
    },
    {
      "epoch": 0.2238,
      "grad_norm": 0.21141879260540009,
      "learning_rate": 0.0001072404,
      "loss": 0.2922,
      "step": 231900
    },
    {
      "epoch": 0.224,
      "grad_norm": 0.19850678741931915,
      "learning_rate": 0.0001072004,
      "loss": 0.2771,
      "step": 232000
    },
    {
      "epoch": 0.2242,
      "grad_norm": 0.24743816256523132,
      "learning_rate": 0.00010716040000000001,
      "loss": 0.2749,
      "step": 232100
    },
    {
      "epoch": 0.2244,
      "grad_norm": 0.3401179015636444,
      "learning_rate": 0.0001071204,
      "loss": 0.2781,
      "step": 232200
    },
    {
      "epoch": 0.2246,
      "grad_norm": 0.2130950391292572,
      "learning_rate": 0.00010708040000000001,
      "loss": 0.27,
      "step": 232300
    },
    {
      "epoch": 0.2248,
      "grad_norm": 0.37853744626045227,
      "learning_rate": 0.00010704039999999999,
      "loss": 0.2838,
      "step": 232400
    },
    {
      "epoch": 0.225,
      "grad_norm": 0.20562714338302612,
      "learning_rate": 0.0001070004,
      "loss": 0.2653,
      "step": 232500
    },
    {
      "epoch": 0.2252,
      "grad_norm": 0.17086434364318848,
      "learning_rate": 0.00010696040000000001,
      "loss": 0.2765,
      "step": 232600
    },
    {
      "epoch": 0.2254,
      "grad_norm": 0.3876626193523407,
      "learning_rate": 0.00010692040000000001,
      "loss": 0.2758,
      "step": 232700
    },
    {
      "epoch": 0.2256,
      "grad_norm": 0.29403626918792725,
      "learning_rate": 0.00010688040000000001,
      "loss": 0.2783,
      "step": 232800
    },
    {
      "epoch": 0.2258,
      "grad_norm": 0.3147401809692383,
      "learning_rate": 0.0001068404,
      "loss": 0.2843,
      "step": 232900
    },
    {
      "epoch": 0.226,
      "grad_norm": 0.38840705156326294,
      "learning_rate": 0.0001068004,
      "loss": 0.3095,
      "step": 233000
    },
    {
      "epoch": 0.2262,
      "grad_norm": 0.2079588621854782,
      "learning_rate": 0.0001067604,
      "loss": 0.2971,
      "step": 233100
    },
    {
      "epoch": 0.2264,
      "grad_norm": 0.293052077293396,
      "learning_rate": 0.00010672040000000001,
      "loss": 0.2787,
      "step": 233200
    },
    {
      "epoch": 0.2266,
      "grad_norm": 0.1692897379398346,
      "learning_rate": 0.00010668040000000002,
      "loss": 0.2939,
      "step": 233300
    },
    {
      "epoch": 0.2268,
      "grad_norm": 0.19356147944927216,
      "learning_rate": 0.0001066404,
      "loss": 0.2901,
      "step": 233400
    },
    {
      "epoch": 0.227,
      "grad_norm": 0.2273617833852768,
      "learning_rate": 0.0001066004,
      "loss": 0.2794,
      "step": 233500
    },
    {
      "epoch": 0.2272,
      "grad_norm": 0.21963585913181305,
      "learning_rate": 0.00010656040000000001,
      "loss": 0.2882,
      "step": 233600
    },
    {
      "epoch": 0.2274,
      "grad_norm": 0.3202444911003113,
      "learning_rate": 0.0001065204,
      "loss": 0.2735,
      "step": 233700
    },
    {
      "epoch": 0.2276,
      "grad_norm": 0.2695014774799347,
      "learning_rate": 0.00010648040000000001,
      "loss": 0.288,
      "step": 233800
    },
    {
      "epoch": 0.2278,
      "grad_norm": 0.1968889981508255,
      "learning_rate": 0.00010644039999999999,
      "loss": 0.2818,
      "step": 233900
    },
    {
      "epoch": 0.228,
      "grad_norm": 0.2076377421617508,
      "learning_rate": 0.0001064004,
      "loss": 0.2806,
      "step": 234000
    },
    {
      "epoch": 0.2282,
      "grad_norm": 0.17942437529563904,
      "learning_rate": 0.0001063604,
      "loss": 0.2874,
      "step": 234100
    },
    {
      "epoch": 0.2284,
      "grad_norm": 0.20164412260055542,
      "learning_rate": 0.00010632040000000001,
      "loss": 0.271,
      "step": 234200
    },
    {
      "epoch": 0.2286,
      "grad_norm": 0.3018455505371094,
      "learning_rate": 0.0001062804,
      "loss": 0.2787,
      "step": 234300
    },
    {
      "epoch": 0.2288,
      "grad_norm": 0.18581052124500275,
      "learning_rate": 0.0001062404,
      "loss": 0.271,
      "step": 234400
    },
    {
      "epoch": 0.229,
      "grad_norm": 0.36110278964042664,
      "learning_rate": 0.0001062004,
      "loss": 0.2856,
      "step": 234500
    },
    {
      "epoch": 0.2292,
      "grad_norm": 0.21756871044635773,
      "learning_rate": 0.0001061604,
      "loss": 0.2872,
      "step": 234600
    },
    {
      "epoch": 0.2294,
      "grad_norm": 0.1998421549797058,
      "learning_rate": 0.00010612040000000001,
      "loss": 0.28,
      "step": 234700
    },
    {
      "epoch": 0.2296,
      "grad_norm": 0.17958033084869385,
      "learning_rate": 0.00010608040000000002,
      "loss": 0.2811,
      "step": 234800
    },
    {
      "epoch": 0.2298,
      "grad_norm": 0.6013987064361572,
      "learning_rate": 0.0001060404,
      "loss": 0.2732,
      "step": 234900
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.20178425312042236,
      "learning_rate": 0.0001060004,
      "loss": 0.2713,
      "step": 235000
    },
    {
      "epoch": 0.2302,
      "grad_norm": 0.2956780791282654,
      "learning_rate": 0.00010596040000000001,
      "loss": 0.2775,
      "step": 235100
    },
    {
      "epoch": 0.2304,
      "grad_norm": 0.2311795949935913,
      "learning_rate": 0.0001059204,
      "loss": 0.2807,
      "step": 235200
    },
    {
      "epoch": 0.2306,
      "grad_norm": 0.24748362600803375,
      "learning_rate": 0.00010588040000000001,
      "loss": 0.2657,
      "step": 235300
    },
    {
      "epoch": 0.2308,
      "grad_norm": 0.38484346866607666,
      "learning_rate": 0.00010584039999999999,
      "loss": 0.2714,
      "step": 235400
    },
    {
      "epoch": 0.231,
      "grad_norm": 0.20243670046329498,
      "learning_rate": 0.0001058004,
      "loss": 0.2783,
      "step": 235500
    },
    {
      "epoch": 0.2312,
      "grad_norm": 0.5278353095054626,
      "learning_rate": 0.0001057604,
      "loss": 0.2779,
      "step": 235600
    },
    {
      "epoch": 0.2314,
      "grad_norm": 0.27309587597846985,
      "learning_rate": 0.00010572040000000001,
      "loss": 0.2695,
      "step": 235700
    },
    {
      "epoch": 0.2316,
      "grad_norm": 0.16521087288856506,
      "learning_rate": 0.00010568040000000002,
      "loss": 0.2732,
      "step": 235800
    },
    {
      "epoch": 0.2318,
      "grad_norm": 0.5062386393547058,
      "learning_rate": 0.0001056404,
      "loss": 0.2878,
      "step": 235900
    },
    {
      "epoch": 0.232,
      "grad_norm": 0.19394774734973907,
      "learning_rate": 0.00010560039999999999,
      "loss": 0.2851,
      "step": 236000
    },
    {
      "epoch": 0.2322,
      "grad_norm": 0.171609029173851,
      "learning_rate": 0.0001055604,
      "loss": 0.2809,
      "step": 236100
    },
    {
      "epoch": 0.2324,
      "grad_norm": 0.30568891763687134,
      "learning_rate": 0.00010552040000000001,
      "loss": 0.2952,
      "step": 236200
    },
    {
      "epoch": 0.2326,
      "grad_norm": 0.32132649421691895,
      "learning_rate": 0.00010548040000000001,
      "loss": 0.2751,
      "step": 236300
    },
    {
      "epoch": 0.2328,
      "grad_norm": 0.30104103684425354,
      "learning_rate": 0.0001054404,
      "loss": 0.285,
      "step": 236400
    },
    {
      "epoch": 0.233,
      "grad_norm": 0.19023063778877258,
      "learning_rate": 0.0001054004,
      "loss": 0.3591,
      "step": 236500
    },
    {
      "epoch": 0.2332,
      "grad_norm": 0.1921398788690567,
      "learning_rate": 0.00010536040000000001,
      "loss": 0.2813,
      "step": 236600
    },
    {
      "epoch": 0.2334,
      "grad_norm": 0.15040914714336395,
      "learning_rate": 0.0001053204,
      "loss": 0.2873,
      "step": 236700
    },
    {
      "epoch": 0.2336,
      "grad_norm": 0.18588508665561676,
      "learning_rate": 0.00010528040000000001,
      "loss": 0.2722,
      "step": 236800
    },
    {
      "epoch": 0.2338,
      "grad_norm": 0.3410860002040863,
      "learning_rate": 0.00010524039999999999,
      "loss": 0.2811,
      "step": 236900
    },
    {
      "epoch": 0.234,
      "grad_norm": 0.21273022890090942,
      "learning_rate": 0.0001052004,
      "loss": 0.2892,
      "step": 237000
    },
    {
      "epoch": 0.2342,
      "grad_norm": 0.20495010912418365,
      "learning_rate": 0.0001051604,
      "loss": 0.2826,
      "step": 237100
    },
    {
      "epoch": 0.2344,
      "grad_norm": 0.3774225413799286,
      "learning_rate": 0.00010512040000000001,
      "loss": 0.2704,
      "step": 237200
    },
    {
      "epoch": 0.2346,
      "grad_norm": 0.1908106654882431,
      "learning_rate": 0.00010508040000000002,
      "loss": 0.2854,
      "step": 237300
    },
    {
      "epoch": 0.2348,
      "grad_norm": 0.333554208278656,
      "learning_rate": 0.0001050404,
      "loss": 0.276,
      "step": 237400
    },
    {
      "epoch": 0.235,
      "grad_norm": 0.2886081337928772,
      "learning_rate": 0.0001050004,
      "loss": 0.2737,
      "step": 237500
    },
    {
      "epoch": 0.2352,
      "grad_norm": 0.2619887888431549,
      "learning_rate": 0.0001049604,
      "loss": 0.2718,
      "step": 237600
    },
    {
      "epoch": 0.2354,
      "grad_norm": 0.2604043483734131,
      "learning_rate": 0.0001049204,
      "loss": 0.2757,
      "step": 237700
    },
    {
      "epoch": 0.2356,
      "grad_norm": 0.20619331300258636,
      "learning_rate": 0.00010488040000000001,
      "loss": 0.2722,
      "step": 237800
    },
    {
      "epoch": 0.2358,
      "grad_norm": 0.1808670163154602,
      "learning_rate": 0.00010484039999999999,
      "loss": 0.2715,
      "step": 237900
    },
    {
      "epoch": 0.236,
      "grad_norm": 0.39049023389816284,
      "learning_rate": 0.0001048004,
      "loss": 0.271,
      "step": 238000
    },
    {
      "epoch": 0.2362,
      "grad_norm": 0.2106139063835144,
      "learning_rate": 0.00010476040000000001,
      "loss": 0.2775,
      "step": 238100
    },
    {
      "epoch": 0.2364,
      "grad_norm": 0.36072278022766113,
      "learning_rate": 0.0001047204,
      "loss": 0.2719,
      "step": 238200
    },
    {
      "epoch": 0.2366,
      "grad_norm": 0.1794702261686325,
      "learning_rate": 0.00010468040000000001,
      "loss": 0.2865,
      "step": 238300
    },
    {
      "epoch": 0.2368,
      "grad_norm": 0.3740937113761902,
      "learning_rate": 0.00010464039999999999,
      "loss": 0.2725,
      "step": 238400
    },
    {
      "epoch": 0.237,
      "grad_norm": 0.2138763815164566,
      "learning_rate": 0.0001046004,
      "loss": 0.2776,
      "step": 238500
    },
    {
      "epoch": 0.2372,
      "grad_norm": 0.21753883361816406,
      "learning_rate": 0.0001045604,
      "loss": 0.2854,
      "step": 238600
    },
    {
      "epoch": 0.2374,
      "grad_norm": 0.19092372059822083,
      "learning_rate": 0.00010452040000000001,
      "loss": 0.2785,
      "step": 238700
    },
    {
      "epoch": 0.2376,
      "grad_norm": 0.18898741900920868,
      "learning_rate": 0.00010448040000000002,
      "loss": 0.2747,
      "step": 238800
    },
    {
      "epoch": 0.2378,
      "grad_norm": 0.3324822187423706,
      "learning_rate": 0.00010444040000000001,
      "loss": 0.2685,
      "step": 238900
    },
    {
      "epoch": 0.238,
      "grad_norm": 0.24389317631721497,
      "learning_rate": 0.0001044004,
      "loss": 0.2748,
      "step": 239000
    },
    {
      "epoch": 0.2382,
      "grad_norm": 0.23654966056346893,
      "learning_rate": 0.0001043604,
      "loss": 0.2845,
      "step": 239100
    },
    {
      "epoch": 0.2384,
      "grad_norm": 0.3116486072540283,
      "learning_rate": 0.0001043204,
      "loss": 0.2732,
      "step": 239200
    },
    {
      "epoch": 0.2386,
      "grad_norm": 0.33123424649238586,
      "learning_rate": 0.00010428040000000001,
      "loss": 0.2974,
      "step": 239300
    },
    {
      "epoch": 0.2388,
      "grad_norm": 0.2512767016887665,
      "learning_rate": 0.00010424040000000002,
      "loss": 0.2799,
      "step": 239400
    },
    {
      "epoch": 0.239,
      "grad_norm": 0.1622258871793747,
      "learning_rate": 0.0001042004,
      "loss": 0.2889,
      "step": 239500
    },
    {
      "epoch": 0.2392,
      "grad_norm": 0.33309224247932434,
      "learning_rate": 0.0001041604,
      "loss": 0.2861,
      "step": 239600
    },
    {
      "epoch": 0.2394,
      "grad_norm": 0.657027542591095,
      "learning_rate": 0.0001041204,
      "loss": 0.2772,
      "step": 239700
    },
    {
      "epoch": 0.2396,
      "grad_norm": 0.33285826444625854,
      "learning_rate": 0.00010408040000000001,
      "loss": 0.2832,
      "step": 239800
    },
    {
      "epoch": 0.2398,
      "grad_norm": 0.3333481550216675,
      "learning_rate": 0.00010404040000000001,
      "loss": 0.2798,
      "step": 239900
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.17029236257076263,
      "learning_rate": 0.0001040004,
      "loss": 0.2781,
      "step": 240000
    },
    {
      "epoch": 0.2402,
      "grad_norm": 0.2667362689971924,
      "learning_rate": 0.0001039604,
      "loss": 0.2745,
      "step": 240100
    },
    {
      "epoch": 0.2404,
      "grad_norm": 0.2005651295185089,
      "learning_rate": 0.00010392040000000001,
      "loss": 0.2708,
      "step": 240200
    },
    {
      "epoch": 0.2406,
      "grad_norm": 0.22746022045612335,
      "learning_rate": 0.00010388040000000002,
      "loss": 0.2826,
      "step": 240300
    },
    {
      "epoch": 0.2408,
      "grad_norm": 0.4452439546585083,
      "learning_rate": 0.00010384040000000001,
      "loss": 0.2786,
      "step": 240400
    },
    {
      "epoch": 0.241,
      "grad_norm": 0.2846798002719879,
      "learning_rate": 0.0001038004,
      "loss": 0.2759,
      "step": 240500
    },
    {
      "epoch": 0.2412,
      "grad_norm": 0.5280228853225708,
      "learning_rate": 0.0001037604,
      "loss": 0.2831,
      "step": 240600
    },
    {
      "epoch": 0.2414,
      "grad_norm": 0.18407422304153442,
      "learning_rate": 0.0001037204,
      "loss": 0.2861,
      "step": 240700
    },
    {
      "epoch": 0.2416,
      "grad_norm": 0.27900245785713196,
      "learning_rate": 0.00010368040000000001,
      "loss": 0.2761,
      "step": 240800
    },
    {
      "epoch": 0.2418,
      "grad_norm": 0.4371165931224823,
      "learning_rate": 0.00010364040000000002,
      "loss": 0.2741,
      "step": 240900
    },
    {
      "epoch": 0.242,
      "grad_norm": 0.1729346215724945,
      "learning_rate": 0.0001036004,
      "loss": 0.2749,
      "step": 241000
    },
    {
      "epoch": 0.2422,
      "grad_norm": 0.3014446794986725,
      "learning_rate": 0.0001035604,
      "loss": 0.2757,
      "step": 241100
    },
    {
      "epoch": 0.2424,
      "grad_norm": 0.2890118956565857,
      "learning_rate": 0.00010352040000000001,
      "loss": 0.2748,
      "step": 241200
    },
    {
      "epoch": 0.2426,
      "grad_norm": 0.18363769352436066,
      "learning_rate": 0.0001034804,
      "loss": 0.2786,
      "step": 241300
    },
    {
      "epoch": 0.2428,
      "grad_norm": 0.23119623959064484,
      "learning_rate": 0.00010344040000000001,
      "loss": 0.2763,
      "step": 241400
    },
    {
      "epoch": 0.243,
      "grad_norm": 0.23859907686710358,
      "learning_rate": 0.0001034004,
      "loss": 0.2766,
      "step": 241500
    },
    {
      "epoch": 0.2432,
      "grad_norm": 0.2567295432090759,
      "learning_rate": 0.0001033604,
      "loss": 0.2733,
      "step": 241600
    },
    {
      "epoch": 0.2434,
      "grad_norm": 0.3161749243736267,
      "learning_rate": 0.00010332040000000001,
      "loss": 0.2697,
      "step": 241700
    },
    {
      "epoch": 0.2436,
      "grad_norm": 0.31291186809539795,
      "learning_rate": 0.00010328040000000001,
      "loss": 0.2612,
      "step": 241800
    },
    {
      "epoch": 0.2438,
      "grad_norm": 0.23509889841079712,
      "learning_rate": 0.00010324040000000001,
      "loss": 0.2825,
      "step": 241900
    },
    {
      "epoch": 0.244,
      "grad_norm": 0.21681292355060577,
      "learning_rate": 0.0001032004,
      "loss": 0.2797,
      "step": 242000
    },
    {
      "epoch": 0.2442,
      "grad_norm": 0.2644917070865631,
      "learning_rate": 0.0001031604,
      "loss": 0.271,
      "step": 242100
    },
    {
      "epoch": 0.2444,
      "grad_norm": 0.536569356918335,
      "learning_rate": 0.0001031204,
      "loss": 0.2731,
      "step": 242200
    },
    {
      "epoch": 0.2446,
      "grad_norm": 0.44594717025756836,
      "learning_rate": 0.00010308040000000001,
      "loss": 0.2774,
      "step": 242300
    },
    {
      "epoch": 0.2448,
      "grad_norm": 0.309821754693985,
      "learning_rate": 0.00010304040000000002,
      "loss": 0.2899,
      "step": 242400
    },
    {
      "epoch": 0.245,
      "grad_norm": 0.28232014179229736,
      "learning_rate": 0.0001030004,
      "loss": 0.2944,
      "step": 242500
    },
    {
      "epoch": 0.2452,
      "grad_norm": 0.29917338490486145,
      "learning_rate": 0.0001029604,
      "loss": 0.2795,
      "step": 242600
    },
    {
      "epoch": 0.2454,
      "grad_norm": 0.23704950511455536,
      "learning_rate": 0.00010292040000000001,
      "loss": 0.2796,
      "step": 242700
    },
    {
      "epoch": 0.2456,
      "grad_norm": 0.21638484299182892,
      "learning_rate": 0.0001028804,
      "loss": 0.2636,
      "step": 242800
    },
    {
      "epoch": 0.2458,
      "grad_norm": 0.18547122180461884,
      "learning_rate": 0.00010284040000000001,
      "loss": 0.2669,
      "step": 242900
    },
    {
      "epoch": 0.246,
      "grad_norm": 0.17270751297473907,
      "learning_rate": 0.00010280039999999999,
      "loss": 0.2593,
      "step": 243000
    },
    {
      "epoch": 0.2462,
      "grad_norm": 0.16788358986377716,
      "learning_rate": 0.0001027604,
      "loss": 0.2592,
      "step": 243100
    },
    {
      "epoch": 0.2464,
      "grad_norm": 0.1689780205488205,
      "learning_rate": 0.0001027204,
      "loss": 0.2647,
      "step": 243200
    },
    {
      "epoch": 0.2466,
      "grad_norm": 0.28541648387908936,
      "learning_rate": 0.00010268040000000001,
      "loss": 0.2598,
      "step": 243300
    },
    {
      "epoch": 0.2468,
      "grad_norm": 0.24988971650600433,
      "learning_rate": 0.00010264040000000002,
      "loss": 0.2606,
      "step": 243400
    },
    {
      "epoch": 0.247,
      "grad_norm": 0.17981068789958954,
      "learning_rate": 0.0001026004,
      "loss": 0.2581,
      "step": 243500
    },
    {
      "epoch": 0.2472,
      "grad_norm": 0.19057908654212952,
      "learning_rate": 0.0001025604,
      "loss": 0.2645,
      "step": 243600
    },
    {
      "epoch": 0.2474,
      "grad_norm": 0.1446358561515808,
      "learning_rate": 0.0001025204,
      "loss": 0.2513,
      "step": 243700
    },
    {
      "epoch": 0.2476,
      "grad_norm": 0.18903489410877228,
      "learning_rate": 0.00010248040000000001,
      "loss": 0.2577,
      "step": 243800
    },
    {
      "epoch": 0.2478,
      "grad_norm": 0.18427778780460358,
      "learning_rate": 0.00010244040000000002,
      "loss": 0.2549,
      "step": 243900
    },
    {
      "epoch": 0.248,
      "grad_norm": 0.17634162306785583,
      "learning_rate": 0.0001024004,
      "loss": 0.2544,
      "step": 244000
    },
    {
      "epoch": 0.2482,
      "grad_norm": 0.15997561812400818,
      "learning_rate": 0.0001023604,
      "loss": 0.257,
      "step": 244100
    },
    {
      "epoch": 0.2484,
      "grad_norm": 0.13847199082374573,
      "learning_rate": 0.00010232040000000001,
      "loss": 0.2518,
      "step": 244200
    },
    {
      "epoch": 0.2486,
      "grad_norm": 0.19378510117530823,
      "learning_rate": 0.0001022804,
      "loss": 0.2608,
      "step": 244300
    },
    {
      "epoch": 0.2488,
      "grad_norm": 0.2129637449979782,
      "learning_rate": 0.00010224040000000001,
      "loss": 0.2586,
      "step": 244400
    },
    {
      "epoch": 0.249,
      "grad_norm": 0.28681081533432007,
      "learning_rate": 0.00010220039999999999,
      "loss": 0.259,
      "step": 244500
    },
    {
      "epoch": 0.2492,
      "grad_norm": 0.20721031725406647,
      "learning_rate": 0.0001021604,
      "loss": 0.2585,
      "step": 244600
    },
    {
      "epoch": 0.2494,
      "grad_norm": 0.2532545328140259,
      "learning_rate": 0.0001021204,
      "loss": 0.2532,
      "step": 244700
    },
    {
      "epoch": 0.2496,
      "grad_norm": 0.28852516412734985,
      "learning_rate": 0.00010208040000000001,
      "loss": 0.255,
      "step": 244800
    },
    {
      "epoch": 0.2498,
      "grad_norm": 0.1656378209590912,
      "learning_rate": 0.00010204040000000002,
      "loss": 0.2537,
      "step": 244900
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.2153751105070114,
      "learning_rate": 0.0001020004,
      "loss": 0.259,
      "step": 245000
    },
    {
      "epoch": 0.2502,
      "grad_norm": 0.19507664442062378,
      "learning_rate": 0.0001019604,
      "loss": 0.2587,
      "step": 245100
    },
    {
      "epoch": 0.2504,
      "grad_norm": 0.1736658811569214,
      "learning_rate": 0.0001019204,
      "loss": 0.2612,
      "step": 245200
    },
    {
      "epoch": 0.2506,
      "grad_norm": 0.18013814091682434,
      "learning_rate": 0.00010188040000000001,
      "loss": 0.2554,
      "step": 245300
    },
    {
      "epoch": 0.2508,
      "grad_norm": 0.2120230495929718,
      "learning_rate": 0.00010184040000000002,
      "loss": 0.2519,
      "step": 245400
    },
    {
      "epoch": 0.251,
      "grad_norm": 0.16198742389678955,
      "learning_rate": 0.0001018004,
      "loss": 0.254,
      "step": 245500
    },
    {
      "epoch": 0.2512,
      "grad_norm": 0.16579264402389526,
      "learning_rate": 0.0001017604,
      "loss": 0.258,
      "step": 245600
    },
    {
      "epoch": 0.2514,
      "grad_norm": 0.23009906709194183,
      "learning_rate": 0.00010172040000000001,
      "loss": 0.254,
      "step": 245700
    },
    {
      "epoch": 0.2516,
      "grad_norm": 0.15772415697574615,
      "learning_rate": 0.0001016804,
      "loss": 0.2532,
      "step": 245800
    },
    {
      "epoch": 0.2518,
      "grad_norm": 0.20444506406784058,
      "learning_rate": 0.00010164040000000001,
      "loss": 0.2524,
      "step": 245900
    },
    {
      "epoch": 0.252,
      "grad_norm": 0.19540543854236603,
      "learning_rate": 0.00010160039999999999,
      "loss": 0.2538,
      "step": 246000
    },
    {
      "epoch": 0.2522,
      "grad_norm": 0.17972715198993683,
      "learning_rate": 0.0001015604,
      "loss": 0.259,
      "step": 246100
    },
    {
      "epoch": 0.2524,
      "grad_norm": 0.17533235251903534,
      "learning_rate": 0.0001015204,
      "loss": 0.2564,
      "step": 246200
    },
    {
      "epoch": 0.2526,
      "grad_norm": 0.1778615266084671,
      "learning_rate": 0.00010148040000000001,
      "loss": 0.2579,
      "step": 246300
    },
    {
      "epoch": 0.2528,
      "grad_norm": 0.27196893095970154,
      "learning_rate": 0.00010144040000000002,
      "loss": 0.2631,
      "step": 246400
    },
    {
      "epoch": 0.253,
      "grad_norm": 0.20036780834197998,
      "learning_rate": 0.0001014004,
      "loss": 0.2619,
      "step": 246500
    },
    {
      "epoch": 0.2532,
      "grad_norm": 0.18973121047019958,
      "learning_rate": 0.0001013604,
      "loss": 0.2637,
      "step": 246600
    },
    {
      "epoch": 0.2534,
      "grad_norm": 0.17328648269176483,
      "learning_rate": 0.0001013204,
      "loss": 0.2516,
      "step": 246700
    },
    {
      "epoch": 0.2536,
      "grad_norm": 0.1752815842628479,
      "learning_rate": 0.00010128040000000001,
      "loss": 0.2572,
      "step": 246800
    },
    {
      "epoch": 0.2538,
      "grad_norm": 0.24637199938297272,
      "learning_rate": 0.00010124040000000001,
      "loss": 0.2543,
      "step": 246900
    },
    {
      "epoch": 0.254,
      "grad_norm": 0.19035212695598602,
      "learning_rate": 0.0001012004,
      "loss": 0.2513,
      "step": 247000
    },
    {
      "epoch": 0.2542,
      "grad_norm": 0.14904499053955078,
      "learning_rate": 0.0001011604,
      "loss": 0.2579,
      "step": 247100
    },
    {
      "epoch": 0.2544,
      "grad_norm": 0.6997934579849243,
      "learning_rate": 0.00010112040000000001,
      "loss": 0.2545,
      "step": 247200
    },
    {
      "epoch": 0.2546,
      "grad_norm": 0.1693519502878189,
      "learning_rate": 0.0001010804,
      "loss": 0.2575,
      "step": 247300
    },
    {
      "epoch": 0.2548,
      "grad_norm": 0.1666778028011322,
      "learning_rate": 0.00010104040000000001,
      "loss": 0.2645,
      "step": 247400
    },
    {
      "epoch": 0.255,
      "grad_norm": 0.24935069680213928,
      "learning_rate": 0.00010100039999999999,
      "loss": 0.2571,
      "step": 247500
    },
    {
      "epoch": 0.2552,
      "grad_norm": 0.1595294177532196,
      "learning_rate": 0.0001009604,
      "loss": 0.2556,
      "step": 247600
    },
    {
      "epoch": 0.2554,
      "grad_norm": 0.1594139188528061,
      "learning_rate": 0.0001009204,
      "loss": 0.2612,
      "step": 247700
    },
    {
      "epoch": 0.2556,
      "grad_norm": 0.2499750703573227,
      "learning_rate": 0.00010088040000000001,
      "loss": 0.2679,
      "step": 247800
    },
    {
      "epoch": 0.2558,
      "grad_norm": 0.1567823588848114,
      "learning_rate": 0.00010084040000000002,
      "loss": 0.2573,
      "step": 247900
    },
    {
      "epoch": 0.256,
      "grad_norm": 0.17192710936069489,
      "learning_rate": 0.0001008004,
      "loss": 0.2543,
      "step": 248000
    },
    {
      "epoch": 0.2562,
      "grad_norm": 0.18040655553340912,
      "learning_rate": 0.0001007604,
      "loss": 0.2632,
      "step": 248100
    },
    {
      "epoch": 0.2564,
      "grad_norm": 0.29884371161460876,
      "learning_rate": 0.0001007204,
      "loss": 0.2545,
      "step": 248200
    },
    {
      "epoch": 0.2566,
      "grad_norm": 0.2246190458536148,
      "learning_rate": 0.0001006804,
      "loss": 0.2507,
      "step": 248300
    },
    {
      "epoch": 0.2568,
      "grad_norm": 0.21513308584690094,
      "learning_rate": 0.00010064040000000001,
      "loss": 0.2517,
      "step": 248400
    },
    {
      "epoch": 0.257,
      "grad_norm": 0.17849339544773102,
      "learning_rate": 0.00010060039999999999,
      "loss": 0.253,
      "step": 248500
    },
    {
      "epoch": 0.2572,
      "grad_norm": 0.2598421573638916,
      "learning_rate": 0.0001005604,
      "loss": 0.2546,
      "step": 248600
    },
    {
      "epoch": 0.2574,
      "grad_norm": 0.21260765194892883,
      "learning_rate": 0.00010052040000000001,
      "loss": 0.2523,
      "step": 248700
    },
    {
      "epoch": 0.2576,
      "grad_norm": 0.9292067289352417,
      "learning_rate": 0.00010048040000000001,
      "loss": 0.2514,
      "step": 248800
    },
    {
      "epoch": 0.2578,
      "grad_norm": 0.1865740567445755,
      "learning_rate": 0.00010044040000000001,
      "loss": 0.2583,
      "step": 248900
    },
    {
      "epoch": 0.258,
      "grad_norm": 0.22997444868087769,
      "learning_rate": 0.00010040039999999999,
      "loss": 0.2542,
      "step": 249000
    },
    {
      "epoch": 0.2582,
      "grad_norm": 0.23303471505641937,
      "learning_rate": 0.0001003604,
      "loss": 0.2538,
      "step": 249100
    },
    {
      "epoch": 0.2584,
      "grad_norm": 0.26385053992271423,
      "learning_rate": 0.0001003204,
      "loss": 0.254,
      "step": 249200
    },
    {
      "epoch": 0.2586,
      "grad_norm": 0.36450210213661194,
      "learning_rate": 0.00010028040000000001,
      "loss": 0.261,
      "step": 249300
    },
    {
      "epoch": 0.2588,
      "grad_norm": 0.20195017755031586,
      "learning_rate": 0.00010024040000000002,
      "loss": 0.2497,
      "step": 249400
    },
    {
      "epoch": 0.259,
      "grad_norm": 0.24559420347213745,
      "learning_rate": 0.0001002004,
      "loss": 0.2556,
      "step": 249500
    },
    {
      "epoch": 0.2592,
      "grad_norm": 0.42150649428367615,
      "learning_rate": 0.0001001604,
      "loss": 0.257,
      "step": 249600
    },
    {
      "epoch": 0.2594,
      "grad_norm": 0.19844648241996765,
      "learning_rate": 0.0001001204,
      "loss": 0.2574,
      "step": 249700
    },
    {
      "epoch": 0.2596,
      "grad_norm": 0.18794132769107819,
      "learning_rate": 0.0001000804,
      "loss": 0.2507,
      "step": 249800
    },
    {
      "epoch": 0.2598,
      "grad_norm": 0.17367449402809143,
      "learning_rate": 0.00010004040000000001,
      "loss": 0.2526,
      "step": 249900
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.23622073233127594,
      "learning_rate": 0.00010000039999999999,
      "loss": 0.2523,
      "step": 250000
    },
    {
      "epoch": 0.2602,
      "grad_norm": 0.1616039127111435,
      "learning_rate": 9.996040000000001e-05,
      "loss": 0.2555,
      "step": 250100
    },
    {
      "epoch": 0.2604,
      "grad_norm": 0.1847952902317047,
      "learning_rate": 9.99204e-05,
      "loss": 0.2545,
      "step": 250200
    },
    {
      "epoch": 0.2606,
      "grad_norm": 0.2138368785381317,
      "learning_rate": 9.988040000000001e-05,
      "loss": 0.2485,
      "step": 250300
    },
    {
      "epoch": 0.2608,
      "grad_norm": 0.17640897631645203,
      "learning_rate": 9.984040000000001e-05,
      "loss": 0.2585,
      "step": 250400
    },
    {
      "epoch": 0.261,
      "grad_norm": 0.20163944363594055,
      "learning_rate": 9.98004e-05,
      "loss": 0.2486,
      "step": 250500
    },
    {
      "epoch": 0.2612,
      "grad_norm": 0.23935359716415405,
      "learning_rate": 9.976040000000001e-05,
      "loss": 0.2603,
      "step": 250600
    },
    {
      "epoch": 0.2614,
      "grad_norm": 0.14974956214427948,
      "learning_rate": 9.97204e-05,
      "loss": 0.2557,
      "step": 250700
    },
    {
      "epoch": 0.2616,
      "grad_norm": 0.17297634482383728,
      "learning_rate": 9.968040000000001e-05,
      "loss": 0.2562,
      "step": 250800
    },
    {
      "epoch": 0.2618,
      "grad_norm": 0.31046923995018005,
      "learning_rate": 9.96404e-05,
      "loss": 0.2569,
      "step": 250900
    },
    {
      "epoch": 0.262,
      "grad_norm": 0.1529819816350937,
      "learning_rate": 9.960040000000001e-05,
      "loss": 0.2559,
      "step": 251000
    },
    {
      "epoch": 0.2622,
      "grad_norm": 0.22081358730793,
      "learning_rate": 9.95604e-05,
      "loss": 0.2595,
      "step": 251100
    },
    {
      "epoch": 0.2624,
      "grad_norm": 0.2596571445465088,
      "learning_rate": 9.95204e-05,
      "loss": 0.2621,
      "step": 251200
    },
    {
      "epoch": 0.2626,
      "grad_norm": 0.20315712690353394,
      "learning_rate": 9.94804e-05,
      "loss": 0.2556,
      "step": 251300
    },
    {
      "epoch": 0.2628,
      "grad_norm": 0.16273678839206696,
      "learning_rate": 9.94404e-05,
      "loss": 0.2538,
      "step": 251400
    },
    {
      "epoch": 0.263,
      "grad_norm": 0.3179258108139038,
      "learning_rate": 9.94004e-05,
      "loss": 0.2547,
      "step": 251500
    },
    {
      "epoch": 0.2632,
      "grad_norm": 0.18335247039794922,
      "learning_rate": 9.936040000000001e-05,
      "loss": 0.275,
      "step": 251600
    },
    {
      "epoch": 0.2634,
      "grad_norm": 0.19399204850196838,
      "learning_rate": 9.93204e-05,
      "loss": 0.253,
      "step": 251700
    },
    {
      "epoch": 0.2636,
      "grad_norm": 0.16826485097408295,
      "learning_rate": 9.928040000000001e-05,
      "loss": 0.2511,
      "step": 251800
    },
    {
      "epoch": 0.2638,
      "grad_norm": 0.16970914602279663,
      "learning_rate": 9.92404e-05,
      "loss": 0.2485,
      "step": 251900
    },
    {
      "epoch": 0.264,
      "grad_norm": 0.19796758890151978,
      "learning_rate": 9.92004e-05,
      "loss": 0.25,
      "step": 252000
    },
    {
      "epoch": 0.2642,
      "grad_norm": 0.23124122619628906,
      "learning_rate": 9.916040000000001e-05,
      "loss": 0.2517,
      "step": 252100
    },
    {
      "epoch": 0.2644,
      "grad_norm": 0.18028624355793,
      "learning_rate": 9.91204e-05,
      "loss": 0.2549,
      "step": 252200
    },
    {
      "epoch": 0.2646,
      "grad_norm": 0.24230654537677765,
      "learning_rate": 9.908040000000001e-05,
      "loss": 0.2511,
      "step": 252300
    },
    {
      "epoch": 0.2648,
      "grad_norm": 0.16083365678787231,
      "learning_rate": 9.90404e-05,
      "loss": 0.2491,
      "step": 252400
    },
    {
      "epoch": 0.265,
      "grad_norm": 0.3353089392185211,
      "learning_rate": 9.900040000000001e-05,
      "loss": 0.2586,
      "step": 252500
    },
    {
      "epoch": 0.2652,
      "grad_norm": 0.29795363545417786,
      "learning_rate": 9.89604e-05,
      "loss": 0.2581,
      "step": 252600
    },
    {
      "epoch": 0.2654,
      "grad_norm": 0.18846677243709564,
      "learning_rate": 9.89204e-05,
      "loss": 0.2596,
      "step": 252700
    },
    {
      "epoch": 0.2656,
      "grad_norm": 0.16122601926326752,
      "learning_rate": 9.88804e-05,
      "loss": 0.2517,
      "step": 252800
    },
    {
      "epoch": 0.2658,
      "grad_norm": 0.19372184574604034,
      "learning_rate": 9.88404e-05,
      "loss": 0.2474,
      "step": 252900
    },
    {
      "epoch": 0.266,
      "grad_norm": 0.1936195194721222,
      "learning_rate": 9.88004e-05,
      "loss": 0.2546,
      "step": 253000
    },
    {
      "epoch": 0.2662,
      "grad_norm": 0.2815191149711609,
      "learning_rate": 9.876040000000001e-05,
      "loss": 0.2507,
      "step": 253100
    },
    {
      "epoch": 0.2664,
      "grad_norm": 0.16391977667808533,
      "learning_rate": 9.87204e-05,
      "loss": 0.2538,
      "step": 253200
    },
    {
      "epoch": 0.2666,
      "grad_norm": 0.17948247492313385,
      "learning_rate": 9.868040000000001e-05,
      "loss": 0.253,
      "step": 253300
    },
    {
      "epoch": 0.2668,
      "grad_norm": 0.18989241123199463,
      "learning_rate": 9.86404e-05,
      "loss": 0.2547,
      "step": 253400
    },
    {
      "epoch": 0.267,
      "grad_norm": 0.20617620646953583,
      "learning_rate": 9.86004e-05,
      "loss": 0.2549,
      "step": 253500
    },
    {
      "epoch": 0.2672,
      "grad_norm": 0.2014150768518448,
      "learning_rate": 9.85604e-05,
      "loss": 0.2588,
      "step": 253600
    },
    {
      "epoch": 0.2674,
      "grad_norm": 0.22781629860401154,
      "learning_rate": 9.85204e-05,
      "loss": 0.2625,
      "step": 253700
    },
    {
      "epoch": 0.2676,
      "grad_norm": 0.179660364985466,
      "learning_rate": 9.84804e-05,
      "loss": 0.2547,
      "step": 253800
    },
    {
      "epoch": 0.2678,
      "grad_norm": 0.18104757368564606,
      "learning_rate": 9.84404e-05,
      "loss": 0.2569,
      "step": 253900
    },
    {
      "epoch": 0.268,
      "grad_norm": 0.42554691433906555,
      "learning_rate": 9.840040000000001e-05,
      "loss": 0.2606,
      "step": 254000
    },
    {
      "epoch": 0.2682,
      "grad_norm": 0.17360180616378784,
      "learning_rate": 9.836040000000001e-05,
      "loss": 0.2583,
      "step": 254100
    },
    {
      "epoch": 0.2684,
      "grad_norm": 0.21442458033561707,
      "learning_rate": 9.83204e-05,
      "loss": 0.2525,
      "step": 254200
    },
    {
      "epoch": 0.2686,
      "grad_norm": 0.17544832825660706,
      "learning_rate": 9.82804e-05,
      "loss": 0.2512,
      "step": 254300
    },
    {
      "epoch": 0.2688,
      "grad_norm": 0.18496903777122498,
      "learning_rate": 9.824040000000001e-05,
      "loss": 0.2501,
      "step": 254400
    },
    {
      "epoch": 0.269,
      "grad_norm": 0.20425285398960114,
      "learning_rate": 9.82004e-05,
      "loss": 0.2529,
      "step": 254500
    },
    {
      "epoch": 0.2692,
      "grad_norm": 0.16543030738830566,
      "learning_rate": 9.816040000000001e-05,
      "loss": 0.2572,
      "step": 254600
    },
    {
      "epoch": 0.2694,
      "grad_norm": 0.18236897885799408,
      "learning_rate": 9.81204e-05,
      "loss": 0.2546,
      "step": 254700
    },
    {
      "epoch": 0.2696,
      "grad_norm": 0.35450029373168945,
      "learning_rate": 9.808040000000001e-05,
      "loss": 0.2519,
      "step": 254800
    },
    {
      "epoch": 0.2698,
      "grad_norm": 0.1963033676147461,
      "learning_rate": 9.80404e-05,
      "loss": 0.274,
      "step": 254900
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.24699929356575012,
      "learning_rate": 9.80004e-05,
      "loss": 0.2546,
      "step": 255000
    },
    {
      "epoch": 0.2702,
      "grad_norm": 0.2196625918149948,
      "learning_rate": 9.79604e-05,
      "loss": 0.2565,
      "step": 255100
    },
    {
      "epoch": 0.2704,
      "grad_norm": 0.2058703750371933,
      "learning_rate": 9.79204e-05,
      "loss": 0.2576,
      "step": 255200
    },
    {
      "epoch": 0.2706,
      "grad_norm": 0.16371659934520721,
      "learning_rate": 9.78804e-05,
      "loss": 0.2534,
      "step": 255300
    },
    {
      "epoch": 0.2708,
      "grad_norm": 0.249703511595726,
      "learning_rate": 9.784040000000001e-05,
      "loss": 0.251,
      "step": 255400
    },
    {
      "epoch": 0.271,
      "grad_norm": 0.18516069650650024,
      "learning_rate": 9.78004e-05,
      "loss": 0.2562,
      "step": 255500
    },
    {
      "epoch": 0.2712,
      "grad_norm": 0.17773735523223877,
      "learning_rate": 9.776040000000001e-05,
      "loss": 0.2583,
      "step": 255600
    },
    {
      "epoch": 0.2714,
      "grad_norm": 0.18748611211776733,
      "learning_rate": 9.772040000000001e-05,
      "loss": 0.2588,
      "step": 255700
    },
    {
      "epoch": 0.2716,
      "grad_norm": 0.16686369478702545,
      "learning_rate": 9.76804e-05,
      "loss": 0.2603,
      "step": 255800
    },
    {
      "epoch": 0.2718,
      "grad_norm": 0.17523817718029022,
      "learning_rate": 9.764040000000001e-05,
      "loss": 0.2466,
      "step": 255900
    },
    {
      "epoch": 0.272,
      "grad_norm": 0.2433302402496338,
      "learning_rate": 9.76004e-05,
      "loss": 0.2528,
      "step": 256000
    },
    {
      "epoch": 0.2722,
      "grad_norm": 0.14348742365837097,
      "learning_rate": 9.756040000000001e-05,
      "loss": 0.2494,
      "step": 256100
    },
    {
      "epoch": 0.2724,
      "grad_norm": 0.2618894577026367,
      "learning_rate": 9.75204e-05,
      "loss": 0.2606,
      "step": 256200
    },
    {
      "epoch": 0.2726,
      "grad_norm": 0.2586542069911957,
      "learning_rate": 9.748040000000001e-05,
      "loss": 0.2476,
      "step": 256300
    },
    {
      "epoch": 0.2728,
      "grad_norm": 0.17183682322502136,
      "learning_rate": 9.74404e-05,
      "loss": 0.253,
      "step": 256400
    },
    {
      "epoch": 0.273,
      "grad_norm": 0.15170541405677795,
      "learning_rate": 9.74004e-05,
      "loss": 0.253,
      "step": 256500
    },
    {
      "epoch": 0.2732,
      "grad_norm": 0.20044559240341187,
      "learning_rate": 9.73604e-05,
      "loss": 0.2594,
      "step": 256600
    },
    {
      "epoch": 0.2734,
      "grad_norm": 0.1726810783147812,
      "learning_rate": 9.73204e-05,
      "loss": 0.2544,
      "step": 256700
    },
    {
      "epoch": 0.2736,
      "grad_norm": 0.17345348000526428,
      "learning_rate": 9.72804e-05,
      "loss": 0.2543,
      "step": 256800
    },
    {
      "epoch": 0.2738,
      "grad_norm": 0.1733126938343048,
      "learning_rate": 9.724040000000001e-05,
      "loss": 0.2516,
      "step": 256900
    },
    {
      "epoch": 0.274,
      "grad_norm": 0.2917121946811676,
      "learning_rate": 9.72004e-05,
      "loss": 0.2605,
      "step": 257000
    },
    {
      "epoch": 0.2742,
      "grad_norm": 0.1882348358631134,
      "learning_rate": 9.716040000000001e-05,
      "loss": 0.2584,
      "step": 257100
    },
    {
      "epoch": 0.2744,
      "grad_norm": 0.1900627166032791,
      "learning_rate": 9.71204e-05,
      "loss": 0.2614,
      "step": 257200
    },
    {
      "epoch": 0.2746,
      "grad_norm": 0.21451936662197113,
      "learning_rate": 9.70804e-05,
      "loss": 0.255,
      "step": 257300
    },
    {
      "epoch": 0.2748,
      "grad_norm": 0.24926543235778809,
      "learning_rate": 9.704040000000001e-05,
      "loss": 0.2502,
      "step": 257400
    },
    {
      "epoch": 0.275,
      "grad_norm": 0.19679979979991913,
      "learning_rate": 9.70004e-05,
      "loss": 0.2504,
      "step": 257500
    },
    {
      "epoch": 0.2752,
      "grad_norm": 0.27472445368766785,
      "learning_rate": 9.696040000000001e-05,
      "loss": 0.2499,
      "step": 257600
    },
    {
      "epoch": 0.2754,
      "grad_norm": 0.1651807725429535,
      "learning_rate": 9.69204e-05,
      "loss": 0.2574,
      "step": 257700
    },
    {
      "epoch": 0.2756,
      "grad_norm": 0.2149466574192047,
      "learning_rate": 9.688040000000001e-05,
      "loss": 0.2466,
      "step": 257800
    },
    {
      "epoch": 0.2758,
      "grad_norm": 0.21232870221138,
      "learning_rate": 9.684040000000002e-05,
      "loss": 0.2539,
      "step": 257900
    },
    {
      "epoch": 0.276,
      "grad_norm": 0.20928317308425903,
      "learning_rate": 9.68004e-05,
      "loss": 0.2496,
      "step": 258000
    },
    {
      "epoch": 0.2762,
      "grad_norm": 0.2095847725868225,
      "learning_rate": 9.67604e-05,
      "loss": 0.2552,
      "step": 258100
    },
    {
      "epoch": 0.2764,
      "grad_norm": 0.1795085370540619,
      "learning_rate": 9.67204e-05,
      "loss": 0.2495,
      "step": 258200
    },
    {
      "epoch": 0.2766,
      "grad_norm": 0.17618554830551147,
      "learning_rate": 9.66804e-05,
      "loss": 0.252,
      "step": 258300
    },
    {
      "epoch": 0.2768,
      "grad_norm": 0.16858777403831482,
      "learning_rate": 9.664040000000001e-05,
      "loss": 0.2585,
      "step": 258400
    },
    {
      "epoch": 0.277,
      "grad_norm": 0.21656472980976105,
      "learning_rate": 9.66004e-05,
      "loss": 0.2498,
      "step": 258500
    },
    {
      "epoch": 0.2772,
      "grad_norm": 0.18165180087089539,
      "learning_rate": 9.656040000000001e-05,
      "loss": 0.2529,
      "step": 258600
    },
    {
      "epoch": 0.2774,
      "grad_norm": 0.2290857881307602,
      "learning_rate": 9.65204e-05,
      "loss": 0.25,
      "step": 258700
    },
    {
      "epoch": 0.2776,
      "grad_norm": 0.1842830330133438,
      "learning_rate": 9.64804e-05,
      "loss": 0.2526,
      "step": 258800
    },
    {
      "epoch": 0.2778,
      "grad_norm": 0.16289664804935455,
      "learning_rate": 9.64404e-05,
      "loss": 0.2546,
      "step": 258900
    },
    {
      "epoch": 0.278,
      "grad_norm": 0.22200186550617218,
      "learning_rate": 9.64004e-05,
      "loss": 0.2487,
      "step": 259000
    },
    {
      "epoch": 0.2782,
      "grad_norm": 0.15263180434703827,
      "learning_rate": 9.63604e-05,
      "loss": 0.2532,
      "step": 259100
    },
    {
      "epoch": 0.2784,
      "grad_norm": 0.19079795479774475,
      "learning_rate": 9.63204e-05,
      "loss": 0.2463,
      "step": 259200
    },
    {
      "epoch": 0.2786,
      "grad_norm": 0.21255415678024292,
      "learning_rate": 9.628040000000001e-05,
      "loss": 0.2485,
      "step": 259300
    },
    {
      "epoch": 0.2788,
      "grad_norm": 0.15365895628929138,
      "learning_rate": 9.624040000000001e-05,
      "loss": 0.2472,
      "step": 259400
    },
    {
      "epoch": 0.279,
      "grad_norm": 0.1517648547887802,
      "learning_rate": 9.620040000000001e-05,
      "loss": 0.2538,
      "step": 259500
    },
    {
      "epoch": 0.2792,
      "grad_norm": 0.2428562343120575,
      "learning_rate": 9.61604e-05,
      "loss": 0.2455,
      "step": 259600
    },
    {
      "epoch": 0.2794,
      "grad_norm": 0.15415999293327332,
      "learning_rate": 9.61204e-05,
      "loss": 0.2472,
      "step": 259700
    },
    {
      "epoch": 0.2796,
      "grad_norm": 0.19108308851718903,
      "learning_rate": 9.60804e-05,
      "loss": 0.2482,
      "step": 259800
    },
    {
      "epoch": 0.2798,
      "grad_norm": 0.1723984181880951,
      "learning_rate": 9.604040000000001e-05,
      "loss": 0.2502,
      "step": 259900
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.24995626509189606,
      "learning_rate": 9.60004e-05,
      "loss": 0.2578,
      "step": 260000
    },
    {
      "epoch": 0.2802,
      "grad_norm": 0.17275862395763397,
      "learning_rate": 9.596040000000001e-05,
      "loss": 0.2583,
      "step": 260100
    },
    {
      "epoch": 0.2804,
      "grad_norm": 0.20496828854084015,
      "learning_rate": 9.59204e-05,
      "loss": 0.2502,
      "step": 260200
    },
    {
      "epoch": 0.2806,
      "grad_norm": 0.15254418551921844,
      "learning_rate": 9.58804e-05,
      "loss": 0.2453,
      "step": 260300
    },
    {
      "epoch": 0.2808,
      "grad_norm": 0.22347435355186462,
      "learning_rate": 9.58404e-05,
      "loss": 0.2655,
      "step": 260400
    },
    {
      "epoch": 0.281,
      "grad_norm": 0.27912843227386475,
      "learning_rate": 9.58004e-05,
      "loss": 0.2595,
      "step": 260500
    },
    {
      "epoch": 0.2812,
      "grad_norm": 0.17391923069953918,
      "learning_rate": 9.57604e-05,
      "loss": 0.2441,
      "step": 260600
    },
    {
      "epoch": 0.2814,
      "grad_norm": 0.20242680609226227,
      "learning_rate": 9.572040000000001e-05,
      "loss": 0.2513,
      "step": 260700
    },
    {
      "epoch": 0.2816,
      "grad_norm": 0.19099730253219604,
      "learning_rate": 9.56804e-05,
      "loss": 0.2474,
      "step": 260800
    },
    {
      "epoch": 0.2818,
      "grad_norm": 0.17173656821250916,
      "learning_rate": 9.564040000000001e-05,
      "loss": 0.2468,
      "step": 260900
    },
    {
      "epoch": 0.282,
      "grad_norm": 0.2755275368690491,
      "learning_rate": 9.560040000000001e-05,
      "loss": 0.2579,
      "step": 261000
    },
    {
      "epoch": 0.2822,
      "grad_norm": 0.1447450965642929,
      "learning_rate": 9.55604e-05,
      "loss": 0.2549,
      "step": 261100
    },
    {
      "epoch": 0.2824,
      "grad_norm": 0.17367517948150635,
      "learning_rate": 9.552040000000001e-05,
      "loss": 0.2493,
      "step": 261200
    },
    {
      "epoch": 0.2826,
      "grad_norm": 0.19174928963184357,
      "learning_rate": 9.54804e-05,
      "loss": 0.2518,
      "step": 261300
    },
    {
      "epoch": 0.2828,
      "grad_norm": 0.20758305490016937,
      "learning_rate": 9.544040000000001e-05,
      "loss": 0.2522,
      "step": 261400
    },
    {
      "epoch": 0.283,
      "grad_norm": 0.172040194272995,
      "learning_rate": 9.54004e-05,
      "loss": 0.2527,
      "step": 261500
    },
    {
      "epoch": 0.2832,
      "grad_norm": 0.19083595275878906,
      "learning_rate": 9.536040000000001e-05,
      "loss": 0.2439,
      "step": 261600
    },
    {
      "epoch": 0.2834,
      "grad_norm": 0.161021426320076,
      "learning_rate": 9.53204e-05,
      "loss": 0.2529,
      "step": 261700
    },
    {
      "epoch": 0.2836,
      "grad_norm": 0.1945277452468872,
      "learning_rate": 9.52804e-05,
      "loss": 0.2508,
      "step": 261800
    },
    {
      "epoch": 0.2838,
      "grad_norm": 0.18637241423130035,
      "learning_rate": 9.52404e-05,
      "loss": 0.2504,
      "step": 261900
    },
    {
      "epoch": 0.284,
      "grad_norm": 0.6245195865631104,
      "learning_rate": 9.52004e-05,
      "loss": 0.2512,
      "step": 262000
    },
    {
      "epoch": 0.2842,
      "grad_norm": 0.26768702268600464,
      "learning_rate": 9.51604e-05,
      "loss": 0.2504,
      "step": 262100
    },
    {
      "epoch": 0.2844,
      "grad_norm": 0.19558322429656982,
      "learning_rate": 9.512040000000001e-05,
      "loss": 0.2543,
      "step": 262200
    },
    {
      "epoch": 0.2846,
      "grad_norm": 0.19235068559646606,
      "learning_rate": 9.50804e-05,
      "loss": 0.2491,
      "step": 262300
    },
    {
      "epoch": 0.2848,
      "grad_norm": 0.19615627825260162,
      "learning_rate": 9.504040000000001e-05,
      "loss": 0.2486,
      "step": 262400
    },
    {
      "epoch": 0.285,
      "grad_norm": 0.4205188751220703,
      "learning_rate": 9.50004e-05,
      "loss": 0.2507,
      "step": 262500
    },
    {
      "epoch": 0.2852,
      "grad_norm": 0.2430608868598938,
      "learning_rate": 9.49604e-05,
      "loss": 0.2509,
      "step": 262600
    },
    {
      "epoch": 0.2854,
      "grad_norm": 0.17347165942192078,
      "learning_rate": 9.492040000000001e-05,
      "loss": 0.2495,
      "step": 262700
    },
    {
      "epoch": 0.2856,
      "grad_norm": 0.256759375333786,
      "learning_rate": 9.48804e-05,
      "loss": 0.2503,
      "step": 262800
    },
    {
      "epoch": 0.2858,
      "grad_norm": 0.16245199739933014,
      "learning_rate": 9.484040000000001e-05,
      "loss": 0.2465,
      "step": 262900
    },
    {
      "epoch": 0.286,
      "grad_norm": 0.17991042137145996,
      "learning_rate": 9.48004e-05,
      "loss": 0.2503,
      "step": 263000
    },
    {
      "epoch": 0.2862,
      "grad_norm": 0.1637287586927414,
      "learning_rate": 9.476040000000001e-05,
      "loss": 0.2472,
      "step": 263100
    },
    {
      "epoch": 0.2864,
      "grad_norm": 0.2055928260087967,
      "learning_rate": 9.472040000000002e-05,
      "loss": 0.2495,
      "step": 263200
    },
    {
      "epoch": 0.2866,
      "grad_norm": 0.19248582422733307,
      "learning_rate": 9.468040000000001e-05,
      "loss": 0.2465,
      "step": 263300
    },
    {
      "epoch": 0.2868,
      "grad_norm": 0.19688032567501068,
      "learning_rate": 9.46404e-05,
      "loss": 0.2486,
      "step": 263400
    },
    {
      "epoch": 0.287,
      "grad_norm": 0.1787942796945572,
      "learning_rate": 9.46004e-05,
      "loss": 0.2483,
      "step": 263500
    },
    {
      "epoch": 0.2872,
      "grad_norm": 0.18415433168411255,
      "learning_rate": 9.45604e-05,
      "loss": 0.2566,
      "step": 263600
    },
    {
      "epoch": 0.2874,
      "grad_norm": 0.19202789664268494,
      "learning_rate": 9.452040000000001e-05,
      "loss": 0.2498,
      "step": 263700
    },
    {
      "epoch": 0.2876,
      "grad_norm": 0.27513137459754944,
      "learning_rate": 9.44804e-05,
      "loss": 0.2508,
      "step": 263800
    },
    {
      "epoch": 0.2878,
      "grad_norm": 0.17416657507419586,
      "learning_rate": 9.444040000000001e-05,
      "loss": 0.2501,
      "step": 263900
    },
    {
      "epoch": 0.288,
      "grad_norm": 0.15521082282066345,
      "learning_rate": 9.44004e-05,
      "loss": 0.2462,
      "step": 264000
    },
    {
      "epoch": 0.2882,
      "grad_norm": 0.17850640416145325,
      "learning_rate": 9.43604e-05,
      "loss": 0.2577,
      "step": 264100
    },
    {
      "epoch": 0.2884,
      "grad_norm": 0.28766483068466187,
      "learning_rate": 9.43204e-05,
      "loss": 0.2539,
      "step": 264200
    },
    {
      "epoch": 0.2886,
      "grad_norm": 0.1784902662038803,
      "learning_rate": 9.42804e-05,
      "loss": 0.2531,
      "step": 264300
    },
    {
      "epoch": 0.2888,
      "grad_norm": 0.19980043172836304,
      "learning_rate": 9.42404e-05,
      "loss": 0.2467,
      "step": 264400
    },
    {
      "epoch": 0.289,
      "grad_norm": 1.2848162651062012,
      "learning_rate": 9.42004e-05,
      "loss": 0.2493,
      "step": 264500
    },
    {
      "epoch": 0.2892,
      "grad_norm": 0.20343329012393951,
      "learning_rate": 9.416040000000001e-05,
      "loss": 0.248,
      "step": 264600
    },
    {
      "epoch": 0.2894,
      "grad_norm": 0.22436699271202087,
      "learning_rate": 9.412040000000001e-05,
      "loss": 0.2492,
      "step": 264700
    },
    {
      "epoch": 0.2896,
      "grad_norm": 0.19231826066970825,
      "learning_rate": 9.408040000000001e-05,
      "loss": 0.2457,
      "step": 264800
    },
    {
      "epoch": 0.2898,
      "grad_norm": 0.20982171595096588,
      "learning_rate": 9.40404e-05,
      "loss": 0.2461,
      "step": 264900
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.25590869784355164,
      "learning_rate": 9.40004e-05,
      "loss": 0.2511,
      "step": 265000
    },
    {
      "epoch": 0.2902,
      "grad_norm": 0.19312362372875214,
      "learning_rate": 9.39604e-05,
      "loss": 0.2466,
      "step": 265100
    },
    {
      "epoch": 0.2904,
      "grad_norm": 0.23638297617435455,
      "learning_rate": 9.392040000000001e-05,
      "loss": 0.2466,
      "step": 265200
    },
    {
      "epoch": 0.2906,
      "grad_norm": 0.16785204410552979,
      "learning_rate": 9.38804e-05,
      "loss": 0.2469,
      "step": 265300
    },
    {
      "epoch": 0.2908,
      "grad_norm": 0.2154064029455185,
      "learning_rate": 9.384040000000001e-05,
      "loss": 0.2706,
      "step": 265400
    },
    {
      "epoch": 0.291,
      "grad_norm": 0.18391206860542297,
      "learning_rate": 9.38004e-05,
      "loss": 0.243,
      "step": 265500
    },
    {
      "epoch": 0.2912,
      "grad_norm": 0.20186170935630798,
      "learning_rate": 9.37604e-05,
      "loss": 0.2555,
      "step": 265600
    },
    {
      "epoch": 0.2914,
      "grad_norm": 0.19941025972366333,
      "learning_rate": 9.37204e-05,
      "loss": 0.248,
      "step": 265700
    },
    {
      "epoch": 0.2916,
      "grad_norm": 0.18308115005493164,
      "learning_rate": 9.36804e-05,
      "loss": 0.2464,
      "step": 265800
    },
    {
      "epoch": 0.2918,
      "grad_norm": 0.24809174239635468,
      "learning_rate": 9.36404e-05,
      "loss": 0.2536,
      "step": 265900
    },
    {
      "epoch": 0.292,
      "grad_norm": 0.2352696657180786,
      "learning_rate": 9.36004e-05,
      "loss": 0.2515,
      "step": 266000
    },
    {
      "epoch": 0.2922,
      "grad_norm": 0.20043088495731354,
      "learning_rate": 9.35604e-05,
      "loss": 0.249,
      "step": 266100
    },
    {
      "epoch": 0.2924,
      "grad_norm": 0.33282342553138733,
      "learning_rate": 9.352040000000001e-05,
      "loss": 0.2588,
      "step": 266200
    },
    {
      "epoch": 0.2926,
      "grad_norm": 0.1646975576877594,
      "learning_rate": 9.348040000000001e-05,
      "loss": 0.2571,
      "step": 266300
    },
    {
      "epoch": 0.2928,
      "grad_norm": 0.2019079327583313,
      "learning_rate": 9.34404e-05,
      "loss": 0.2501,
      "step": 266400
    },
    {
      "epoch": 0.293,
      "grad_norm": 0.3303513526916504,
      "learning_rate": 9.34004e-05,
      "loss": 0.2547,
      "step": 266500
    },
    {
      "epoch": 0.2932,
      "grad_norm": 0.1550179123878479,
      "learning_rate": 9.33604e-05,
      "loss": 0.2554,
      "step": 266600
    },
    {
      "epoch": 0.2934,
      "grad_norm": 0.3752208948135376,
      "learning_rate": 9.332040000000001e-05,
      "loss": 0.2493,
      "step": 266700
    },
    {
      "epoch": 0.2936,
      "grad_norm": 0.18411359190940857,
      "learning_rate": 9.32804e-05,
      "loss": 0.2569,
      "step": 266800
    },
    {
      "epoch": 0.2938,
      "grad_norm": 0.17852115631103516,
      "learning_rate": 9.324040000000001e-05,
      "loss": 0.2479,
      "step": 266900
    },
    {
      "epoch": 0.294,
      "grad_norm": 0.17118707299232483,
      "learning_rate": 9.320040000000002e-05,
      "loss": 0.254,
      "step": 267000
    },
    {
      "epoch": 0.2942,
      "grad_norm": 0.35995370149612427,
      "learning_rate": 9.31604e-05,
      "loss": 0.2512,
      "step": 267100
    },
    {
      "epoch": 0.2944,
      "grad_norm": 0.27355167269706726,
      "learning_rate": 9.31204e-05,
      "loss": 0.2549,
      "step": 267200
    },
    {
      "epoch": 0.2946,
      "grad_norm": 0.1623479276895523,
      "learning_rate": 9.30804e-05,
      "loss": 0.2522,
      "step": 267300
    },
    {
      "epoch": 0.2948,
      "grad_norm": 0.26281481981277466,
      "learning_rate": 9.30404e-05,
      "loss": 0.2505,
      "step": 267400
    },
    {
      "epoch": 0.295,
      "grad_norm": 0.17927655577659607,
      "learning_rate": 9.300040000000001e-05,
      "loss": 0.2497,
      "step": 267500
    },
    {
      "epoch": 0.2952,
      "grad_norm": 0.1781659871339798,
      "learning_rate": 9.29604e-05,
      "loss": 0.2497,
      "step": 267600
    },
    {
      "epoch": 0.2954,
      "grad_norm": 0.16235920786857605,
      "learning_rate": 9.292040000000001e-05,
      "loss": 0.2567,
      "step": 267700
    },
    {
      "epoch": 0.2956,
      "grad_norm": 0.1638641655445099,
      "learning_rate": 9.28804e-05,
      "loss": 0.2469,
      "step": 267800
    },
    {
      "epoch": 0.2958,
      "grad_norm": 0.2075788974761963,
      "learning_rate": 9.28404e-05,
      "loss": 0.2482,
      "step": 267900
    },
    {
      "epoch": 0.296,
      "grad_norm": 0.1806693822145462,
      "learning_rate": 9.280040000000001e-05,
      "loss": 0.2516,
      "step": 268000
    },
    {
      "epoch": 0.2962,
      "grad_norm": 0.22953234612941742,
      "learning_rate": 9.27604e-05,
      "loss": 0.2631,
      "step": 268100
    },
    {
      "epoch": 0.2964,
      "grad_norm": 0.22682808339595795,
      "learning_rate": 9.272040000000001e-05,
      "loss": 0.2532,
      "step": 268200
    },
    {
      "epoch": 0.2966,
      "grad_norm": 0.3644724488258362,
      "learning_rate": 9.26804e-05,
      "loss": 0.2521,
      "step": 268300
    },
    {
      "epoch": 0.2968,
      "grad_norm": 0.18433807790279388,
      "learning_rate": 9.264040000000001e-05,
      "loss": 0.2578,
      "step": 268400
    },
    {
      "epoch": 0.297,
      "grad_norm": 0.1955367773771286,
      "learning_rate": 9.260040000000002e-05,
      "loss": 0.2493,
      "step": 268500
    },
    {
      "epoch": 0.2972,
      "grad_norm": 0.2631000578403473,
      "learning_rate": 9.256040000000001e-05,
      "loss": 0.2487,
      "step": 268600
    },
    {
      "epoch": 0.2974,
      "grad_norm": 0.24312226474285126,
      "learning_rate": 9.25204e-05,
      "loss": 0.2501,
      "step": 268700
    },
    {
      "epoch": 0.2976,
      "grad_norm": 0.19950313866138458,
      "learning_rate": 9.24804e-05,
      "loss": 0.2504,
      "step": 268800
    },
    {
      "epoch": 0.2978,
      "grad_norm": 0.21298249065876007,
      "learning_rate": 9.24404e-05,
      "loss": 0.2492,
      "step": 268900
    },
    {
      "epoch": 0.298,
      "grad_norm": 0.27059680223464966,
      "learning_rate": 9.240040000000001e-05,
      "loss": 0.2585,
      "step": 269000
    },
    {
      "epoch": 0.2982,
      "grad_norm": 0.17913617193698883,
      "learning_rate": 9.23604e-05,
      "loss": 0.2426,
      "step": 269100
    },
    {
      "epoch": 0.2984,
      "grad_norm": 0.19647064805030823,
      "learning_rate": 9.232040000000001e-05,
      "loss": 0.2485,
      "step": 269200
    },
    {
      "epoch": 0.2986,
      "grad_norm": 0.14924323558807373,
      "learning_rate": 9.22804e-05,
      "loss": 0.2514,
      "step": 269300
    },
    {
      "epoch": 0.2988,
      "grad_norm": 0.18103660643100739,
      "learning_rate": 9.22404e-05,
      "loss": 0.253,
      "step": 269400
    },
    {
      "epoch": 0.299,
      "grad_norm": 0.16981446743011475,
      "learning_rate": 9.22004e-05,
      "loss": 0.2506,
      "step": 269500
    },
    {
      "epoch": 0.2992,
      "grad_norm": 0.17313528060913086,
      "learning_rate": 9.21604e-05,
      "loss": 0.255,
      "step": 269600
    },
    {
      "epoch": 0.2994,
      "grad_norm": 0.19472962617874146,
      "learning_rate": 9.21204e-05,
      "loss": 0.2462,
      "step": 269700
    },
    {
      "epoch": 0.2996,
      "grad_norm": 0.20072679221630096,
      "learning_rate": 9.20804e-05,
      "loss": 0.2499,
      "step": 269800
    },
    {
      "epoch": 0.2998,
      "grad_norm": 0.20446142554283142,
      "learning_rate": 9.204040000000001e-05,
      "loss": 0.2484,
      "step": 269900
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.23795530200004578,
      "learning_rate": 9.200040000000001e-05,
      "loss": 0.2488,
      "step": 270000
    },
    {
      "epoch": 0.3002,
      "grad_norm": 0.5230748057365417,
      "learning_rate": 9.196040000000001e-05,
      "loss": 0.2517,
      "step": 270100
    },
    {
      "epoch": 0.3004,
      "grad_norm": 0.22013844549655914,
      "learning_rate": 9.19204e-05,
      "loss": 0.2482,
      "step": 270200
    },
    {
      "epoch": 0.3006,
      "grad_norm": 0.20569878816604614,
      "learning_rate": 9.18804e-05,
      "loss": 0.2592,
      "step": 270300
    },
    {
      "epoch": 0.3008,
      "grad_norm": 0.16592691838741302,
      "learning_rate": 9.18404e-05,
      "loss": 0.2539,
      "step": 270400
    },
    {
      "epoch": 0.301,
      "grad_norm": 0.20824261009693146,
      "learning_rate": 9.180040000000001e-05,
      "loss": 0.2523,
      "step": 270500
    },
    {
      "epoch": 0.3012,
      "grad_norm": 0.21853414177894592,
      "learning_rate": 9.17604e-05,
      "loss": 0.2476,
      "step": 270600
    },
    {
      "epoch": 0.3014,
      "grad_norm": 0.15546807646751404,
      "learning_rate": 9.172040000000001e-05,
      "loss": 0.2489,
      "step": 270700
    },
    {
      "epoch": 0.3016,
      "grad_norm": 0.19030147790908813,
      "learning_rate": 9.16804e-05,
      "loss": 0.2444,
      "step": 270800
    },
    {
      "epoch": 0.3018,
      "grad_norm": 0.20716990530490875,
      "learning_rate": 9.16404e-05,
      "loss": 0.2479,
      "step": 270900
    },
    {
      "epoch": 0.302,
      "grad_norm": 0.1584189087152481,
      "learning_rate": 9.16004e-05,
      "loss": 0.2461,
      "step": 271000
    },
    {
      "epoch": 0.3022,
      "grad_norm": 0.15803031623363495,
      "learning_rate": 9.15604e-05,
      "loss": 0.2492,
      "step": 271100
    },
    {
      "epoch": 0.3024,
      "grad_norm": 0.18134865164756775,
      "learning_rate": 9.15204e-05,
      "loss": 0.2458,
      "step": 271200
    },
    {
      "epoch": 0.3026,
      "grad_norm": 0.17175163328647614,
      "learning_rate": 9.14804e-05,
      "loss": 0.2507,
      "step": 271300
    },
    {
      "epoch": 0.3028,
      "grad_norm": 0.13778148591518402,
      "learning_rate": 9.14404e-05,
      "loss": 0.2517,
      "step": 271400
    },
    {
      "epoch": 0.303,
      "grad_norm": 0.18867234885692596,
      "learning_rate": 9.140040000000001e-05,
      "loss": 0.2477,
      "step": 271500
    },
    {
      "epoch": 0.3032,
      "grad_norm": 0.18188384175300598,
      "learning_rate": 9.136040000000001e-05,
      "loss": 0.2564,
      "step": 271600
    },
    {
      "epoch": 0.3034,
      "grad_norm": 0.18826958537101746,
      "learning_rate": 9.13204e-05,
      "loss": 0.2468,
      "step": 271700
    },
    {
      "epoch": 0.3036,
      "grad_norm": 0.18992917239665985,
      "learning_rate": 9.12804e-05,
      "loss": 0.2521,
      "step": 271800
    },
    {
      "epoch": 0.3038,
      "grad_norm": 0.2327507585287094,
      "learning_rate": 9.12404e-05,
      "loss": 0.2514,
      "step": 271900
    },
    {
      "epoch": 0.304,
      "grad_norm": 0.24102598428726196,
      "learning_rate": 9.120040000000001e-05,
      "loss": 0.2539,
      "step": 272000
    },
    {
      "epoch": 0.3042,
      "grad_norm": 0.16849716007709503,
      "learning_rate": 9.11604e-05,
      "loss": 0.2486,
      "step": 272100
    },
    {
      "epoch": 0.3044,
      "grad_norm": 0.34158793091773987,
      "learning_rate": 9.112040000000001e-05,
      "loss": 0.2524,
      "step": 272200
    },
    {
      "epoch": 0.3046,
      "grad_norm": 0.29653236269950867,
      "learning_rate": 9.10804e-05,
      "loss": 0.259,
      "step": 272300
    },
    {
      "epoch": 0.3048,
      "grad_norm": 0.19480286538600922,
      "learning_rate": 9.104040000000001e-05,
      "loss": 0.2505,
      "step": 272400
    },
    {
      "epoch": 0.305,
      "grad_norm": 0.19145742058753967,
      "learning_rate": 9.10004e-05,
      "loss": 0.2539,
      "step": 272500
    },
    {
      "epoch": 0.3052,
      "grad_norm": 0.5713980793952942,
      "learning_rate": 9.09604e-05,
      "loss": 0.2558,
      "step": 272600
    },
    {
      "epoch": 0.3054,
      "grad_norm": 0.2311251163482666,
      "learning_rate": 9.09204e-05,
      "loss": 0.2525,
      "step": 272700
    },
    {
      "epoch": 0.3056,
      "grad_norm": 0.17040377855300903,
      "learning_rate": 9.08804e-05,
      "loss": 0.2611,
      "step": 272800
    },
    {
      "epoch": 0.3058,
      "grad_norm": 0.16640836000442505,
      "learning_rate": 9.08404e-05,
      "loss": 0.2485,
      "step": 272900
    },
    {
      "epoch": 0.306,
      "grad_norm": 0.18391956388950348,
      "learning_rate": 9.080040000000001e-05,
      "loss": 0.2515,
      "step": 273000
    },
    {
      "epoch": 0.3062,
      "grad_norm": 0.20270375907421112,
      "learning_rate": 9.07604e-05,
      "loss": 0.2495,
      "step": 273100
    },
    {
      "epoch": 0.3064,
      "grad_norm": 0.1664918065071106,
      "learning_rate": 9.07204e-05,
      "loss": 0.253,
      "step": 273200
    },
    {
      "epoch": 0.3066,
      "grad_norm": 0.2487817406654358,
      "learning_rate": 9.068040000000001e-05,
      "loss": 0.2534,
      "step": 273300
    },
    {
      "epoch": 0.3068,
      "grad_norm": 0.19376716017723083,
      "learning_rate": 9.06404e-05,
      "loss": 0.2447,
      "step": 273400
    },
    {
      "epoch": 0.307,
      "grad_norm": 1.2538508176803589,
      "learning_rate": 9.060040000000001e-05,
      "loss": 0.2626,
      "step": 273500
    },
    {
      "epoch": 0.3072,
      "grad_norm": 0.18994612991809845,
      "learning_rate": 9.05604e-05,
      "loss": 0.269,
      "step": 273600
    },
    {
      "epoch": 0.3074,
      "grad_norm": 0.1983320116996765,
      "learning_rate": 9.052040000000001e-05,
      "loss": 0.2484,
      "step": 273700
    },
    {
      "epoch": 0.3076,
      "grad_norm": 0.17373573780059814,
      "learning_rate": 9.048040000000002e-05,
      "loss": 0.245,
      "step": 273800
    },
    {
      "epoch": 0.3078,
      "grad_norm": 0.2375088334083557,
      "learning_rate": 9.044040000000001e-05,
      "loss": 0.2467,
      "step": 273900
    },
    {
      "epoch": 0.308,
      "grad_norm": 0.16651317477226257,
      "learning_rate": 9.04004e-05,
      "loss": 0.2551,
      "step": 274000
    },
    {
      "epoch": 0.3082,
      "grad_norm": 0.27858221530914307,
      "learning_rate": 9.03604e-05,
      "loss": 0.2434,
      "step": 274100
    },
    {
      "epoch": 0.3084,
      "grad_norm": 0.18154625594615936,
      "learning_rate": 9.03204e-05,
      "loss": 0.2591,
      "step": 274200
    },
    {
      "epoch": 0.3086,
      "grad_norm": 0.2042841613292694,
      "learning_rate": 9.028040000000001e-05,
      "loss": 0.2435,
      "step": 274300
    },
    {
      "epoch": 0.3088,
      "grad_norm": 0.2147081345319748,
      "learning_rate": 9.02404e-05,
      "loss": 0.2484,
      "step": 274400
    },
    {
      "epoch": 0.309,
      "grad_norm": 0.1780577152967453,
      "learning_rate": 9.020040000000001e-05,
      "loss": 0.2476,
      "step": 274500
    },
    {
      "epoch": 0.3092,
      "grad_norm": 0.2056666761636734,
      "learning_rate": 9.01604e-05,
      "loss": 0.256,
      "step": 274600
    },
    {
      "epoch": 0.3094,
      "grad_norm": 0.27037736773490906,
      "learning_rate": 9.01204e-05,
      "loss": 0.2486,
      "step": 274700
    },
    {
      "epoch": 0.3096,
      "grad_norm": 0.17876963317394257,
      "learning_rate": 9.00804e-05,
      "loss": 0.253,
      "step": 274800
    },
    {
      "epoch": 0.3098,
      "grad_norm": 0.1852559596300125,
      "learning_rate": 9.00404e-05,
      "loss": 0.2535,
      "step": 274900
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.2233329564332962,
      "learning_rate": 9.00004e-05,
      "loss": 0.2478,
      "step": 275000
    },
    {
      "epoch": 0.3102,
      "grad_norm": 0.2136661559343338,
      "learning_rate": 8.99604e-05,
      "loss": 0.2555,
      "step": 275100
    },
    {
      "epoch": 0.3104,
      "grad_norm": 0.2538791000843048,
      "learning_rate": 8.992040000000001e-05,
      "loss": 0.2453,
      "step": 275200
    },
    {
      "epoch": 0.3106,
      "grad_norm": 0.1559320092201233,
      "learning_rate": 8.988040000000001e-05,
      "loss": 0.2462,
      "step": 275300
    },
    {
      "epoch": 0.3108,
      "grad_norm": 0.21376411616802216,
      "learning_rate": 8.984040000000001e-05,
      "loss": 0.247,
      "step": 275400
    },
    {
      "epoch": 0.311,
      "grad_norm": 0.20133382081985474,
      "learning_rate": 8.98004e-05,
      "loss": 0.254,
      "step": 275500
    },
    {
      "epoch": 0.3112,
      "grad_norm": 0.17593859136104584,
      "learning_rate": 8.97604e-05,
      "loss": 0.2535,
      "step": 275600
    },
    {
      "epoch": 0.3114,
      "grad_norm": 0.20064404606819153,
      "learning_rate": 8.97204e-05,
      "loss": 0.2527,
      "step": 275700
    },
    {
      "epoch": 0.3116,
      "grad_norm": 0.17532220482826233,
      "learning_rate": 8.968040000000001e-05,
      "loss": 0.2533,
      "step": 275800
    },
    {
      "epoch": 0.3118,
      "grad_norm": 0.21646389365196228,
      "learning_rate": 8.96404e-05,
      "loss": 0.25,
      "step": 275900
    },
    {
      "epoch": 0.312,
      "grad_norm": 0.2135733962059021,
      "learning_rate": 8.960040000000001e-05,
      "loss": 0.2463,
      "step": 276000
    },
    {
      "epoch": 0.3122,
      "grad_norm": 0.24991312623023987,
      "learning_rate": 8.95604e-05,
      "loss": 0.2506,
      "step": 276100
    },
    {
      "epoch": 0.3124,
      "grad_norm": 0.20160618424415588,
      "learning_rate": 8.95204e-05,
      "loss": 0.2508,
      "step": 276200
    },
    {
      "epoch": 0.3126,
      "grad_norm": 0.3549211323261261,
      "learning_rate": 8.94804e-05,
      "loss": 0.2452,
      "step": 276300
    },
    {
      "epoch": 0.3128,
      "grad_norm": 0.18709681928157806,
      "learning_rate": 8.94404e-05,
      "loss": 0.2463,
      "step": 276400
    },
    {
      "epoch": 0.313,
      "grad_norm": 0.1972089558839798,
      "learning_rate": 8.94004e-05,
      "loss": 0.2434,
      "step": 276500
    },
    {
      "epoch": 0.3132,
      "grad_norm": 0.21796953678131104,
      "learning_rate": 8.93604e-05,
      "loss": 0.2577,
      "step": 276600
    },
    {
      "epoch": 0.3134,
      "grad_norm": 0.19307708740234375,
      "learning_rate": 8.93204e-05,
      "loss": 0.2485,
      "step": 276700
    },
    {
      "epoch": 0.3136,
      "grad_norm": 0.24761377274990082,
      "learning_rate": 8.928040000000001e-05,
      "loss": 0.2482,
      "step": 276800
    },
    {
      "epoch": 0.3138,
      "grad_norm": 0.1852557212114334,
      "learning_rate": 8.924040000000001e-05,
      "loss": 0.2549,
      "step": 276900
    },
    {
      "epoch": 0.314,
      "grad_norm": 0.18273688852787018,
      "learning_rate": 8.92004e-05,
      "loss": 0.2463,
      "step": 277000
    },
    {
      "epoch": 0.3142,
      "grad_norm": 0.25522246956825256,
      "learning_rate": 8.91604e-05,
      "loss": 0.2467,
      "step": 277100
    },
    {
      "epoch": 0.3144,
      "grad_norm": 0.1901063621044159,
      "learning_rate": 8.91204e-05,
      "loss": 0.2592,
      "step": 277200
    },
    {
      "epoch": 0.3146,
      "grad_norm": 0.21189343929290771,
      "learning_rate": 8.908040000000001e-05,
      "loss": 0.2549,
      "step": 277300
    },
    {
      "epoch": 0.3148,
      "grad_norm": 0.17540982365608215,
      "learning_rate": 8.90404e-05,
      "loss": 0.2534,
      "step": 277400
    },
    {
      "epoch": 0.315,
      "grad_norm": 0.1724601536989212,
      "learning_rate": 8.900040000000001e-05,
      "loss": 0.2529,
      "step": 277500
    },
    {
      "epoch": 0.3152,
      "grad_norm": 0.21971330046653748,
      "learning_rate": 8.89604e-05,
      "loss": 0.2542,
      "step": 277600
    },
    {
      "epoch": 0.3154,
      "grad_norm": 0.18612246215343475,
      "learning_rate": 8.892040000000001e-05,
      "loss": 0.249,
      "step": 277700
    },
    {
      "epoch": 0.3156,
      "grad_norm": 0.15609164535999298,
      "learning_rate": 8.88804e-05,
      "loss": 0.2486,
      "step": 277800
    },
    {
      "epoch": 0.3158,
      "grad_norm": 0.26058584451675415,
      "learning_rate": 8.88404e-05,
      "loss": 0.2489,
      "step": 277900
    },
    {
      "epoch": 0.316,
      "grad_norm": 0.2343665212392807,
      "learning_rate": 8.88004e-05,
      "loss": 0.2434,
      "step": 278000
    },
    {
      "epoch": 0.3162,
      "grad_norm": 0.26184457540512085,
      "learning_rate": 8.87604e-05,
      "loss": 0.2558,
      "step": 278100
    },
    {
      "epoch": 0.3164,
      "grad_norm": 0.23153603076934814,
      "learning_rate": 8.87204e-05,
      "loss": 0.2515,
      "step": 278200
    },
    {
      "epoch": 0.3166,
      "grad_norm": 0.16645345091819763,
      "learning_rate": 8.868040000000001e-05,
      "loss": 0.2506,
      "step": 278300
    },
    {
      "epoch": 0.3168,
      "grad_norm": 0.21093203127384186,
      "learning_rate": 8.86404e-05,
      "loss": 0.2501,
      "step": 278400
    },
    {
      "epoch": 0.317,
      "grad_norm": 0.2181226760149002,
      "learning_rate": 8.86004e-05,
      "loss": 0.2469,
      "step": 278500
    },
    {
      "epoch": 0.3172,
      "grad_norm": 0.27833208441734314,
      "learning_rate": 8.85604e-05,
      "loss": 0.2493,
      "step": 278600
    },
    {
      "epoch": 0.3174,
      "grad_norm": 0.18451841175556183,
      "learning_rate": 8.85204e-05,
      "loss": 0.2476,
      "step": 278700
    },
    {
      "epoch": 0.3176,
      "grad_norm": 0.15855015814304352,
      "learning_rate": 8.848040000000001e-05,
      "loss": 0.2562,
      "step": 278800
    },
    {
      "epoch": 0.3178,
      "grad_norm": 0.19361542165279388,
      "learning_rate": 8.84404e-05,
      "loss": 0.2461,
      "step": 278900
    },
    {
      "epoch": 0.318,
      "grad_norm": 0.1782364845275879,
      "learning_rate": 8.840040000000001e-05,
      "loss": 0.2504,
      "step": 279000
    },
    {
      "epoch": 0.3182,
      "grad_norm": 0.1721198856830597,
      "learning_rate": 8.83604e-05,
      "loss": 0.2489,
      "step": 279100
    },
    {
      "epoch": 0.3184,
      "grad_norm": 0.17030316591262817,
      "learning_rate": 8.832040000000001e-05,
      "loss": 0.2485,
      "step": 279200
    },
    {
      "epoch": 0.3186,
      "grad_norm": 0.2248735874891281,
      "learning_rate": 8.82804e-05,
      "loss": 0.252,
      "step": 279300
    },
    {
      "epoch": 0.3188,
      "grad_norm": 0.18986405432224274,
      "learning_rate": 8.82404e-05,
      "loss": 0.2536,
      "step": 279400
    },
    {
      "epoch": 0.319,
      "grad_norm": 0.48810842633247375,
      "learning_rate": 8.82004e-05,
      "loss": 0.2584,
      "step": 279500
    },
    {
      "epoch": 0.3192,
      "grad_norm": 0.16459937393665314,
      "learning_rate": 8.816040000000001e-05,
      "loss": 0.2482,
      "step": 279600
    },
    {
      "epoch": 0.3194,
      "grad_norm": 0.17048123478889465,
      "learning_rate": 8.81204e-05,
      "loss": 0.2472,
      "step": 279700
    },
    {
      "epoch": 0.3196,
      "grad_norm": 0.20386305451393127,
      "learning_rate": 8.808040000000001e-05,
      "loss": 0.2453,
      "step": 279800
    },
    {
      "epoch": 0.3198,
      "grad_norm": 0.2072535902261734,
      "learning_rate": 8.80404e-05,
      "loss": 0.2536,
      "step": 279900
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.21061697602272034,
      "learning_rate": 8.80004e-05,
      "loss": 0.25,
      "step": 280000
    },
    {
      "epoch": 0.3202,
      "grad_norm": 0.18452088534832,
      "learning_rate": 8.79604e-05,
      "loss": 0.251,
      "step": 280100
    },
    {
      "epoch": 0.3204,
      "grad_norm": 0.27870404720306396,
      "learning_rate": 8.79204e-05,
      "loss": 0.2523,
      "step": 280200
    },
    {
      "epoch": 0.3206,
      "grad_norm": 0.20520621538162231,
      "learning_rate": 8.78804e-05,
      "loss": 0.2463,
      "step": 280300
    },
    {
      "epoch": 0.3208,
      "grad_norm": 0.23470567166805267,
      "learning_rate": 8.78404e-05,
      "loss": 0.2507,
      "step": 280400
    },
    {
      "epoch": 0.321,
      "grad_norm": 0.17464961111545563,
      "learning_rate": 8.780040000000001e-05,
      "loss": 0.2463,
      "step": 280500
    },
    {
      "epoch": 0.3212,
      "grad_norm": 0.2375008910894394,
      "learning_rate": 8.776040000000001e-05,
      "loss": 0.2457,
      "step": 280600
    },
    {
      "epoch": 0.3214,
      "grad_norm": 0.17004603147506714,
      "learning_rate": 8.772040000000001e-05,
      "loss": 0.2511,
      "step": 280700
    },
    {
      "epoch": 0.3216,
      "grad_norm": 0.3254378139972687,
      "learning_rate": 8.76804e-05,
      "loss": 0.2513,
      "step": 280800
    },
    {
      "epoch": 0.3218,
      "grad_norm": 0.1874452382326126,
      "learning_rate": 8.76404e-05,
      "loss": 0.2451,
      "step": 280900
    },
    {
      "epoch": 0.322,
      "grad_norm": 0.18145489692687988,
      "learning_rate": 8.76004e-05,
      "loss": 0.2469,
      "step": 281000
    },
    {
      "epoch": 0.3222,
      "grad_norm": 0.15016548335552216,
      "learning_rate": 8.756040000000001e-05,
      "loss": 0.2496,
      "step": 281100
    },
    {
      "epoch": 0.3224,
      "grad_norm": 0.3445073962211609,
      "learning_rate": 8.75204e-05,
      "loss": 0.2484,
      "step": 281200
    },
    {
      "epoch": 0.3226,
      "grad_norm": 0.17580726742744446,
      "learning_rate": 8.748040000000001e-05,
      "loss": 0.2482,
      "step": 281300
    },
    {
      "epoch": 0.3228,
      "grad_norm": 0.17913737893104553,
      "learning_rate": 8.74404e-05,
      "loss": 0.2457,
      "step": 281400
    },
    {
      "epoch": 0.323,
      "grad_norm": 0.20740285515785217,
      "learning_rate": 8.740040000000001e-05,
      "loss": 0.2519,
      "step": 281500
    },
    {
      "epoch": 0.3232,
      "grad_norm": 0.15794052183628082,
      "learning_rate": 8.73604e-05,
      "loss": 0.2478,
      "step": 281600
    },
    {
      "epoch": 0.3234,
      "grad_norm": 0.204098641872406,
      "learning_rate": 8.73204e-05,
      "loss": 0.2478,
      "step": 281700
    },
    {
      "epoch": 0.3236,
      "grad_norm": 0.17985448241233826,
      "learning_rate": 8.72804e-05,
      "loss": 0.2511,
      "step": 281800
    },
    {
      "epoch": 0.3238,
      "grad_norm": 0.20703667402267456,
      "learning_rate": 8.72404e-05,
      "loss": 0.249,
      "step": 281900
    },
    {
      "epoch": 0.324,
      "grad_norm": 0.3906717002391815,
      "learning_rate": 8.72004e-05,
      "loss": 0.2498,
      "step": 282000
    },
    {
      "epoch": 0.3242,
      "grad_norm": 0.32392579317092896,
      "learning_rate": 8.716040000000001e-05,
      "loss": 0.2529,
      "step": 282100
    },
    {
      "epoch": 0.3244,
      "grad_norm": 0.2961254119873047,
      "learning_rate": 8.712040000000001e-05,
      "loss": 0.2528,
      "step": 282200
    },
    {
      "epoch": 0.3246,
      "grad_norm": 0.18365971744060516,
      "learning_rate": 8.70804e-05,
      "loss": 0.2508,
      "step": 282300
    },
    {
      "epoch": 0.3248,
      "grad_norm": 0.30931851267814636,
      "learning_rate": 8.70404e-05,
      "loss": 0.2557,
      "step": 282400
    },
    {
      "epoch": 0.325,
      "grad_norm": 0.18898595869541168,
      "learning_rate": 8.70004e-05,
      "loss": 0.2444,
      "step": 282500
    },
    {
      "epoch": 0.3252,
      "grad_norm": 0.27550897002220154,
      "learning_rate": 8.696040000000001e-05,
      "loss": 0.2427,
      "step": 282600
    },
    {
      "epoch": 0.3254,
      "grad_norm": 0.19707511365413666,
      "learning_rate": 8.69204e-05,
      "loss": 0.2517,
      "step": 282700
    },
    {
      "epoch": 0.3256,
      "grad_norm": 0.18959122896194458,
      "learning_rate": 8.688040000000001e-05,
      "loss": 0.2474,
      "step": 282800
    },
    {
      "epoch": 0.3258,
      "grad_norm": 0.263849675655365,
      "learning_rate": 8.68404e-05,
      "loss": 0.2497,
      "step": 282900
    },
    {
      "epoch": 0.326,
      "grad_norm": 0.17215219140052795,
      "learning_rate": 8.680040000000001e-05,
      "loss": 0.256,
      "step": 283000
    },
    {
      "epoch": 0.3262,
      "grad_norm": 0.1411217749118805,
      "learning_rate": 8.67604e-05,
      "loss": 0.2485,
      "step": 283100
    },
    {
      "epoch": 0.3264,
      "grad_norm": 0.19203592836856842,
      "learning_rate": 8.67204e-05,
      "loss": 0.2478,
      "step": 283200
    },
    {
      "epoch": 0.3266,
      "grad_norm": 0.18021759390830994,
      "learning_rate": 8.66804e-05,
      "loss": 0.2478,
      "step": 283300
    },
    {
      "epoch": 0.3268,
      "grad_norm": 0.15332086384296417,
      "learning_rate": 8.66404e-05,
      "loss": 0.2443,
      "step": 283400
    },
    {
      "epoch": 0.327,
      "grad_norm": 0.1966552585363388,
      "learning_rate": 8.66004e-05,
      "loss": 0.248,
      "step": 283500
    },
    {
      "epoch": 0.3272,
      "grad_norm": 0.21253438293933868,
      "learning_rate": 8.656040000000001e-05,
      "loss": 0.2472,
      "step": 283600
    },
    {
      "epoch": 0.3274,
      "grad_norm": 0.253312349319458,
      "learning_rate": 8.65204e-05,
      "loss": 0.2522,
      "step": 283700
    },
    {
      "epoch": 0.3276,
      "grad_norm": 0.17193368077278137,
      "learning_rate": 8.64804e-05,
      "loss": 0.2564,
      "step": 283800
    },
    {
      "epoch": 0.3278,
      "grad_norm": 0.17592443525791168,
      "learning_rate": 8.644039999999999e-05,
      "loss": 0.2484,
      "step": 283900
    },
    {
      "epoch": 0.328,
      "grad_norm": 0.200908824801445,
      "learning_rate": 8.64004e-05,
      "loss": 0.2473,
      "step": 284000
    },
    {
      "epoch": 0.3282,
      "grad_norm": 0.21171993017196655,
      "learning_rate": 8.636040000000001e-05,
      "loss": 0.2482,
      "step": 284100
    },
    {
      "epoch": 0.3284,
      "grad_norm": 0.21706272661685944,
      "learning_rate": 8.63204e-05,
      "loss": 0.2537,
      "step": 284200
    },
    {
      "epoch": 0.3286,
      "grad_norm": 0.21344873309135437,
      "learning_rate": 8.628040000000001e-05,
      "loss": 0.2501,
      "step": 284300
    },
    {
      "epoch": 0.3288,
      "grad_norm": 0.19360865652561188,
      "learning_rate": 8.62404e-05,
      "loss": 0.2499,
      "step": 284400
    },
    {
      "epoch": 0.329,
      "grad_norm": 0.16247357428073883,
      "learning_rate": 8.620040000000001e-05,
      "loss": 0.246,
      "step": 284500
    },
    {
      "epoch": 0.3292,
      "grad_norm": 0.205808624625206,
      "learning_rate": 8.61604e-05,
      "loss": 0.2592,
      "step": 284600
    },
    {
      "epoch": 0.3294,
      "grad_norm": 0.21998731791973114,
      "learning_rate": 8.61204e-05,
      "loss": 0.2512,
      "step": 284700
    },
    {
      "epoch": 0.3296,
      "grad_norm": 0.17610786855220795,
      "learning_rate": 8.60804e-05,
      "loss": 0.2474,
      "step": 284800
    },
    {
      "epoch": 0.3298,
      "grad_norm": 0.1882948875427246,
      "learning_rate": 8.60404e-05,
      "loss": 0.2512,
      "step": 284900
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.1443343162536621,
      "learning_rate": 8.60004e-05,
      "loss": 0.244,
      "step": 285000
    },
    {
      "epoch": 0.3302,
      "grad_norm": 0.22007732093334198,
      "learning_rate": 8.596040000000001e-05,
      "loss": 0.252,
      "step": 285100
    },
    {
      "epoch": 0.3304,
      "grad_norm": 0.1846906691789627,
      "learning_rate": 8.59204e-05,
      "loss": 0.2482,
      "step": 285200
    },
    {
      "epoch": 0.3306,
      "grad_norm": 0.19831278920173645,
      "learning_rate": 8.58804e-05,
      "loss": 0.2463,
      "step": 285300
    },
    {
      "epoch": 0.3308,
      "grad_norm": 0.16694426536560059,
      "learning_rate": 8.584039999999999e-05,
      "loss": 0.2461,
      "step": 285400
    },
    {
      "epoch": 0.331,
      "grad_norm": 0.22869306802749634,
      "learning_rate": 8.58004e-05,
      "loss": 0.2461,
      "step": 285500
    },
    {
      "epoch": 0.3312,
      "grad_norm": 0.29163485765457153,
      "learning_rate": 8.57604e-05,
      "loss": 0.2514,
      "step": 285600
    },
    {
      "epoch": 0.3314,
      "grad_norm": 0.20715934038162231,
      "learning_rate": 8.57204e-05,
      "loss": 0.2494,
      "step": 285700
    },
    {
      "epoch": 0.3316,
      "grad_norm": 0.2089393138885498,
      "learning_rate": 8.568040000000001e-05,
      "loss": 0.255,
      "step": 285800
    },
    {
      "epoch": 0.3318,
      "grad_norm": 0.1605835258960724,
      "learning_rate": 8.564040000000001e-05,
      "loss": 0.2502,
      "step": 285900
    },
    {
      "epoch": 0.332,
      "grad_norm": 0.21266242861747742,
      "learning_rate": 8.560040000000001e-05,
      "loss": 0.2506,
      "step": 286000
    },
    {
      "epoch": 0.3322,
      "grad_norm": 0.18077294528484344,
      "learning_rate": 8.55604e-05,
      "loss": 0.2463,
      "step": 286100
    },
    {
      "epoch": 0.3324,
      "grad_norm": 0.2691899538040161,
      "learning_rate": 8.55204e-05,
      "loss": 0.2484,
      "step": 286200
    },
    {
      "epoch": 0.3326,
      "grad_norm": 0.20911413431167603,
      "learning_rate": 8.54804e-05,
      "loss": 0.2492,
      "step": 286300
    },
    {
      "epoch": 0.3328,
      "grad_norm": 0.20955608785152435,
      "learning_rate": 8.544040000000001e-05,
      "loss": 0.2501,
      "step": 286400
    },
    {
      "epoch": 0.333,
      "grad_norm": 0.16261516511440277,
      "learning_rate": 8.54004e-05,
      "loss": 0.2481,
      "step": 286500
    },
    {
      "epoch": 0.3332,
      "grad_norm": 0.3086774945259094,
      "learning_rate": 8.536040000000001e-05,
      "loss": 0.2505,
      "step": 286600
    },
    {
      "epoch": 0.3334,
      "grad_norm": 0.16996116936206818,
      "learning_rate": 8.53204e-05,
      "loss": 0.25,
      "step": 286700
    },
    {
      "epoch": 0.3336,
      "grad_norm": 0.2675207853317261,
      "learning_rate": 8.528040000000001e-05,
      "loss": 0.2459,
      "step": 286800
    },
    {
      "epoch": 0.3338,
      "grad_norm": 0.26065778732299805,
      "learning_rate": 8.52404e-05,
      "loss": 0.2468,
      "step": 286900
    },
    {
      "epoch": 0.334,
      "grad_norm": 0.15775251388549805,
      "learning_rate": 8.52004e-05,
      "loss": 0.248,
      "step": 287000
    },
    {
      "epoch": 0.3342,
      "grad_norm": 0.17829616367816925,
      "learning_rate": 8.51604e-05,
      "loss": 0.2522,
      "step": 287100
    },
    {
      "epoch": 0.3344,
      "grad_norm": 0.283382385969162,
      "learning_rate": 8.51204e-05,
      "loss": 0.2504,
      "step": 287200
    },
    {
      "epoch": 0.3346,
      "grad_norm": 0.21750444173812866,
      "learning_rate": 8.50804e-05,
      "loss": 0.2458,
      "step": 287300
    },
    {
      "epoch": 0.3348,
      "grad_norm": 0.17604978382587433,
      "learning_rate": 8.504040000000001e-05,
      "loss": 0.2543,
      "step": 287400
    },
    {
      "epoch": 0.335,
      "grad_norm": 0.17278654873371124,
      "learning_rate": 8.500040000000001e-05,
      "loss": 0.2437,
      "step": 287500
    },
    {
      "epoch": 0.3352,
      "grad_norm": 0.1866379976272583,
      "learning_rate": 8.49604e-05,
      "loss": 0.2473,
      "step": 287600
    },
    {
      "epoch": 0.3354,
      "grad_norm": 0.18013985455036163,
      "learning_rate": 8.49204e-05,
      "loss": 0.2504,
      "step": 287700
    },
    {
      "epoch": 0.3356,
      "grad_norm": 0.1813124716281891,
      "learning_rate": 8.48804e-05,
      "loss": 0.25,
      "step": 287800
    },
    {
      "epoch": 0.3358,
      "grad_norm": 0.1880805343389511,
      "learning_rate": 8.484040000000001e-05,
      "loss": 0.2506,
      "step": 287900
    },
    {
      "epoch": 0.336,
      "grad_norm": 0.20966345071792603,
      "learning_rate": 8.48004e-05,
      "loss": 0.2508,
      "step": 288000
    },
    {
      "epoch": 0.3362,
      "grad_norm": 0.23822133243083954,
      "learning_rate": 8.476040000000001e-05,
      "loss": 0.2549,
      "step": 288100
    },
    {
      "epoch": 0.3364,
      "grad_norm": 0.17772118747234344,
      "learning_rate": 8.47204e-05,
      "loss": 0.2441,
      "step": 288200
    },
    {
      "epoch": 0.3366,
      "grad_norm": 0.22590018808841705,
      "learning_rate": 8.468040000000001e-05,
      "loss": 0.2517,
      "step": 288300
    },
    {
      "epoch": 0.3368,
      "grad_norm": 0.19911415874958038,
      "learning_rate": 8.46404e-05,
      "loss": 0.2495,
      "step": 288400
    },
    {
      "epoch": 0.337,
      "grad_norm": 0.19281023740768433,
      "learning_rate": 8.46004e-05,
      "loss": 0.2535,
      "step": 288500
    },
    {
      "epoch": 0.3372,
      "grad_norm": 0.18460847437381744,
      "learning_rate": 8.45604e-05,
      "loss": 0.2541,
      "step": 288600
    },
    {
      "epoch": 0.3374,
      "grad_norm": 0.2108038067817688,
      "learning_rate": 8.45204e-05,
      "loss": 0.252,
      "step": 288700
    },
    {
      "epoch": 0.3376,
      "grad_norm": 0.24235336482524872,
      "learning_rate": 8.44804e-05,
      "loss": 0.2482,
      "step": 288800
    },
    {
      "epoch": 0.3378,
      "grad_norm": 0.16646993160247803,
      "learning_rate": 8.444040000000001e-05,
      "loss": 0.254,
      "step": 288900
    },
    {
      "epoch": 0.338,
      "grad_norm": 0.2231166511774063,
      "learning_rate": 8.44004e-05,
      "loss": 0.2479,
      "step": 289000
    },
    {
      "epoch": 0.3382,
      "grad_norm": 0.24372661113739014,
      "learning_rate": 8.43604e-05,
      "loss": 0.2488,
      "step": 289100
    },
    {
      "epoch": 0.3384,
      "grad_norm": 0.1555611491203308,
      "learning_rate": 8.432039999999999e-05,
      "loss": 0.2532,
      "step": 289200
    },
    {
      "epoch": 0.3386,
      "grad_norm": 0.22455960512161255,
      "learning_rate": 8.42804e-05,
      "loss": 0.2474,
      "step": 289300
    },
    {
      "epoch": 0.3388,
      "grad_norm": 0.2038494348526001,
      "learning_rate": 8.424040000000001e-05,
      "loss": 0.2468,
      "step": 289400
    },
    {
      "epoch": 0.339,
      "grad_norm": 0.2508559226989746,
      "learning_rate": 8.42004e-05,
      "loss": 0.2489,
      "step": 289500
    },
    {
      "epoch": 0.3392,
      "grad_norm": 0.2195146679878235,
      "learning_rate": 8.416040000000001e-05,
      "loss": 0.2452,
      "step": 289600
    },
    {
      "epoch": 0.3394,
      "grad_norm": 0.19063441455364227,
      "learning_rate": 8.41204e-05,
      "loss": 0.2491,
      "step": 289700
    },
    {
      "epoch": 0.3396,
      "grad_norm": 0.2436915785074234,
      "learning_rate": 8.408040000000001e-05,
      "loss": 0.2508,
      "step": 289800
    },
    {
      "epoch": 0.3398,
      "grad_norm": 0.20626668632030487,
      "learning_rate": 8.40404e-05,
      "loss": 0.2497,
      "step": 289900
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.1684507131576538,
      "learning_rate": 8.40004e-05,
      "loss": 0.2493,
      "step": 290000
    },
    {
      "epoch": 0.3402,
      "grad_norm": 0.16648483276367188,
      "learning_rate": 8.39604e-05,
      "loss": 0.2526,
      "step": 290100
    },
    {
      "epoch": 0.3404,
      "grad_norm": 0.1838679313659668,
      "learning_rate": 8.39204e-05,
      "loss": 0.2477,
      "step": 290200
    },
    {
      "epoch": 0.3406,
      "grad_norm": 0.204168438911438,
      "learning_rate": 8.38804e-05,
      "loss": 0.247,
      "step": 290300
    },
    {
      "epoch": 0.3408,
      "grad_norm": 0.18529953062534332,
      "learning_rate": 8.384040000000001e-05,
      "loss": 0.25,
      "step": 290400
    },
    {
      "epoch": 0.341,
      "grad_norm": 0.17540723085403442,
      "learning_rate": 8.38004e-05,
      "loss": 0.2467,
      "step": 290500
    },
    {
      "epoch": 0.3412,
      "grad_norm": 0.15867659449577332,
      "learning_rate": 8.376040000000001e-05,
      "loss": 0.2442,
      "step": 290600
    },
    {
      "epoch": 0.3414,
      "grad_norm": 0.24484454095363617,
      "learning_rate": 8.372039999999999e-05,
      "loss": 0.2486,
      "step": 290700
    },
    {
      "epoch": 0.3416,
      "grad_norm": 0.21966548264026642,
      "learning_rate": 8.36804e-05,
      "loss": 0.2487,
      "step": 290800
    },
    {
      "epoch": 0.3418,
      "grad_norm": 0.16184256970882416,
      "learning_rate": 8.36404e-05,
      "loss": 0.2464,
      "step": 290900
    },
    {
      "epoch": 0.342,
      "grad_norm": 0.2605167329311371,
      "learning_rate": 8.36004e-05,
      "loss": 0.2456,
      "step": 291000
    },
    {
      "epoch": 0.3422,
      "grad_norm": 0.15184715390205383,
      "learning_rate": 8.356040000000001e-05,
      "loss": 0.2544,
      "step": 291100
    },
    {
      "epoch": 0.3424,
      "grad_norm": 0.18287336826324463,
      "learning_rate": 8.35204e-05,
      "loss": 0.2461,
      "step": 291200
    },
    {
      "epoch": 0.3426,
      "grad_norm": 0.1705942153930664,
      "learning_rate": 8.348040000000001e-05,
      "loss": 0.2468,
      "step": 291300
    },
    {
      "epoch": 0.3428,
      "grad_norm": 0.4577696919441223,
      "learning_rate": 8.34404e-05,
      "loss": 0.2488,
      "step": 291400
    },
    {
      "epoch": 0.343,
      "grad_norm": 0.21272404491901398,
      "learning_rate": 8.34004e-05,
      "loss": 0.2522,
      "step": 291500
    },
    {
      "epoch": 0.3432,
      "grad_norm": 0.24788586795330048,
      "learning_rate": 8.33604e-05,
      "loss": 0.2437,
      "step": 291600
    },
    {
      "epoch": 0.3434,
      "grad_norm": 0.16285791993141174,
      "learning_rate": 8.33204e-05,
      "loss": 0.2455,
      "step": 291700
    },
    {
      "epoch": 0.3436,
      "grad_norm": 0.1585230529308319,
      "learning_rate": 8.32804e-05,
      "loss": 0.2454,
      "step": 291800
    },
    {
      "epoch": 0.3438,
      "grad_norm": 0.1985248178243637,
      "learning_rate": 8.324040000000001e-05,
      "loss": 0.2475,
      "step": 291900
    },
    {
      "epoch": 0.344,
      "grad_norm": 0.3097187280654907,
      "learning_rate": 8.32004e-05,
      "loss": 0.2479,
      "step": 292000
    },
    {
      "epoch": 0.3442,
      "grad_norm": 0.21880201995372772,
      "learning_rate": 8.316040000000001e-05,
      "loss": 0.2495,
      "step": 292100
    },
    {
      "epoch": 0.3444,
      "grad_norm": 0.1829034388065338,
      "learning_rate": 8.31204e-05,
      "loss": 0.2501,
      "step": 292200
    },
    {
      "epoch": 0.3446,
      "grad_norm": 0.1782861053943634,
      "learning_rate": 8.30804e-05,
      "loss": 0.2532,
      "step": 292300
    },
    {
      "epoch": 0.3448,
      "grad_norm": 0.22069762647151947,
      "learning_rate": 8.30404e-05,
      "loss": 0.2608,
      "step": 292400
    },
    {
      "epoch": 0.345,
      "grad_norm": 0.18594609200954437,
      "learning_rate": 8.30004e-05,
      "loss": 0.2552,
      "step": 292500
    },
    {
      "epoch": 0.3452,
      "grad_norm": 0.17892573773860931,
      "learning_rate": 8.29604e-05,
      "loss": 0.2488,
      "step": 292600
    },
    {
      "epoch": 0.3454,
      "grad_norm": 0.18807783722877502,
      "learning_rate": 8.292040000000001e-05,
      "loss": 0.2513,
      "step": 292700
    },
    {
      "epoch": 0.3456,
      "grad_norm": 0.19882932305335999,
      "learning_rate": 8.288040000000001e-05,
      "loss": 0.2489,
      "step": 292800
    },
    {
      "epoch": 0.3458,
      "grad_norm": 0.19057603180408478,
      "learning_rate": 8.28404e-05,
      "loss": 0.2479,
      "step": 292900
    },
    {
      "epoch": 0.346,
      "grad_norm": 0.2387564480304718,
      "learning_rate": 8.28004e-05,
      "loss": 0.2488,
      "step": 293000
    },
    {
      "epoch": 0.3462,
      "grad_norm": 0.22695794701576233,
      "learning_rate": 8.27604e-05,
      "loss": 0.2452,
      "step": 293100
    },
    {
      "epoch": 0.3464,
      "grad_norm": 0.15103566646575928,
      "learning_rate": 8.272040000000001e-05,
      "loss": 0.2488,
      "step": 293200
    },
    {
      "epoch": 0.3466,
      "grad_norm": 0.16485248506069183,
      "learning_rate": 8.26804e-05,
      "loss": 0.2457,
      "step": 293300
    },
    {
      "epoch": 0.3468,
      "grad_norm": 0.1788826584815979,
      "learning_rate": 8.264040000000001e-05,
      "loss": 0.2479,
      "step": 293400
    },
    {
      "epoch": 0.347,
      "grad_norm": 0.17442439496517181,
      "learning_rate": 8.26004e-05,
      "loss": 0.2471,
      "step": 293500
    },
    {
      "epoch": 0.3472,
      "grad_norm": 0.5051277875900269,
      "learning_rate": 8.256040000000001e-05,
      "loss": 0.2474,
      "step": 293600
    },
    {
      "epoch": 0.3474,
      "grad_norm": 0.40286529064178467,
      "learning_rate": 8.25204e-05,
      "loss": 0.2482,
      "step": 293700
    },
    {
      "epoch": 0.3476,
      "grad_norm": 0.2842089831829071,
      "learning_rate": 8.24804e-05,
      "loss": 0.2494,
      "step": 293800
    },
    {
      "epoch": 0.3478,
      "grad_norm": 0.17578501999378204,
      "learning_rate": 8.24404e-05,
      "loss": 0.2527,
      "step": 293900
    },
    {
      "epoch": 0.348,
      "grad_norm": 0.17574481666088104,
      "learning_rate": 8.24004e-05,
      "loss": 0.2463,
      "step": 294000
    },
    {
      "epoch": 0.3482,
      "grad_norm": 0.1960979551076889,
      "learning_rate": 8.23604e-05,
      "loss": 0.2495,
      "step": 294100
    },
    {
      "epoch": 0.3484,
      "grad_norm": 0.1842304915189743,
      "learning_rate": 8.232040000000001e-05,
      "loss": 0.25,
      "step": 294200
    },
    {
      "epoch": 0.3486,
      "grad_norm": 0.17555032670497894,
      "learning_rate": 8.22804e-05,
      "loss": 0.2487,
      "step": 294300
    },
    {
      "epoch": 0.3488,
      "grad_norm": 0.18717283010482788,
      "learning_rate": 8.224040000000001e-05,
      "loss": 0.2462,
      "step": 294400
    },
    {
      "epoch": 0.349,
      "grad_norm": 0.18235217034816742,
      "learning_rate": 8.220039999999999e-05,
      "loss": 0.2472,
      "step": 294500
    },
    {
      "epoch": 0.3492,
      "grad_norm": 0.19905789196491241,
      "learning_rate": 8.21604e-05,
      "loss": 0.2485,
      "step": 294600
    },
    {
      "epoch": 0.3494,
      "grad_norm": 0.19601567089557648,
      "learning_rate": 8.212040000000001e-05,
      "loss": 0.2519,
      "step": 294700
    },
    {
      "epoch": 0.3496,
      "grad_norm": 0.16348402202129364,
      "learning_rate": 8.20804e-05,
      "loss": 0.2463,
      "step": 294800
    },
    {
      "epoch": 0.3498,
      "grad_norm": 0.16451190412044525,
      "learning_rate": 8.204040000000001e-05,
      "loss": 0.251,
      "step": 294900
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.20977216958999634,
      "learning_rate": 8.20004e-05,
      "loss": 0.2465,
      "step": 295000
    },
    {
      "epoch": 0.3502,
      "grad_norm": 0.3382442891597748,
      "learning_rate": 8.196040000000001e-05,
      "loss": 0.2456,
      "step": 295100
    },
    {
      "epoch": 0.3504,
      "grad_norm": 0.22302016615867615,
      "learning_rate": 8.19204e-05,
      "loss": 0.2469,
      "step": 295200
    },
    {
      "epoch": 0.3506,
      "grad_norm": 0.1620716005563736,
      "learning_rate": 8.18804e-05,
      "loss": 0.2435,
      "step": 295300
    },
    {
      "epoch": 0.3508,
      "grad_norm": 0.2126971334218979,
      "learning_rate": 8.18404e-05,
      "loss": 0.2444,
      "step": 295400
    },
    {
      "epoch": 0.351,
      "grad_norm": 0.1620679646730423,
      "learning_rate": 8.18004e-05,
      "loss": 0.2402,
      "step": 295500
    },
    {
      "epoch": 0.3512,
      "grad_norm": 0.19541557133197784,
      "learning_rate": 8.17604e-05,
      "loss": 0.2426,
      "step": 295600
    },
    {
      "epoch": 0.3514,
      "grad_norm": 0.18142946064472198,
      "learning_rate": 8.172040000000001e-05,
      "loss": 0.2452,
      "step": 295700
    },
    {
      "epoch": 0.3516,
      "grad_norm": 0.18419720232486725,
      "learning_rate": 8.16804e-05,
      "loss": 0.2466,
      "step": 295800
    },
    {
      "epoch": 0.3518,
      "grad_norm": 0.19173642992973328,
      "learning_rate": 8.164040000000001e-05,
      "loss": 0.2487,
      "step": 295900
    },
    {
      "epoch": 0.352,
      "grad_norm": 0.16842658817768097,
      "learning_rate": 8.16004e-05,
      "loss": 0.2508,
      "step": 296000
    },
    {
      "epoch": 0.3522,
      "grad_norm": 0.23482634127140045,
      "learning_rate": 8.15604e-05,
      "loss": 0.2421,
      "step": 296100
    },
    {
      "epoch": 0.3524,
      "grad_norm": 0.17169761657714844,
      "learning_rate": 8.15204e-05,
      "loss": 0.2454,
      "step": 296200
    },
    {
      "epoch": 0.3526,
      "grad_norm": 0.1711745262145996,
      "learning_rate": 8.14804e-05,
      "loss": 0.2424,
      "step": 296300
    },
    {
      "epoch": 0.3528,
      "grad_norm": 0.18024346232414246,
      "learning_rate": 8.144040000000001e-05,
      "loss": 0.244,
      "step": 296400
    },
    {
      "epoch": 0.353,
      "grad_norm": 0.468720018863678,
      "learning_rate": 8.14004e-05,
      "loss": 0.2452,
      "step": 296500
    },
    {
      "epoch": 0.3532,
      "grad_norm": 0.16344809532165527,
      "learning_rate": 8.136040000000001e-05,
      "loss": 0.2426,
      "step": 296600
    },
    {
      "epoch": 0.3534,
      "grad_norm": 0.22236470878124237,
      "learning_rate": 8.13204e-05,
      "loss": 0.2466,
      "step": 296700
    },
    {
      "epoch": 0.3536,
      "grad_norm": 0.19982776045799255,
      "learning_rate": 8.12804e-05,
      "loss": 0.2506,
      "step": 296800
    },
    {
      "epoch": 0.3538,
      "grad_norm": 0.1859310418367386,
      "learning_rate": 8.12404e-05,
      "loss": 0.251,
      "step": 296900
    },
    {
      "epoch": 0.354,
      "grad_norm": 0.21040932834148407,
      "learning_rate": 8.12004e-05,
      "loss": 0.2457,
      "step": 297000
    },
    {
      "epoch": 0.3542,
      "grad_norm": 0.1694370061159134,
      "learning_rate": 8.11604e-05,
      "loss": 0.2444,
      "step": 297100
    },
    {
      "epoch": 0.3544,
      "grad_norm": 0.17156536877155304,
      "learning_rate": 8.112040000000001e-05,
      "loss": 0.2424,
      "step": 297200
    },
    {
      "epoch": 0.3546,
      "grad_norm": 0.17718739807605743,
      "learning_rate": 8.10804e-05,
      "loss": 0.2411,
      "step": 297300
    },
    {
      "epoch": 0.3548,
      "grad_norm": 0.23720037937164307,
      "learning_rate": 8.104040000000001e-05,
      "loss": 0.2469,
      "step": 297400
    },
    {
      "epoch": 0.355,
      "grad_norm": 0.2064405083656311,
      "learning_rate": 8.10004e-05,
      "loss": 0.2415,
      "step": 297500
    },
    {
      "epoch": 0.3552,
      "grad_norm": 0.1649145632982254,
      "learning_rate": 8.09604e-05,
      "loss": 0.2421,
      "step": 297600
    },
    {
      "epoch": 0.3554,
      "grad_norm": 0.1422872543334961,
      "learning_rate": 8.09204e-05,
      "loss": 0.2414,
      "step": 297700
    },
    {
      "epoch": 0.3556,
      "grad_norm": 0.42385339736938477,
      "learning_rate": 8.08804e-05,
      "loss": 0.2425,
      "step": 297800
    },
    {
      "epoch": 0.3558,
      "grad_norm": 0.1627955287694931,
      "learning_rate": 8.08404e-05,
      "loss": 0.2481,
      "step": 297900
    },
    {
      "epoch": 0.356,
      "grad_norm": 0.1613747775554657,
      "learning_rate": 8.08004e-05,
      "loss": 0.2478,
      "step": 298000
    },
    {
      "epoch": 0.3562,
      "grad_norm": 0.23526862263679504,
      "learning_rate": 8.076040000000001e-05,
      "loss": 0.248,
      "step": 298100
    },
    {
      "epoch": 0.3564,
      "grad_norm": 0.18639631569385529,
      "learning_rate": 8.07204e-05,
      "loss": 0.2567,
      "step": 298200
    },
    {
      "epoch": 0.3566,
      "grad_norm": 0.2338380515575409,
      "learning_rate": 8.06804e-05,
      "loss": 0.2494,
      "step": 298300
    },
    {
      "epoch": 0.3568,
      "grad_norm": 0.165755957365036,
      "learning_rate": 8.06404e-05,
      "loss": 0.2518,
      "step": 298400
    },
    {
      "epoch": 0.357,
      "grad_norm": 0.1789596974849701,
      "learning_rate": 8.060040000000001e-05,
      "loss": 0.2483,
      "step": 298500
    },
    {
      "epoch": 0.3572,
      "grad_norm": 0.174707293510437,
      "learning_rate": 8.05604e-05,
      "loss": 0.2445,
      "step": 298600
    },
    {
      "epoch": 0.3574,
      "grad_norm": 0.15874457359313965,
      "learning_rate": 8.052040000000001e-05,
      "loss": 0.2526,
      "step": 298700
    },
    {
      "epoch": 0.3576,
      "grad_norm": 0.15661177039146423,
      "learning_rate": 8.04804e-05,
      "loss": 0.2447,
      "step": 298800
    },
    {
      "epoch": 0.3578,
      "grad_norm": 0.185741126537323,
      "learning_rate": 8.044040000000001e-05,
      "loss": 0.254,
      "step": 298900
    },
    {
      "epoch": 0.358,
      "grad_norm": 0.1975129246711731,
      "learning_rate": 8.04004e-05,
      "loss": 0.2453,
      "step": 299000
    },
    {
      "epoch": 0.3582,
      "grad_norm": 0.21340824663639069,
      "learning_rate": 8.03604e-05,
      "loss": 0.2436,
      "step": 299100
    },
    {
      "epoch": 0.3584,
      "grad_norm": 0.1729285567998886,
      "learning_rate": 8.03204e-05,
      "loss": 0.2495,
      "step": 299200
    },
    {
      "epoch": 0.3586,
      "grad_norm": 0.2982395589351654,
      "learning_rate": 8.02804e-05,
      "loss": 0.2443,
      "step": 299300
    },
    {
      "epoch": 0.3588,
      "grad_norm": 0.18718211352825165,
      "learning_rate": 8.02404e-05,
      "loss": 0.2421,
      "step": 299400
    },
    {
      "epoch": 0.359,
      "grad_norm": 0.18755905330181122,
      "learning_rate": 8.020040000000001e-05,
      "loss": 0.2517,
      "step": 299500
    },
    {
      "epoch": 0.3592,
      "grad_norm": 0.15775266289710999,
      "learning_rate": 8.01604e-05,
      "loss": 0.2471,
      "step": 299600
    },
    {
      "epoch": 0.3594,
      "grad_norm": 0.14446353912353516,
      "learning_rate": 8.012040000000001e-05,
      "loss": 0.2433,
      "step": 299700
    },
    {
      "epoch": 0.3596,
      "grad_norm": 0.17306815087795258,
      "learning_rate": 8.008039999999999e-05,
      "loss": 0.2464,
      "step": 299800
    },
    {
      "epoch": 0.3598,
      "grad_norm": 0.22161149978637695,
      "learning_rate": 8.00404e-05,
      "loss": 0.2467,
      "step": 299900
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.15656442940235138,
      "learning_rate": 8.000040000000001e-05,
      "loss": 0.2445,
      "step": 300000
    },
    {
      "epoch": 0.3602,
      "grad_norm": 0.17966248095035553,
      "learning_rate": 7.99604e-05,
      "loss": 0.2451,
      "step": 300100
    },
    {
      "epoch": 0.3604,
      "grad_norm": 0.18749384582042694,
      "learning_rate": 7.992040000000001e-05,
      "loss": 0.242,
      "step": 300200
    },
    {
      "epoch": 0.3606,
      "grad_norm": 0.18940359354019165,
      "learning_rate": 7.98804e-05,
      "loss": 0.241,
      "step": 300300
    },
    {
      "epoch": 0.3608,
      "grad_norm": 0.1695726066827774,
      "learning_rate": 7.984040000000001e-05,
      "loss": 0.247,
      "step": 300400
    },
    {
      "epoch": 0.361,
      "grad_norm": 0.23196519911289215,
      "learning_rate": 7.98004e-05,
      "loss": 0.2518,
      "step": 300500
    },
    {
      "epoch": 0.3612,
      "grad_norm": 0.1548641175031662,
      "learning_rate": 7.97604e-05,
      "loss": 0.2424,
      "step": 300600
    },
    {
      "epoch": 0.3614,
      "grad_norm": 0.25332915782928467,
      "learning_rate": 7.97204e-05,
      "loss": 0.2428,
      "step": 300700
    },
    {
      "epoch": 0.3616,
      "grad_norm": 0.21986594796180725,
      "learning_rate": 7.96804e-05,
      "loss": 0.2447,
      "step": 300800
    },
    {
      "epoch": 0.3618,
      "grad_norm": 0.18515200912952423,
      "learning_rate": 7.96404e-05,
      "loss": 0.2424,
      "step": 300900
    },
    {
      "epoch": 0.362,
      "grad_norm": 0.23712636530399323,
      "learning_rate": 7.960040000000001e-05,
      "loss": 0.2513,
      "step": 301000
    },
    {
      "epoch": 0.3622,
      "grad_norm": 0.21687628328800201,
      "learning_rate": 7.95604e-05,
      "loss": 0.249,
      "step": 301100
    },
    {
      "epoch": 0.3624,
      "grad_norm": 0.22492177784442902,
      "learning_rate": 7.952040000000001e-05,
      "loss": 0.2463,
      "step": 301200
    },
    {
      "epoch": 0.3626,
      "grad_norm": 0.19002969563007355,
      "learning_rate": 7.94804e-05,
      "loss": 0.2455,
      "step": 301300
    },
    {
      "epoch": 0.3628,
      "grad_norm": 0.1988956481218338,
      "learning_rate": 7.94404e-05,
      "loss": 0.2405,
      "step": 301400
    },
    {
      "epoch": 0.363,
      "grad_norm": 0.1998603492975235,
      "learning_rate": 7.94004e-05,
      "loss": 0.2454,
      "step": 301500
    },
    {
      "epoch": 0.3632,
      "grad_norm": 0.23634324967861176,
      "learning_rate": 7.93604e-05,
      "loss": 0.2481,
      "step": 301600
    },
    {
      "epoch": 0.3634,
      "grad_norm": 0.17416614294052124,
      "learning_rate": 7.932040000000001e-05,
      "loss": 0.2418,
      "step": 301700
    },
    {
      "epoch": 0.3636,
      "grad_norm": 0.2007831186056137,
      "learning_rate": 7.92804e-05,
      "loss": 0.2538,
      "step": 301800
    },
    {
      "epoch": 0.3638,
      "grad_norm": 0.1813400238752365,
      "learning_rate": 7.924040000000001e-05,
      "loss": 0.2553,
      "step": 301900
    },
    {
      "epoch": 0.364,
      "grad_norm": 0.18407954275608063,
      "learning_rate": 7.92004e-05,
      "loss": 0.2429,
      "step": 302000
    },
    {
      "epoch": 0.3642,
      "grad_norm": 0.2030446082353592,
      "learning_rate": 7.91604e-05,
      "loss": 0.243,
      "step": 302100
    },
    {
      "epoch": 0.3644,
      "grad_norm": 0.18323349952697754,
      "learning_rate": 7.91204e-05,
      "loss": 0.2449,
      "step": 302200
    },
    {
      "epoch": 0.3646,
      "grad_norm": 0.1829667091369629,
      "learning_rate": 7.90804e-05,
      "loss": 0.2497,
      "step": 302300
    },
    {
      "epoch": 0.3648,
      "grad_norm": 0.21287867426872253,
      "learning_rate": 7.90404e-05,
      "loss": 0.2463,
      "step": 302400
    },
    {
      "epoch": 0.365,
      "grad_norm": 0.17376112937927246,
      "learning_rate": 7.900040000000001e-05,
      "loss": 0.241,
      "step": 302500
    },
    {
      "epoch": 0.3652,
      "grad_norm": 0.19770017266273499,
      "learning_rate": 7.89604e-05,
      "loss": 0.247,
      "step": 302600
    },
    {
      "epoch": 0.3654,
      "grad_norm": 0.2480800300836563,
      "learning_rate": 7.892040000000001e-05,
      "loss": 0.2408,
      "step": 302700
    },
    {
      "epoch": 0.3656,
      "grad_norm": 0.1723504662513733,
      "learning_rate": 7.88804e-05,
      "loss": 0.2475,
      "step": 302800
    },
    {
      "epoch": 0.3658,
      "grad_norm": 0.19746597111225128,
      "learning_rate": 7.88404e-05,
      "loss": 0.2387,
      "step": 302900
    },
    {
      "epoch": 0.366,
      "grad_norm": 0.2700865864753723,
      "learning_rate": 7.88004e-05,
      "loss": 0.2431,
      "step": 303000
    },
    {
      "epoch": 0.3662,
      "grad_norm": 0.18088355660438538,
      "learning_rate": 7.87604e-05,
      "loss": 0.2425,
      "step": 303100
    },
    {
      "epoch": 0.3664,
      "grad_norm": 0.24252882599830627,
      "learning_rate": 7.87204e-05,
      "loss": 0.2456,
      "step": 303200
    },
    {
      "epoch": 0.3666,
      "grad_norm": 0.18904949724674225,
      "learning_rate": 7.86804e-05,
      "loss": 0.2485,
      "step": 303300
    },
    {
      "epoch": 0.3668,
      "grad_norm": 0.2009989470243454,
      "learning_rate": 7.864040000000001e-05,
      "loss": 0.2457,
      "step": 303400
    },
    {
      "epoch": 0.367,
      "grad_norm": 0.21063704788684845,
      "learning_rate": 7.860040000000001e-05,
      "loss": 0.2437,
      "step": 303500
    },
    {
      "epoch": 0.3672,
      "grad_norm": 0.24128037691116333,
      "learning_rate": 7.85604e-05,
      "loss": 0.2519,
      "step": 303600
    },
    {
      "epoch": 0.3674,
      "grad_norm": 0.18868161737918854,
      "learning_rate": 7.85204e-05,
      "loss": 0.2456,
      "step": 303700
    },
    {
      "epoch": 0.3676,
      "grad_norm": 0.24962688982486725,
      "learning_rate": 7.84804e-05,
      "loss": 0.2483,
      "step": 303800
    },
    {
      "epoch": 0.3678,
      "grad_norm": 0.15526580810546875,
      "learning_rate": 7.84404e-05,
      "loss": 0.2507,
      "step": 303900
    },
    {
      "epoch": 0.368,
      "grad_norm": 0.21151374280452728,
      "learning_rate": 7.840040000000001e-05,
      "loss": 0.2587,
      "step": 304000
    },
    {
      "epoch": 0.3682,
      "grad_norm": 0.18684406578540802,
      "learning_rate": 7.83604e-05,
      "loss": 0.2472,
      "step": 304100
    },
    {
      "epoch": 0.3684,
      "grad_norm": 0.5539630651473999,
      "learning_rate": 7.832040000000001e-05,
      "loss": 0.2514,
      "step": 304200
    },
    {
      "epoch": 0.3686,
      "grad_norm": 0.31068259477615356,
      "learning_rate": 7.82804e-05,
      "loss": 0.2521,
      "step": 304300
    },
    {
      "epoch": 0.3688,
      "grad_norm": 0.15399712324142456,
      "learning_rate": 7.82404e-05,
      "loss": 0.245,
      "step": 304400
    },
    {
      "epoch": 0.369,
      "grad_norm": 0.16748294234275818,
      "learning_rate": 7.82004e-05,
      "loss": 0.2458,
      "step": 304500
    },
    {
      "epoch": 0.3692,
      "grad_norm": 0.35579004883766174,
      "learning_rate": 7.81604e-05,
      "loss": 0.2753,
      "step": 304600
    },
    {
      "epoch": 0.3694,
      "grad_norm": 0.28640273213386536,
      "learning_rate": 7.81204e-05,
      "loss": 0.3173,
      "step": 304700
    },
    {
      "epoch": 0.3696,
      "grad_norm": 0.5388140082359314,
      "learning_rate": 7.808040000000001e-05,
      "loss": 0.3084,
      "step": 304800
    },
    {
      "epoch": 0.3698,
      "grad_norm": 0.28529220819473267,
      "learning_rate": 7.80404e-05,
      "loss": 0.2922,
      "step": 304900
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.465318888425827,
      "learning_rate": 7.800040000000001e-05,
      "loss": 0.3043,
      "step": 305000
    },
    {
      "epoch": 0.3702,
      "grad_norm": 0.25032395124435425,
      "learning_rate": 7.79604e-05,
      "loss": 0.3063,
      "step": 305100
    },
    {
      "epoch": 0.3704,
      "grad_norm": 0.4543182849884033,
      "learning_rate": 7.79204e-05,
      "loss": 0.2866,
      "step": 305200
    },
    {
      "epoch": 0.3706,
      "grad_norm": 0.305585652589798,
      "learning_rate": 7.788040000000001e-05,
      "loss": 0.2914,
      "step": 305300
    },
    {
      "epoch": 0.3708,
      "grad_norm": 0.9315662980079651,
      "learning_rate": 7.78404e-05,
      "loss": 0.2972,
      "step": 305400
    },
    {
      "epoch": 0.371,
      "grad_norm": 0.32617905735969543,
      "learning_rate": 7.780040000000001e-05,
      "loss": 0.3116,
      "step": 305500
    },
    {
      "epoch": 0.3712,
      "grad_norm": 0.3560982346534729,
      "learning_rate": 7.77604e-05,
      "loss": 0.3046,
      "step": 305600
    },
    {
      "epoch": 0.3714,
      "grad_norm": 0.23608624935150146,
      "learning_rate": 7.772040000000001e-05,
      "loss": 0.2764,
      "step": 305700
    },
    {
      "epoch": 0.3716,
      "grad_norm": 0.429745614528656,
      "learning_rate": 7.76804e-05,
      "loss": 0.3084,
      "step": 305800
    },
    {
      "epoch": 0.3718,
      "grad_norm": 0.181761234998703,
      "learning_rate": 7.76404e-05,
      "loss": 0.2848,
      "step": 305900
    },
    {
      "epoch": 0.372,
      "grad_norm": 0.2306230366230011,
      "learning_rate": 7.76004e-05,
      "loss": 0.2956,
      "step": 306000
    },
    {
      "epoch": 0.3722,
      "grad_norm": 0.22008122503757477,
      "learning_rate": 7.75604e-05,
      "loss": 0.2669,
      "step": 306100
    },
    {
      "epoch": 0.3724,
      "grad_norm": 0.1795935481786728,
      "learning_rate": 7.75204e-05,
      "loss": 0.2864,
      "step": 306200
    },
    {
      "epoch": 0.3726,
      "grad_norm": 0.28549113869667053,
      "learning_rate": 7.748040000000001e-05,
      "loss": 0.2774,
      "step": 306300
    },
    {
      "epoch": 0.3728,
      "grad_norm": 0.2546445429325104,
      "learning_rate": 7.74404e-05,
      "loss": 0.2714,
      "step": 306400
    },
    {
      "epoch": 0.373,
      "grad_norm": 0.23946426808834076,
      "learning_rate": 7.740040000000001e-05,
      "loss": 0.271,
      "step": 306500
    },
    {
      "epoch": 0.3732,
      "grad_norm": 0.4045504927635193,
      "learning_rate": 7.73604e-05,
      "loss": 0.3021,
      "step": 306600
    },
    {
      "epoch": 0.3734,
      "grad_norm": 0.32764193415641785,
      "learning_rate": 7.73204e-05,
      "loss": 0.2769,
      "step": 306700
    },
    {
      "epoch": 0.3736,
      "grad_norm": 0.2573786973953247,
      "learning_rate": 7.72804e-05,
      "loss": 0.2901,
      "step": 306800
    },
    {
      "epoch": 0.3738,
      "grad_norm": 0.34804463386535645,
      "learning_rate": 7.72404e-05,
      "loss": 0.2925,
      "step": 306900
    },
    {
      "epoch": 0.374,
      "grad_norm": 0.3770942986011505,
      "learning_rate": 7.720040000000001e-05,
      "loss": 0.2738,
      "step": 307000
    },
    {
      "epoch": 0.3742,
      "grad_norm": 0.27824512124061584,
      "learning_rate": 7.71604e-05,
      "loss": 0.28,
      "step": 307100
    },
    {
      "epoch": 0.3744,
      "grad_norm": 0.24019598960876465,
      "learning_rate": 7.712040000000001e-05,
      "loss": 0.2983,
      "step": 307200
    },
    {
      "epoch": 0.3746,
      "grad_norm": 0.283953458070755,
      "learning_rate": 7.70804e-05,
      "loss": 0.293,
      "step": 307300
    },
    {
      "epoch": 0.3748,
      "grad_norm": 0.2683544456958771,
      "learning_rate": 7.70404e-05,
      "loss": 0.2729,
      "step": 307400
    },
    {
      "epoch": 0.375,
      "grad_norm": 0.3089542090892792,
      "learning_rate": 7.70004e-05,
      "loss": 0.2916,
      "step": 307500
    },
    {
      "epoch": 0.3752,
      "grad_norm": 0.5444771647453308,
      "learning_rate": 7.69604e-05,
      "loss": 0.2795,
      "step": 307600
    },
    {
      "epoch": 0.3754,
      "grad_norm": 0.333789587020874,
      "learning_rate": 7.69204e-05,
      "loss": 0.2751,
      "step": 307700
    },
    {
      "epoch": 0.3756,
      "grad_norm": 0.7448511123657227,
      "learning_rate": 7.688040000000001e-05,
      "loss": 0.29,
      "step": 307800
    },
    {
      "epoch": 0.3758,
      "grad_norm": 0.17118731141090393,
      "learning_rate": 7.68404e-05,
      "loss": 0.2819,
      "step": 307900
    },
    {
      "epoch": 0.376,
      "grad_norm": 0.1759895384311676,
      "learning_rate": 7.680040000000001e-05,
      "loss": 0.2877,
      "step": 308000
    },
    {
      "epoch": 0.3762,
      "grad_norm": 0.26918289065361023,
      "learning_rate": 7.67604e-05,
      "loss": 0.2738,
      "step": 308100
    },
    {
      "epoch": 0.3764,
      "grad_norm": 0.2078232765197754,
      "learning_rate": 7.67204e-05,
      "loss": 0.2881,
      "step": 308200
    },
    {
      "epoch": 0.3766,
      "grad_norm": 0.6741946339607239,
      "learning_rate": 7.66804e-05,
      "loss": 0.2742,
      "step": 308300
    },
    {
      "epoch": 0.3768,
      "grad_norm": 0.2831641137599945,
      "learning_rate": 7.66404e-05,
      "loss": 0.3325,
      "step": 308400
    },
    {
      "epoch": 0.377,
      "grad_norm": 0.16423487663269043,
      "learning_rate": 7.66004e-05,
      "loss": 0.2887,
      "step": 308500
    },
    {
      "epoch": 0.3772,
      "grad_norm": 0.2626672089099884,
      "learning_rate": 7.65604e-05,
      "loss": 0.2761,
      "step": 308600
    },
    {
      "epoch": 0.3774,
      "grad_norm": 0.23534747958183289,
      "learning_rate": 7.652040000000001e-05,
      "loss": 0.2843,
      "step": 308700
    },
    {
      "epoch": 0.3776,
      "grad_norm": 0.4902610182762146,
      "learning_rate": 7.648040000000001e-05,
      "loss": 0.2959,
      "step": 308800
    },
    {
      "epoch": 0.3778,
      "grad_norm": 0.29607897996902466,
      "learning_rate": 7.644040000000001e-05,
      "loss": 0.2678,
      "step": 308900
    },
    {
      "epoch": 0.378,
      "grad_norm": 0.3284173011779785,
      "learning_rate": 7.64004e-05,
      "loss": 0.2764,
      "step": 309000
    },
    {
      "epoch": 0.3782,
      "grad_norm": 0.24199332296848297,
      "learning_rate": 7.63604e-05,
      "loss": 0.2766,
      "step": 309100
    },
    {
      "epoch": 0.3784,
      "grad_norm": 0.3074169158935547,
      "learning_rate": 7.63204e-05,
      "loss": 0.2901,
      "step": 309200
    },
    {
      "epoch": 0.3786,
      "grad_norm": 0.3491748869419098,
      "learning_rate": 7.628040000000001e-05,
      "loss": 0.2846,
      "step": 309300
    },
    {
      "epoch": 0.3788,
      "grad_norm": 0.46956080198287964,
      "learning_rate": 7.62404e-05,
      "loss": 0.2853,
      "step": 309400
    },
    {
      "epoch": 0.379,
      "grad_norm": 0.2703664004802704,
      "learning_rate": 7.620040000000001e-05,
      "loss": 0.3082,
      "step": 309500
    },
    {
      "epoch": 0.3792,
      "grad_norm": 0.25405412912368774,
      "learning_rate": 7.61604e-05,
      "loss": 0.2806,
      "step": 309600
    },
    {
      "epoch": 0.3794,
      "grad_norm": 0.39644813537597656,
      "learning_rate": 7.61204e-05,
      "loss": 0.2921,
      "step": 309700
    },
    {
      "epoch": 0.3796,
      "grad_norm": 0.16149291396141052,
      "learning_rate": 7.60804e-05,
      "loss": 0.2622,
      "step": 309800
    },
    {
      "epoch": 0.3798,
      "grad_norm": 0.314878910779953,
      "learning_rate": 7.60404e-05,
      "loss": 0.2703,
      "step": 309900
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.2966255247592926,
      "learning_rate": 7.60004e-05,
      "loss": 0.2854,
      "step": 310000
    },
    {
      "epoch": 0.3802,
      "grad_norm": 0.20654204487800598,
      "learning_rate": 7.59604e-05,
      "loss": 0.29,
      "step": 310100
    },
    {
      "epoch": 0.3804,
      "grad_norm": 0.2614297568798065,
      "learning_rate": 7.59204e-05,
      "loss": 0.2918,
      "step": 310200
    },
    {
      "epoch": 0.3806,
      "grad_norm": 0.22196181118488312,
      "learning_rate": 7.588040000000001e-05,
      "loss": 0.2773,
      "step": 310300
    },
    {
      "epoch": 0.3808,
      "grad_norm": 0.19726967811584473,
      "learning_rate": 7.58404e-05,
      "loss": 0.2817,
      "step": 310400
    },
    {
      "epoch": 0.381,
      "grad_norm": 0.24383938312530518,
      "learning_rate": 7.58004e-05,
      "loss": 0.2837,
      "step": 310500
    },
    {
      "epoch": 0.3812,
      "grad_norm": 0.21394483745098114,
      "learning_rate": 7.576040000000001e-05,
      "loss": 0.2759,
      "step": 310600
    },
    {
      "epoch": 0.3814,
      "grad_norm": 0.2858792245388031,
      "learning_rate": 7.57204e-05,
      "loss": 0.291,
      "step": 310700
    },
    {
      "epoch": 0.3816,
      "grad_norm": 0.19695499539375305,
      "learning_rate": 7.568040000000001e-05,
      "loss": 0.2819,
      "step": 310800
    },
    {
      "epoch": 0.3818,
      "grad_norm": 0.21151651442050934,
      "learning_rate": 7.56404e-05,
      "loss": 0.2702,
      "step": 310900
    },
    {
      "epoch": 0.382,
      "grad_norm": 0.22047953307628632,
      "learning_rate": 7.560040000000001e-05,
      "loss": 0.2866,
      "step": 311000
    },
    {
      "epoch": 0.3822,
      "grad_norm": 0.5067945718765259,
      "learning_rate": 7.55604e-05,
      "loss": 0.2756,
      "step": 311100
    },
    {
      "epoch": 0.3824,
      "grad_norm": 0.43224334716796875,
      "learning_rate": 7.55204e-05,
      "loss": 0.288,
      "step": 311200
    },
    {
      "epoch": 0.3826,
      "grad_norm": 0.2836945652961731,
      "learning_rate": 7.54804e-05,
      "loss": 0.2803,
      "step": 311300
    },
    {
      "epoch": 0.3828,
      "grad_norm": 0.6460776925086975,
      "learning_rate": 7.54404e-05,
      "loss": 0.2686,
      "step": 311400
    },
    {
      "epoch": 0.383,
      "grad_norm": 0.3450317978858948,
      "learning_rate": 7.54004e-05,
      "loss": 0.2747,
      "step": 311500
    },
    {
      "epoch": 0.3832,
      "grad_norm": 0.2429281622171402,
      "learning_rate": 7.536040000000001e-05,
      "loss": 0.2859,
      "step": 311600
    },
    {
      "epoch": 0.3834,
      "grad_norm": 0.1564161479473114,
      "learning_rate": 7.53204e-05,
      "loss": 0.282,
      "step": 311700
    },
    {
      "epoch": 0.3836,
      "grad_norm": 0.242931067943573,
      "learning_rate": 7.528040000000001e-05,
      "loss": 0.2638,
      "step": 311800
    },
    {
      "epoch": 0.3838,
      "grad_norm": 0.22181962430477142,
      "learning_rate": 7.52404e-05,
      "loss": 0.2862,
      "step": 311900
    },
    {
      "epoch": 0.384,
      "grad_norm": 0.23662132024765015,
      "learning_rate": 7.52004e-05,
      "loss": 0.282,
      "step": 312000
    },
    {
      "epoch": 0.3842,
      "grad_norm": 0.25929373502731323,
      "learning_rate": 7.51604e-05,
      "loss": 0.2827,
      "step": 312100
    },
    {
      "epoch": 0.3844,
      "grad_norm": 0.3989902436733246,
      "learning_rate": 7.51204e-05,
      "loss": 0.2733,
      "step": 312200
    },
    {
      "epoch": 0.3846,
      "grad_norm": 0.26393795013427734,
      "learning_rate": 7.508040000000001e-05,
      "loss": 0.2827,
      "step": 312300
    },
    {
      "epoch": 0.3848,
      "grad_norm": 0.6005264520645142,
      "learning_rate": 7.50404e-05,
      "loss": 0.2917,
      "step": 312400
    },
    {
      "epoch": 0.385,
      "grad_norm": 0.23418720066547394,
      "learning_rate": 7.500040000000001e-05,
      "loss": 0.2657,
      "step": 312500
    },
    {
      "epoch": 0.3852,
      "grad_norm": 0.1971062570810318,
      "learning_rate": 7.496040000000002e-05,
      "loss": 0.2741,
      "step": 312600
    },
    {
      "epoch": 0.3854,
      "grad_norm": 0.9465832710266113,
      "learning_rate": 7.49204e-05,
      "loss": 0.2761,
      "step": 312700
    },
    {
      "epoch": 0.3856,
      "grad_norm": 0.2490047663450241,
      "learning_rate": 7.48804e-05,
      "loss": 0.2843,
      "step": 312800
    },
    {
      "epoch": 0.3858,
      "grad_norm": 0.26003503799438477,
      "learning_rate": 7.48404e-05,
      "loss": 0.2663,
      "step": 312900
    },
    {
      "epoch": 0.386,
      "grad_norm": 1.0622795820236206,
      "learning_rate": 7.48004e-05,
      "loss": 0.2784,
      "step": 313000
    },
    {
      "epoch": 0.3862,
      "grad_norm": 0.25771743059158325,
      "learning_rate": 7.476040000000001e-05,
      "loss": 0.2726,
      "step": 313100
    },
    {
      "epoch": 0.3864,
      "grad_norm": 0.3834233283996582,
      "learning_rate": 7.47204e-05,
      "loss": 0.2598,
      "step": 313200
    },
    {
      "epoch": 0.3866,
      "grad_norm": 0.19467617571353912,
      "learning_rate": 7.468040000000001e-05,
      "loss": 0.2667,
      "step": 313300
    },
    {
      "epoch": 0.3868,
      "grad_norm": 0.23708763718605042,
      "learning_rate": 7.46404e-05,
      "loss": 0.2619,
      "step": 313400
    },
    {
      "epoch": 0.387,
      "grad_norm": 0.2725387215614319,
      "learning_rate": 7.46004e-05,
      "loss": 0.2723,
      "step": 313500
    },
    {
      "epoch": 0.3872,
      "grad_norm": 0.23830577731132507,
      "learning_rate": 7.45604e-05,
      "loss": 0.26,
      "step": 313600
    },
    {
      "epoch": 0.3874,
      "grad_norm": 0.20995588600635529,
      "learning_rate": 7.45204e-05,
      "loss": 0.2563,
      "step": 313700
    },
    {
      "epoch": 0.3876,
      "grad_norm": 0.20928217470645905,
      "learning_rate": 7.44804e-05,
      "loss": 0.2839,
      "step": 313800
    },
    {
      "epoch": 0.3878,
      "grad_norm": 0.21170674264431,
      "learning_rate": 7.44404e-05,
      "loss": 0.2693,
      "step": 313900
    },
    {
      "epoch": 0.388,
      "grad_norm": 0.16965176165103912,
      "learning_rate": 7.440040000000001e-05,
      "loss": 0.2739,
      "step": 314000
    },
    {
      "epoch": 0.3882,
      "grad_norm": 0.32784977555274963,
      "learning_rate": 7.436040000000001e-05,
      "loss": 0.2638,
      "step": 314100
    },
    {
      "epoch": 0.3884,
      "grad_norm": 0.45630523562431335,
      "learning_rate": 7.432040000000001e-05,
      "loss": 0.2681,
      "step": 314200
    },
    {
      "epoch": 0.3886,
      "grad_norm": 0.30572786927223206,
      "learning_rate": 7.42804e-05,
      "loss": 0.2632,
      "step": 314300
    },
    {
      "epoch": 0.3888,
      "grad_norm": 0.21931806206703186,
      "learning_rate": 7.42404e-05,
      "loss": 0.2717,
      "step": 314400
    },
    {
      "epoch": 0.389,
      "grad_norm": 0.28573718667030334,
      "learning_rate": 7.42004e-05,
      "loss": 0.2684,
      "step": 314500
    },
    {
      "epoch": 0.3892,
      "grad_norm": 2.636169672012329,
      "learning_rate": 7.416040000000001e-05,
      "loss": 0.2664,
      "step": 314600
    },
    {
      "epoch": 0.3894,
      "grad_norm": 0.19433589279651642,
      "learning_rate": 7.41204e-05,
      "loss": 0.264,
      "step": 314700
    },
    {
      "epoch": 0.3896,
      "grad_norm": 0.2967517673969269,
      "learning_rate": 7.408040000000001e-05,
      "loss": 0.2744,
      "step": 314800
    },
    {
      "epoch": 0.3898,
      "grad_norm": 0.29649585485458374,
      "learning_rate": 7.40404e-05,
      "loss": 0.2656,
      "step": 314900
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.2209979146718979,
      "learning_rate": 7.40004e-05,
      "loss": 0.275,
      "step": 315000
    },
    {
      "epoch": 0.3902,
      "grad_norm": 0.1720481663942337,
      "learning_rate": 7.39604e-05,
      "loss": 0.2581,
      "step": 315100
    },
    {
      "epoch": 0.3904,
      "grad_norm": 0.2020689994096756,
      "learning_rate": 7.39204e-05,
      "loss": 0.2617,
      "step": 315200
    },
    {
      "epoch": 0.3906,
      "grad_norm": 0.22698251903057098,
      "learning_rate": 7.38804e-05,
      "loss": 0.269,
      "step": 315300
    },
    {
      "epoch": 0.3908,
      "grad_norm": 0.2073138952255249,
      "learning_rate": 7.38404e-05,
      "loss": 0.2801,
      "step": 315400
    },
    {
      "epoch": 0.391,
      "grad_norm": 0.4935136139392853,
      "learning_rate": 7.38004e-05,
      "loss": 0.2653,
      "step": 315500
    },
    {
      "epoch": 0.3912,
      "grad_norm": 0.1566108763217926,
      "learning_rate": 7.376040000000001e-05,
      "loss": 0.258,
      "step": 315600
    },
    {
      "epoch": 0.3914,
      "grad_norm": 0.24738989770412445,
      "learning_rate": 7.37204e-05,
      "loss": 0.2634,
      "step": 315700
    },
    {
      "epoch": 0.3916,
      "grad_norm": 0.7126500010490417,
      "learning_rate": 7.36804e-05,
      "loss": 0.2653,
      "step": 315800
    },
    {
      "epoch": 0.3918,
      "grad_norm": 0.23705486953258514,
      "learning_rate": 7.36404e-05,
      "loss": 0.2677,
      "step": 315900
    },
    {
      "epoch": 0.392,
      "grad_norm": 0.3229648470878601,
      "learning_rate": 7.36004e-05,
      "loss": 0.2816,
      "step": 316000
    },
    {
      "epoch": 0.3922,
      "grad_norm": 0.20382308959960938,
      "learning_rate": 7.356040000000001e-05,
      "loss": 0.2606,
      "step": 316100
    },
    {
      "epoch": 0.3924,
      "grad_norm": 0.19805587828159332,
      "learning_rate": 7.35204e-05,
      "loss": 0.2673,
      "step": 316200
    },
    {
      "epoch": 0.3926,
      "grad_norm": 0.24583861231803894,
      "learning_rate": 7.348040000000001e-05,
      "loss": 0.273,
      "step": 316300
    },
    {
      "epoch": 0.3928,
      "grad_norm": 0.2229291796684265,
      "learning_rate": 7.34404e-05,
      "loss": 0.2791,
      "step": 316400
    },
    {
      "epoch": 0.393,
      "grad_norm": 0.21293219923973083,
      "learning_rate": 7.34004e-05,
      "loss": 0.2672,
      "step": 316500
    },
    {
      "epoch": 0.3932,
      "grad_norm": 0.20457857847213745,
      "learning_rate": 7.33604e-05,
      "loss": 0.261,
      "step": 316600
    },
    {
      "epoch": 0.3934,
      "grad_norm": 0.21731828153133392,
      "learning_rate": 7.33204e-05,
      "loss": 0.2681,
      "step": 316700
    },
    {
      "epoch": 0.3936,
      "grad_norm": 0.26696541905403137,
      "learning_rate": 7.32804e-05,
      "loss": 0.2708,
      "step": 316800
    },
    {
      "epoch": 0.3938,
      "grad_norm": 0.21308811008930206,
      "learning_rate": 7.324040000000001e-05,
      "loss": 0.2642,
      "step": 316900
    },
    {
      "epoch": 0.394,
      "grad_norm": 0.2161969393491745,
      "learning_rate": 7.32004e-05,
      "loss": 0.2633,
      "step": 317000
    },
    {
      "epoch": 0.3942,
      "grad_norm": 0.21060483157634735,
      "learning_rate": 7.316040000000001e-05,
      "loss": 0.2669,
      "step": 317100
    },
    {
      "epoch": 0.3944,
      "grad_norm": 0.2582820653915405,
      "learning_rate": 7.31204e-05,
      "loss": 0.2644,
      "step": 317200
    },
    {
      "epoch": 0.3946,
      "grad_norm": 0.24517767131328583,
      "learning_rate": 7.30804e-05,
      "loss": 0.2691,
      "step": 317300
    },
    {
      "epoch": 0.3948,
      "grad_norm": 0.2194310426712036,
      "learning_rate": 7.30404e-05,
      "loss": 0.2653,
      "step": 317400
    },
    {
      "epoch": 0.395,
      "grad_norm": 0.28793978691101074,
      "learning_rate": 7.30004e-05,
      "loss": 0.2643,
      "step": 317500
    },
    {
      "epoch": 0.3952,
      "grad_norm": 0.6977545022964478,
      "learning_rate": 7.296040000000001e-05,
      "loss": 0.2723,
      "step": 317600
    },
    {
      "epoch": 0.3954,
      "grad_norm": 0.2802855372428894,
      "learning_rate": 7.29204e-05,
      "loss": 0.3051,
      "step": 317700
    },
    {
      "epoch": 0.3956,
      "grad_norm": 0.2557845413684845,
      "learning_rate": 7.288040000000001e-05,
      "loss": 0.2639,
      "step": 317800
    },
    {
      "epoch": 0.3958,
      "grad_norm": 0.3409139811992645,
      "learning_rate": 7.284040000000001e-05,
      "loss": 0.279,
      "step": 317900
    },
    {
      "epoch": 0.396,
      "grad_norm": 0.18252715468406677,
      "learning_rate": 7.280040000000001e-05,
      "loss": 0.2664,
      "step": 318000
    },
    {
      "epoch": 0.3962,
      "grad_norm": 0.2673168182373047,
      "learning_rate": 7.27604e-05,
      "loss": 0.2685,
      "step": 318100
    },
    {
      "epoch": 0.3964,
      "grad_norm": 0.37415772676467896,
      "learning_rate": 7.27204e-05,
      "loss": 0.2564,
      "step": 318200
    },
    {
      "epoch": 0.3966,
      "grad_norm": 0.23668460547924042,
      "learning_rate": 7.26804e-05,
      "loss": 0.2792,
      "step": 318300
    },
    {
      "epoch": 0.3968,
      "grad_norm": 0.4940067231655121,
      "learning_rate": 7.264040000000001e-05,
      "loss": 0.2757,
      "step": 318400
    },
    {
      "epoch": 0.397,
      "grad_norm": 0.18225626647472382,
      "learning_rate": 7.26004e-05,
      "loss": 0.2789,
      "step": 318500
    },
    {
      "epoch": 0.3972,
      "grad_norm": 0.19703717529773712,
      "learning_rate": 7.256040000000001e-05,
      "loss": 0.2761,
      "step": 318600
    },
    {
      "epoch": 0.3974,
      "grad_norm": 0.21940729022026062,
      "learning_rate": 7.25204e-05,
      "loss": 0.291,
      "step": 318700
    },
    {
      "epoch": 0.3976,
      "grad_norm": 0.293050080537796,
      "learning_rate": 7.24804e-05,
      "loss": 0.2809,
      "step": 318800
    },
    {
      "epoch": 0.3978,
      "grad_norm": 0.45579320192337036,
      "learning_rate": 7.24404e-05,
      "loss": 0.2659,
      "step": 318900
    },
    {
      "epoch": 0.398,
      "grad_norm": 0.20021362602710724,
      "learning_rate": 7.24004e-05,
      "loss": 0.2687,
      "step": 319000
    },
    {
      "epoch": 0.3982,
      "grad_norm": 0.22239625453948975,
      "learning_rate": 7.23604e-05,
      "loss": 0.2789,
      "step": 319100
    },
    {
      "epoch": 0.3984,
      "grad_norm": 0.2077815979719162,
      "learning_rate": 7.23204e-05,
      "loss": 0.2563,
      "step": 319200
    },
    {
      "epoch": 0.3986,
      "grad_norm": 0.2690514028072357,
      "learning_rate": 7.228040000000001e-05,
      "loss": 0.2593,
      "step": 319300
    },
    {
      "epoch": 0.3988,
      "grad_norm": 0.18208642303943634,
      "learning_rate": 7.224040000000001e-05,
      "loss": 0.2709,
      "step": 319400
    },
    {
      "epoch": 0.399,
      "grad_norm": 0.16601015627384186,
      "learning_rate": 7.220040000000001e-05,
      "loss": 0.2738,
      "step": 319500
    },
    {
      "epoch": 0.3992,
      "grad_norm": 0.18831057846546173,
      "learning_rate": 7.21604e-05,
      "loss": 0.2703,
      "step": 319600
    },
    {
      "epoch": 0.3994,
      "grad_norm": 0.19717268645763397,
      "learning_rate": 7.21204e-05,
      "loss": 0.2563,
      "step": 319700
    },
    {
      "epoch": 0.3996,
      "grad_norm": 0.2667396366596222,
      "learning_rate": 7.20804e-05,
      "loss": 0.2578,
      "step": 319800
    },
    {
      "epoch": 0.3998,
      "grad_norm": 0.2868683934211731,
      "learning_rate": 7.204040000000001e-05,
      "loss": 0.2691,
      "step": 319900
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.21218955516815186,
      "learning_rate": 7.20004e-05,
      "loss": 0.2637,
      "step": 320000
    },
    {
      "epoch": 0.4002,
      "grad_norm": 0.4244702160358429,
      "learning_rate": 7.196040000000001e-05,
      "loss": 0.2767,
      "step": 320100
    },
    {
      "epoch": 0.4004,
      "grad_norm": 0.23811425268650055,
      "learning_rate": 7.19204e-05,
      "loss": 0.2627,
      "step": 320200
    },
    {
      "epoch": 0.4006,
      "grad_norm": 0.20318666100502014,
      "learning_rate": 7.18804e-05,
      "loss": 0.2616,
      "step": 320300
    },
    {
      "epoch": 0.4008,
      "grad_norm": 0.3025484085083008,
      "learning_rate": 7.18404e-05,
      "loss": 0.2685,
      "step": 320400
    },
    {
      "epoch": 0.401,
      "grad_norm": 0.17960281670093536,
      "learning_rate": 7.18004e-05,
      "loss": 0.2733,
      "step": 320500
    },
    {
      "epoch": 0.4012,
      "grad_norm": 0.25798577070236206,
      "learning_rate": 7.17604e-05,
      "loss": 0.264,
      "step": 320600
    },
    {
      "epoch": 0.4014,
      "grad_norm": 0.1824919730424881,
      "learning_rate": 7.17204e-05,
      "loss": 0.2684,
      "step": 320700
    },
    {
      "epoch": 0.4016,
      "grad_norm": 0.23640932142734528,
      "learning_rate": 7.16804e-05,
      "loss": 0.265,
      "step": 320800
    },
    {
      "epoch": 0.4018,
      "grad_norm": 0.22346696257591248,
      "learning_rate": 7.164040000000001e-05,
      "loss": 0.2574,
      "step": 320900
    },
    {
      "epoch": 0.402,
      "grad_norm": 0.33641451597213745,
      "learning_rate": 7.16004e-05,
      "loss": 0.2644,
      "step": 321000
    },
    {
      "epoch": 0.4022,
      "grad_norm": 0.24027025699615479,
      "learning_rate": 7.15604e-05,
      "loss": 0.2803,
      "step": 321100
    },
    {
      "epoch": 0.4024,
      "grad_norm": 0.2125771939754486,
      "learning_rate": 7.15204e-05,
      "loss": 0.2733,
      "step": 321200
    },
    {
      "epoch": 0.4026,
      "grad_norm": 0.2774077355861664,
      "learning_rate": 7.14804e-05,
      "loss": 0.2781,
      "step": 321300
    },
    {
      "epoch": 0.4028,
      "grad_norm": 0.1831909418106079,
      "learning_rate": 7.144040000000001e-05,
      "loss": 0.2661,
      "step": 321400
    },
    {
      "epoch": 0.403,
      "grad_norm": 0.2540408968925476,
      "learning_rate": 7.14004e-05,
      "loss": 0.2801,
      "step": 321500
    },
    {
      "epoch": 0.4032,
      "grad_norm": 0.2425178438425064,
      "learning_rate": 7.136040000000001e-05,
      "loss": 0.2584,
      "step": 321600
    },
    {
      "epoch": 0.4034,
      "grad_norm": 0.1953432857990265,
      "learning_rate": 7.13204e-05,
      "loss": 0.2674,
      "step": 321700
    },
    {
      "epoch": 0.4036,
      "grad_norm": 0.31084179878234863,
      "learning_rate": 7.12804e-05,
      "loss": 0.2599,
      "step": 321800
    },
    {
      "epoch": 0.4038,
      "grad_norm": 0.17004917562007904,
      "learning_rate": 7.12404e-05,
      "loss": 0.272,
      "step": 321900
    },
    {
      "epoch": 0.404,
      "grad_norm": 0.2186584621667862,
      "learning_rate": 7.12004e-05,
      "loss": 0.2615,
      "step": 322000
    },
    {
      "epoch": 0.4042,
      "grad_norm": 0.2131323516368866,
      "learning_rate": 7.11604e-05,
      "loss": 0.2734,
      "step": 322100
    },
    {
      "epoch": 0.4044,
      "grad_norm": 0.29444190859794617,
      "learning_rate": 7.11204e-05,
      "loss": 0.2577,
      "step": 322200
    },
    {
      "epoch": 0.4046,
      "grad_norm": 0.2686418890953064,
      "learning_rate": 7.10804e-05,
      "loss": 0.2636,
      "step": 322300
    },
    {
      "epoch": 0.4048,
      "grad_norm": 0.2958160936832428,
      "learning_rate": 7.104040000000001e-05,
      "loss": 0.2611,
      "step": 322400
    },
    {
      "epoch": 0.405,
      "grad_norm": 0.24784350395202637,
      "learning_rate": 7.10004e-05,
      "loss": 0.2658,
      "step": 322500
    },
    {
      "epoch": 0.4052,
      "grad_norm": 0.21654903888702393,
      "learning_rate": 7.09604e-05,
      "loss": 0.2625,
      "step": 322600
    },
    {
      "epoch": 0.4054,
      "grad_norm": 0.17862387001514435,
      "learning_rate": 7.092039999999999e-05,
      "loss": 0.2611,
      "step": 322700
    },
    {
      "epoch": 0.4056,
      "grad_norm": 0.2753390371799469,
      "learning_rate": 7.08804e-05,
      "loss": 0.2626,
      "step": 322800
    },
    {
      "epoch": 0.4058,
      "grad_norm": 0.41417863965034485,
      "learning_rate": 7.084040000000001e-05,
      "loss": 0.262,
      "step": 322900
    },
    {
      "epoch": 0.406,
      "grad_norm": 0.6197725534439087,
      "learning_rate": 7.08004e-05,
      "loss": 0.2647,
      "step": 323000
    },
    {
      "epoch": 0.4062,
      "grad_norm": 0.31697189807891846,
      "learning_rate": 7.076040000000001e-05,
      "loss": 0.2633,
      "step": 323100
    },
    {
      "epoch": 0.4064,
      "grad_norm": 0.7414258718490601,
      "learning_rate": 7.072040000000001e-05,
      "loss": 0.271,
      "step": 323200
    },
    {
      "epoch": 0.4066,
      "grad_norm": 0.22181887924671173,
      "learning_rate": 7.068040000000001e-05,
      "loss": 0.2637,
      "step": 323300
    },
    {
      "epoch": 0.4068,
      "grad_norm": 0.24176335334777832,
      "learning_rate": 7.06404e-05,
      "loss": 0.2641,
      "step": 323400
    },
    {
      "epoch": 0.407,
      "grad_norm": 0.2450632005929947,
      "learning_rate": 7.06004e-05,
      "loss": 0.2634,
      "step": 323500
    },
    {
      "epoch": 0.4072,
      "grad_norm": 0.1967063993215561,
      "learning_rate": 7.05604e-05,
      "loss": 0.273,
      "step": 323600
    },
    {
      "epoch": 0.4074,
      "grad_norm": 0.16844309866428375,
      "learning_rate": 7.052040000000001e-05,
      "loss": 0.2664,
      "step": 323700
    },
    {
      "epoch": 0.4076,
      "grad_norm": 0.24496600031852722,
      "learning_rate": 7.04804e-05,
      "loss": 0.2706,
      "step": 323800
    },
    {
      "epoch": 0.4078,
      "grad_norm": 0.24819956719875336,
      "learning_rate": 7.044040000000001e-05,
      "loss": 0.2573,
      "step": 323900
    },
    {
      "epoch": 0.408,
      "grad_norm": 0.31527596712112427,
      "learning_rate": 7.04004e-05,
      "loss": 0.2602,
      "step": 324000
    },
    {
      "epoch": 0.4082,
      "grad_norm": 0.22740913927555084,
      "learning_rate": 7.03604e-05,
      "loss": 0.2614,
      "step": 324100
    },
    {
      "epoch": 0.4084,
      "grad_norm": 0.2220146805047989,
      "learning_rate": 7.03204e-05,
      "loss": 0.2682,
      "step": 324200
    },
    {
      "epoch": 0.4086,
      "grad_norm": 0.22316837310791016,
      "learning_rate": 7.02804e-05,
      "loss": 0.2612,
      "step": 324300
    },
    {
      "epoch": 0.4088,
      "grad_norm": 0.2431160807609558,
      "learning_rate": 7.02404e-05,
      "loss": 0.2753,
      "step": 324400
    },
    {
      "epoch": 0.409,
      "grad_norm": 0.2228887677192688,
      "learning_rate": 7.02004e-05,
      "loss": 0.2612,
      "step": 324500
    },
    {
      "epoch": 0.4092,
      "grad_norm": 0.25735148787498474,
      "learning_rate": 7.01604e-05,
      "loss": 0.2639,
      "step": 324600
    },
    {
      "epoch": 0.4094,
      "grad_norm": 0.19601066410541534,
      "learning_rate": 7.012040000000001e-05,
      "loss": 0.2515,
      "step": 324700
    },
    {
      "epoch": 0.4096,
      "grad_norm": 0.2028639167547226,
      "learning_rate": 7.008040000000001e-05,
      "loss": 0.2704,
      "step": 324800
    },
    {
      "epoch": 0.4098,
      "grad_norm": 0.18071521818637848,
      "learning_rate": 7.00404e-05,
      "loss": 0.2657,
      "step": 324900
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.22790144383907318,
      "learning_rate": 7.00004e-05,
      "loss": 0.2852,
      "step": 325000
    },
    {
      "epoch": 0.4102,
      "grad_norm": 0.2082364410161972,
      "learning_rate": 6.99604e-05,
      "loss": 0.2622,
      "step": 325100
    },
    {
      "epoch": 0.4104,
      "grad_norm": 0.1667306274175644,
      "learning_rate": 6.992040000000001e-05,
      "loss": 0.2555,
      "step": 325200
    },
    {
      "epoch": 0.4106,
      "grad_norm": 0.1831902712583542,
      "learning_rate": 6.98804e-05,
      "loss": 0.2698,
      "step": 325300
    },
    {
      "epoch": 0.4108,
      "grad_norm": 0.18110473453998566,
      "learning_rate": 6.984040000000001e-05,
      "loss": 0.2496,
      "step": 325400
    },
    {
      "epoch": 0.411,
      "grad_norm": 0.19504287838935852,
      "learning_rate": 6.98004e-05,
      "loss": 0.2556,
      "step": 325500
    },
    {
      "epoch": 0.4112,
      "grad_norm": 0.16663411259651184,
      "learning_rate": 6.97604e-05,
      "loss": 0.2577,
      "step": 325600
    },
    {
      "epoch": 0.4114,
      "grad_norm": 0.43907126784324646,
      "learning_rate": 6.97204e-05,
      "loss": 0.2585,
      "step": 325700
    },
    {
      "epoch": 0.4116,
      "grad_norm": 0.20249086618423462,
      "learning_rate": 6.96804e-05,
      "loss": 0.25,
      "step": 325800
    },
    {
      "epoch": 0.4118,
      "grad_norm": 0.15998288989067078,
      "learning_rate": 6.96404e-05,
      "loss": 0.2482,
      "step": 325900
    },
    {
      "epoch": 0.412,
      "grad_norm": 0.19670312106609344,
      "learning_rate": 6.96004e-05,
      "loss": 0.2445,
      "step": 326000
    },
    {
      "epoch": 0.4122,
      "grad_norm": 0.18469685316085815,
      "learning_rate": 6.95604e-05,
      "loss": 0.2502,
      "step": 326100
    },
    {
      "epoch": 0.4124,
      "grad_norm": 0.18897442519664764,
      "learning_rate": 6.952040000000001e-05,
      "loss": 0.2474,
      "step": 326200
    },
    {
      "epoch": 0.4126,
      "grad_norm": 0.17963925004005432,
      "learning_rate": 6.94804e-05,
      "loss": 0.2457,
      "step": 326300
    },
    {
      "epoch": 0.4128,
      "grad_norm": 0.19767078757286072,
      "learning_rate": 6.94404e-05,
      "loss": 0.2515,
      "step": 326400
    },
    {
      "epoch": 0.413,
      "grad_norm": 0.20545931160449982,
      "learning_rate": 6.94004e-05,
      "loss": 0.2459,
      "step": 326500
    },
    {
      "epoch": 0.4132,
      "grad_norm": 0.17647592723369598,
      "learning_rate": 6.93604e-05,
      "loss": 0.2457,
      "step": 326600
    },
    {
      "epoch": 0.4134,
      "grad_norm": 0.19674290716648102,
      "learning_rate": 6.932040000000001e-05,
      "loss": 0.2469,
      "step": 326700
    },
    {
      "epoch": 0.4136,
      "grad_norm": 0.25124111771583557,
      "learning_rate": 6.92804e-05,
      "loss": 0.2434,
      "step": 326800
    },
    {
      "epoch": 0.4138,
      "grad_norm": 0.19026201963424683,
      "learning_rate": 6.924040000000001e-05,
      "loss": 0.2536,
      "step": 326900
    },
    {
      "epoch": 0.414,
      "grad_norm": 0.20437213778495789,
      "learning_rate": 6.92004e-05,
      "loss": 0.2459,
      "step": 327000
    },
    {
      "epoch": 0.4142,
      "grad_norm": 0.24830076098442078,
      "learning_rate": 6.916040000000001e-05,
      "loss": 0.2454,
      "step": 327100
    },
    {
      "epoch": 0.4144,
      "grad_norm": 0.19002242386341095,
      "learning_rate": 6.91204e-05,
      "loss": 0.2472,
      "step": 327200
    },
    {
      "epoch": 0.4146,
      "grad_norm": 0.1964949518442154,
      "learning_rate": 6.90804e-05,
      "loss": 0.2435,
      "step": 327300
    },
    {
      "epoch": 0.4148,
      "grad_norm": 0.20290331542491913,
      "learning_rate": 6.90404e-05,
      "loss": 0.2452,
      "step": 327400
    },
    {
      "epoch": 0.415,
      "grad_norm": 0.17645521461963654,
      "learning_rate": 6.90004e-05,
      "loss": 0.2441,
      "step": 327500
    },
    {
      "epoch": 0.4152,
      "grad_norm": 0.19181664288043976,
      "learning_rate": 6.89604e-05,
      "loss": 0.2487,
      "step": 327600
    },
    {
      "epoch": 0.4154,
      "grad_norm": 0.18161311745643616,
      "learning_rate": 6.892040000000001e-05,
      "loss": 0.2446,
      "step": 327700
    },
    {
      "epoch": 0.4156,
      "grad_norm": 0.16001363098621368,
      "learning_rate": 6.88804e-05,
      "loss": 0.2443,
      "step": 327800
    },
    {
      "epoch": 0.4158,
      "grad_norm": 0.1850643903017044,
      "learning_rate": 6.88404e-05,
      "loss": 0.253,
      "step": 327900
    },
    {
      "epoch": 0.416,
      "grad_norm": 0.1478988528251648,
      "learning_rate": 6.880039999999999e-05,
      "loss": 0.2445,
      "step": 328000
    },
    {
      "epoch": 0.4162,
      "grad_norm": 0.18644775450229645,
      "learning_rate": 6.87604e-05,
      "loss": 0.243,
      "step": 328100
    },
    {
      "epoch": 0.4164,
      "grad_norm": 0.18201524019241333,
      "learning_rate": 6.872040000000001e-05,
      "loss": 0.2423,
      "step": 328200
    },
    {
      "epoch": 0.4166,
      "grad_norm": 0.25159624218940735,
      "learning_rate": 6.86804e-05,
      "loss": 0.2461,
      "step": 328300
    },
    {
      "epoch": 0.4168,
      "grad_norm": 0.21256084740161896,
      "learning_rate": 6.864040000000001e-05,
      "loss": 0.2443,
      "step": 328400
    },
    {
      "epoch": 0.417,
      "grad_norm": 0.27206477522850037,
      "learning_rate": 6.86004e-05,
      "loss": 0.2421,
      "step": 328500
    },
    {
      "epoch": 0.4172,
      "grad_norm": 0.1862238049507141,
      "learning_rate": 6.856040000000001e-05,
      "loss": 0.253,
      "step": 328600
    },
    {
      "epoch": 0.4174,
      "grad_norm": 0.1940353810787201,
      "learning_rate": 6.85204e-05,
      "loss": 0.2454,
      "step": 328700
    },
    {
      "epoch": 0.4176,
      "grad_norm": 0.14966708421707153,
      "learning_rate": 6.84804e-05,
      "loss": 0.2471,
      "step": 328800
    },
    {
      "epoch": 0.4178,
      "grad_norm": 0.15949209034442902,
      "learning_rate": 6.84404e-05,
      "loss": 0.2497,
      "step": 328900
    },
    {
      "epoch": 0.418,
      "grad_norm": 0.198212668299675,
      "learning_rate": 6.84004e-05,
      "loss": 0.2496,
      "step": 329000
    },
    {
      "epoch": 0.4182,
      "grad_norm": 0.22681452333927155,
      "learning_rate": 6.83604e-05,
      "loss": 0.2464,
      "step": 329100
    },
    {
      "epoch": 0.4184,
      "grad_norm": 0.17633529007434845,
      "learning_rate": 6.832040000000001e-05,
      "loss": 0.2519,
      "step": 329200
    },
    {
      "epoch": 0.4186,
      "grad_norm": 0.2030135840177536,
      "learning_rate": 6.82804e-05,
      "loss": 0.2462,
      "step": 329300
    },
    {
      "epoch": 0.4188,
      "grad_norm": 0.16612933576107025,
      "learning_rate": 6.82404e-05,
      "loss": 0.2453,
      "step": 329400
    },
    {
      "epoch": 0.419,
      "grad_norm": 0.2159220576286316,
      "learning_rate": 6.82004e-05,
      "loss": 0.2478,
      "step": 329500
    },
    {
      "epoch": 0.4192,
      "grad_norm": 0.24220001697540283,
      "learning_rate": 6.81604e-05,
      "loss": 0.2477,
      "step": 329600
    },
    {
      "epoch": 0.4194,
      "grad_norm": 0.1778249442577362,
      "learning_rate": 6.81204e-05,
      "loss": 0.2516,
      "step": 329700
    },
    {
      "epoch": 0.4196,
      "grad_norm": 0.2725713551044464,
      "learning_rate": 6.80804e-05,
      "loss": 0.2514,
      "step": 329800
    },
    {
      "epoch": 0.4198,
      "grad_norm": 0.18258759379386902,
      "learning_rate": 6.80404e-05,
      "loss": 0.2479,
      "step": 329900
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.24233517050743103,
      "learning_rate": 6.800040000000001e-05,
      "loss": 0.2501,
      "step": 330000
    },
    {
      "epoch": 0.4202,
      "grad_norm": 0.16992133855819702,
      "learning_rate": 6.796040000000001e-05,
      "loss": 0.2422,
      "step": 330100
    },
    {
      "epoch": 0.4204,
      "grad_norm": 0.19478961825370789,
      "learning_rate": 6.79204e-05,
      "loss": 0.2504,
      "step": 330200
    },
    {
      "epoch": 0.4206,
      "grad_norm": 0.3001626133918762,
      "learning_rate": 6.78804e-05,
      "loss": 0.2555,
      "step": 330300
    },
    {
      "epoch": 0.4208,
      "grad_norm": 0.15378372371196747,
      "learning_rate": 6.78404e-05,
      "loss": 0.2442,
      "step": 330400
    },
    {
      "epoch": 0.421,
      "grad_norm": 0.17064020037651062,
      "learning_rate": 6.780040000000001e-05,
      "loss": 0.2448,
      "step": 330500
    },
    {
      "epoch": 0.4212,
      "grad_norm": 0.15659099817276,
      "learning_rate": 6.77604e-05,
      "loss": 0.2453,
      "step": 330600
    },
    {
      "epoch": 0.4214,
      "grad_norm": 0.274606317281723,
      "learning_rate": 6.772040000000001e-05,
      "loss": 0.2458,
      "step": 330700
    },
    {
      "epoch": 0.4216,
      "grad_norm": 0.19441047310829163,
      "learning_rate": 6.76804e-05,
      "loss": 0.2455,
      "step": 330800
    },
    {
      "epoch": 0.4218,
      "grad_norm": 0.16700878739356995,
      "learning_rate": 6.76404e-05,
      "loss": 0.2433,
      "step": 330900
    },
    {
      "epoch": 0.422,
      "grad_norm": 0.15905675292015076,
      "learning_rate": 6.76004e-05,
      "loss": 0.2447,
      "step": 331000
    },
    {
      "epoch": 0.4222,
      "grad_norm": 0.2215052992105484,
      "learning_rate": 6.75604e-05,
      "loss": 0.2492,
      "step": 331100
    },
    {
      "epoch": 0.4224,
      "grad_norm": 0.17576748132705688,
      "learning_rate": 6.75204e-05,
      "loss": 0.2476,
      "step": 331200
    },
    {
      "epoch": 0.4226,
      "grad_norm": 0.17451365292072296,
      "learning_rate": 6.74804e-05,
      "loss": 0.2487,
      "step": 331300
    },
    {
      "epoch": 0.4228,
      "grad_norm": 0.5440964102745056,
      "learning_rate": 6.74404e-05,
      "loss": 0.2479,
      "step": 331400
    },
    {
      "epoch": 0.423,
      "grad_norm": 0.19704194366931915,
      "learning_rate": 6.740040000000001e-05,
      "loss": 0.247,
      "step": 331500
    },
    {
      "epoch": 0.4232,
      "grad_norm": 0.20427988469600677,
      "learning_rate": 6.73604e-05,
      "loss": 0.2481,
      "step": 331600
    },
    {
      "epoch": 0.4234,
      "grad_norm": 0.16605882346630096,
      "learning_rate": 6.73204e-05,
      "loss": 0.2466,
      "step": 331700
    },
    {
      "epoch": 0.4236,
      "grad_norm": 0.22239907085895538,
      "learning_rate": 6.72804e-05,
      "loss": 0.243,
      "step": 331800
    },
    {
      "epoch": 0.4238,
      "grad_norm": 0.17800769209861755,
      "learning_rate": 6.72404e-05,
      "loss": 0.2489,
      "step": 331900
    },
    {
      "epoch": 0.424,
      "grad_norm": 0.2153862714767456,
      "learning_rate": 6.720040000000001e-05,
      "loss": 0.2489,
      "step": 332000
    },
    {
      "epoch": 0.4242,
      "grad_norm": 0.3256576359272003,
      "learning_rate": 6.71604e-05,
      "loss": 0.242,
      "step": 332100
    },
    {
      "epoch": 0.4244,
      "grad_norm": 0.20492444932460785,
      "learning_rate": 6.712040000000001e-05,
      "loss": 0.242,
      "step": 332200
    },
    {
      "epoch": 0.4246,
      "grad_norm": 0.276932954788208,
      "learning_rate": 6.70804e-05,
      "loss": 0.2524,
      "step": 332300
    },
    {
      "epoch": 0.4248,
      "grad_norm": 0.17257800698280334,
      "learning_rate": 6.704040000000001e-05,
      "loss": 0.2448,
      "step": 332400
    },
    {
      "epoch": 0.425,
      "grad_norm": 0.1662578284740448,
      "learning_rate": 6.70004e-05,
      "loss": 0.2423,
      "step": 332500
    },
    {
      "epoch": 0.4252,
      "grad_norm": 0.17013122141361237,
      "learning_rate": 6.69604e-05,
      "loss": 0.2418,
      "step": 332600
    },
    {
      "epoch": 0.4254,
      "grad_norm": 0.18194365501403809,
      "learning_rate": 6.69204e-05,
      "loss": 0.2484,
      "step": 332700
    },
    {
      "epoch": 0.4256,
      "grad_norm": 0.17675453424453735,
      "learning_rate": 6.68804e-05,
      "loss": 0.2432,
      "step": 332800
    },
    {
      "epoch": 0.4258,
      "grad_norm": 0.16768479347229004,
      "learning_rate": 6.68404e-05,
      "loss": 0.2446,
      "step": 332900
    },
    {
      "epoch": 0.426,
      "grad_norm": 0.28635919094085693,
      "learning_rate": 6.680040000000001e-05,
      "loss": 0.2471,
      "step": 333000
    },
    {
      "epoch": 0.4262,
      "grad_norm": 0.26237931847572327,
      "learning_rate": 6.67604e-05,
      "loss": 0.2461,
      "step": 333100
    },
    {
      "epoch": 0.4264,
      "grad_norm": 0.15941229462623596,
      "learning_rate": 6.67204e-05,
      "loss": 0.2517,
      "step": 333200
    },
    {
      "epoch": 0.4266,
      "grad_norm": 0.20639023184776306,
      "learning_rate": 6.668039999999999e-05,
      "loss": 0.2439,
      "step": 333300
    },
    {
      "epoch": 0.4268,
      "grad_norm": 0.1907823532819748,
      "learning_rate": 6.66404e-05,
      "loss": 0.2441,
      "step": 333400
    },
    {
      "epoch": 0.427,
      "grad_norm": 0.4512256979942322,
      "learning_rate": 6.660040000000001e-05,
      "loss": 0.2555,
      "step": 333500
    },
    {
      "epoch": 0.4272,
      "grad_norm": 0.2021334320306778,
      "learning_rate": 6.65604e-05,
      "loss": 0.2445,
      "step": 333600
    },
    {
      "epoch": 0.4274,
      "grad_norm": 0.2123294472694397,
      "learning_rate": 6.652040000000001e-05,
      "loss": 0.2421,
      "step": 333700
    },
    {
      "epoch": 0.4276,
      "grad_norm": 0.2353043407201767,
      "learning_rate": 6.64804e-05,
      "loss": 0.246,
      "step": 333800
    },
    {
      "epoch": 0.4278,
      "grad_norm": 0.18522989749908447,
      "learning_rate": 6.644040000000001e-05,
      "loss": 0.2466,
      "step": 333900
    },
    {
      "epoch": 0.428,
      "grad_norm": 0.29194357991218567,
      "learning_rate": 6.64004e-05,
      "loss": 0.2472,
      "step": 334000
    },
    {
      "epoch": 0.4282,
      "grad_norm": 0.21177667379379272,
      "learning_rate": 6.63604e-05,
      "loss": 0.2451,
      "step": 334100
    },
    {
      "epoch": 0.4284,
      "grad_norm": 0.2588765323162079,
      "learning_rate": 6.63204e-05,
      "loss": 0.2426,
      "step": 334200
    },
    {
      "epoch": 0.4286,
      "grad_norm": 0.16683948040008545,
      "learning_rate": 6.62804e-05,
      "loss": 0.2385,
      "step": 334300
    },
    {
      "epoch": 0.4288,
      "grad_norm": 0.2573770582675934,
      "learning_rate": 6.62404e-05,
      "loss": 0.2441,
      "step": 334400
    },
    {
      "epoch": 0.429,
      "grad_norm": 0.24478641152381897,
      "learning_rate": 6.620040000000001e-05,
      "loss": 0.2426,
      "step": 334500
    },
    {
      "epoch": 0.4292,
      "grad_norm": 0.2206483632326126,
      "learning_rate": 6.61604e-05,
      "loss": 0.2465,
      "step": 334600
    },
    {
      "epoch": 0.4294,
      "grad_norm": 0.16202546656131744,
      "learning_rate": 6.61204e-05,
      "loss": 0.2448,
      "step": 334700
    },
    {
      "epoch": 0.4296,
      "grad_norm": 0.29942259192466736,
      "learning_rate": 6.608039999999999e-05,
      "loss": 0.2507,
      "step": 334800
    },
    {
      "epoch": 0.4298,
      "grad_norm": 0.15078793466091156,
      "learning_rate": 6.60404e-05,
      "loss": 0.2448,
      "step": 334900
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.19433638453483582,
      "learning_rate": 6.60004e-05,
      "loss": 0.2399,
      "step": 335000
    },
    {
      "epoch": 0.4302,
      "grad_norm": 0.19410523772239685,
      "learning_rate": 6.59604e-05,
      "loss": 0.2456,
      "step": 335100
    },
    {
      "epoch": 0.4304,
      "grad_norm": 0.2051495611667633,
      "learning_rate": 6.59204e-05,
      "loss": 0.245,
      "step": 335200
    },
    {
      "epoch": 0.4306,
      "grad_norm": 0.1647467315196991,
      "learning_rate": 6.58804e-05,
      "loss": 0.2444,
      "step": 335300
    },
    {
      "epoch": 0.4308,
      "grad_norm": 0.25535687804222107,
      "learning_rate": 6.584040000000001e-05,
      "loss": 0.2452,
      "step": 335400
    },
    {
      "epoch": 0.431,
      "grad_norm": 0.16099178791046143,
      "learning_rate": 6.58004e-05,
      "loss": 0.2415,
      "step": 335500
    },
    {
      "epoch": 0.4312,
      "grad_norm": 0.15489977598190308,
      "learning_rate": 6.57604e-05,
      "loss": 0.2452,
      "step": 335600
    },
    {
      "epoch": 0.4314,
      "grad_norm": 0.18465201556682587,
      "learning_rate": 6.57204e-05,
      "loss": 0.2441,
      "step": 335700
    },
    {
      "epoch": 0.4316,
      "grad_norm": 0.24972856044769287,
      "learning_rate": 6.568040000000001e-05,
      "loss": 0.2441,
      "step": 335800
    },
    {
      "epoch": 0.4318,
      "grad_norm": 0.1736900806427002,
      "learning_rate": 6.56404e-05,
      "loss": 0.2452,
      "step": 335900
    },
    {
      "epoch": 0.432,
      "grad_norm": 0.18592572212219238,
      "learning_rate": 6.560040000000001e-05,
      "loss": 0.2397,
      "step": 336000
    },
    {
      "epoch": 0.4322,
      "grad_norm": 0.1998361349105835,
      "learning_rate": 6.55604e-05,
      "loss": 0.2399,
      "step": 336100
    },
    {
      "epoch": 0.4324,
      "grad_norm": 0.25269144773483276,
      "learning_rate": 6.552040000000001e-05,
      "loss": 0.2429,
      "step": 336200
    },
    {
      "epoch": 0.4326,
      "grad_norm": 0.2118110954761505,
      "learning_rate": 6.54804e-05,
      "loss": 0.2397,
      "step": 336300
    },
    {
      "epoch": 0.4328,
      "grad_norm": 0.20308533310890198,
      "learning_rate": 6.54404e-05,
      "loss": 0.248,
      "step": 336400
    },
    {
      "epoch": 0.433,
      "grad_norm": 0.18508122861385345,
      "learning_rate": 6.54004e-05,
      "loss": 0.2446,
      "step": 336500
    },
    {
      "epoch": 0.4332,
      "grad_norm": 0.1768845021724701,
      "learning_rate": 6.53604e-05,
      "loss": 0.2472,
      "step": 336600
    },
    {
      "epoch": 0.4334,
      "grad_norm": 0.16436360776424408,
      "learning_rate": 6.53204e-05,
      "loss": 0.2479,
      "step": 336700
    },
    {
      "epoch": 0.4336,
      "grad_norm": 0.20555828511714935,
      "learning_rate": 6.528040000000001e-05,
      "loss": 0.2475,
      "step": 336800
    },
    {
      "epoch": 0.4338,
      "grad_norm": 0.2154914289712906,
      "learning_rate": 6.52404e-05,
      "loss": 0.2488,
      "step": 336900
    },
    {
      "epoch": 0.434,
      "grad_norm": 0.18624047935009003,
      "learning_rate": 6.52004e-05,
      "loss": 0.2405,
      "step": 337000
    },
    {
      "epoch": 0.4342,
      "grad_norm": 0.19469185173511505,
      "learning_rate": 6.51604e-05,
      "loss": 0.2421,
      "step": 337100
    },
    {
      "epoch": 0.4344,
      "grad_norm": 0.22191110253334045,
      "learning_rate": 6.51204e-05,
      "loss": 0.2495,
      "step": 337200
    },
    {
      "epoch": 0.4346,
      "grad_norm": 0.15489134192466736,
      "learning_rate": 6.508040000000001e-05,
      "loss": 0.2424,
      "step": 337300
    },
    {
      "epoch": 0.4348,
      "grad_norm": 0.17435604333877563,
      "learning_rate": 6.50404e-05,
      "loss": 0.2435,
      "step": 337400
    },
    {
      "epoch": 0.435,
      "grad_norm": 0.2414373904466629,
      "learning_rate": 6.500040000000001e-05,
      "loss": 0.2406,
      "step": 337500
    },
    {
      "epoch": 0.4352,
      "grad_norm": 0.15996411442756653,
      "learning_rate": 6.49604e-05,
      "loss": 0.2411,
      "step": 337600
    },
    {
      "epoch": 0.4354,
      "grad_norm": 0.20958919823169708,
      "learning_rate": 6.492040000000001e-05,
      "loss": 0.2454,
      "step": 337700
    },
    {
      "epoch": 0.4356,
      "grad_norm": 0.20292189717292786,
      "learning_rate": 6.48804e-05,
      "loss": 0.2455,
      "step": 337800
    },
    {
      "epoch": 0.4358,
      "grad_norm": 0.15378180146217346,
      "learning_rate": 6.48404e-05,
      "loss": 0.2476,
      "step": 337900
    },
    {
      "epoch": 0.436,
      "grad_norm": 0.20952782034873962,
      "learning_rate": 6.48004e-05,
      "loss": 0.2502,
      "step": 338000
    },
    {
      "epoch": 0.4362,
      "grad_norm": 0.17655488848686218,
      "learning_rate": 6.47604e-05,
      "loss": 0.243,
      "step": 338100
    },
    {
      "epoch": 0.4364,
      "grad_norm": 0.18256178498268127,
      "learning_rate": 6.47204e-05,
      "loss": 0.2421,
      "step": 338200
    },
    {
      "epoch": 0.4366,
      "grad_norm": 0.23415111005306244,
      "learning_rate": 6.468040000000001e-05,
      "loss": 0.2425,
      "step": 338300
    },
    {
      "epoch": 0.4368,
      "grad_norm": 0.1419600397348404,
      "learning_rate": 6.46404e-05,
      "loss": 0.247,
      "step": 338400
    },
    {
      "epoch": 0.437,
      "grad_norm": 0.18709716200828552,
      "learning_rate": 6.46004e-05,
      "loss": 0.2456,
      "step": 338500
    },
    {
      "epoch": 0.4372,
      "grad_norm": 0.1568467766046524,
      "learning_rate": 6.456039999999999e-05,
      "loss": 0.2381,
      "step": 338600
    },
    {
      "epoch": 0.4374,
      "grad_norm": 0.17499247193336487,
      "learning_rate": 6.45204e-05,
      "loss": 0.2415,
      "step": 338700
    },
    {
      "epoch": 0.4376,
      "grad_norm": 0.23707623779773712,
      "learning_rate": 6.448040000000001e-05,
      "loss": 0.2548,
      "step": 338800
    },
    {
      "epoch": 0.4378,
      "grad_norm": 0.3676884174346924,
      "learning_rate": 6.44404e-05,
      "loss": 0.2466,
      "step": 338900
    },
    {
      "epoch": 0.438,
      "grad_norm": 0.21501903235912323,
      "learning_rate": 6.440040000000001e-05,
      "loss": 0.2501,
      "step": 339000
    },
    {
      "epoch": 0.4382,
      "grad_norm": 0.19455668330192566,
      "learning_rate": 6.43604e-05,
      "loss": 0.2418,
      "step": 339100
    },
    {
      "epoch": 0.4384,
      "grad_norm": 0.24228647351264954,
      "learning_rate": 6.432040000000001e-05,
      "loss": 0.2399,
      "step": 339200
    },
    {
      "epoch": 0.4386,
      "grad_norm": 0.16758756339550018,
      "learning_rate": 6.42804e-05,
      "loss": 0.2457,
      "step": 339300
    },
    {
      "epoch": 0.4388,
      "grad_norm": 0.24889026582241058,
      "learning_rate": 6.42404e-05,
      "loss": 0.2414,
      "step": 339400
    },
    {
      "epoch": 0.439,
      "grad_norm": 0.2294233739376068,
      "learning_rate": 6.42004e-05,
      "loss": 0.2416,
      "step": 339500
    },
    {
      "epoch": 0.4392,
      "grad_norm": 0.24796201288700104,
      "learning_rate": 6.41604e-05,
      "loss": 0.2407,
      "step": 339600
    },
    {
      "epoch": 0.4394,
      "grad_norm": 0.16139519214630127,
      "learning_rate": 6.41204e-05,
      "loss": 0.2453,
      "step": 339700
    },
    {
      "epoch": 0.4396,
      "grad_norm": 0.1596374362707138,
      "learning_rate": 6.408040000000001e-05,
      "loss": 0.2394,
      "step": 339800
    },
    {
      "epoch": 0.4398,
      "grad_norm": 0.18892623484134674,
      "learning_rate": 6.40404e-05,
      "loss": 0.2465,
      "step": 339900
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.16438473761081696,
      "learning_rate": 6.400040000000001e-05,
      "loss": 0.2437,
      "step": 340000
    },
    {
      "epoch": 0.4402,
      "grad_norm": 0.2160944789648056,
      "learning_rate": 6.396039999999999e-05,
      "loss": 0.2451,
      "step": 340100
    },
    {
      "epoch": 0.4404,
      "grad_norm": 0.1667250096797943,
      "learning_rate": 6.39204e-05,
      "loss": 0.245,
      "step": 340200
    },
    {
      "epoch": 0.4406,
      "grad_norm": 0.15988707542419434,
      "learning_rate": 6.38804e-05,
      "loss": 0.2431,
      "step": 340300
    },
    {
      "epoch": 0.4408,
      "grad_norm": 0.1956913322210312,
      "learning_rate": 6.38404e-05,
      "loss": 0.2461,
      "step": 340400
    },
    {
      "epoch": 0.441,
      "grad_norm": 0.3029384911060333,
      "learning_rate": 6.38004e-05,
      "loss": 0.2427,
      "step": 340500
    },
    {
      "epoch": 0.4412,
      "grad_norm": 0.1784927248954773,
      "learning_rate": 6.37604e-05,
      "loss": 0.2403,
      "step": 340600
    },
    {
      "epoch": 0.4414,
      "grad_norm": 0.18993079662322998,
      "learning_rate": 6.372040000000001e-05,
      "loss": 0.2453,
      "step": 340700
    },
    {
      "epoch": 0.4416,
      "grad_norm": 0.17982354760169983,
      "learning_rate": 6.36804e-05,
      "loss": 0.2386,
      "step": 340800
    },
    {
      "epoch": 0.4418,
      "grad_norm": 0.15973594784736633,
      "learning_rate": 6.36404e-05,
      "loss": 0.2412,
      "step": 340900
    },
    {
      "epoch": 0.442,
      "grad_norm": 0.2465885430574417,
      "learning_rate": 6.36004e-05,
      "loss": 0.2431,
      "step": 341000
    },
    {
      "epoch": 0.4422,
      "grad_norm": 0.2246682345867157,
      "learning_rate": 6.35604e-05,
      "loss": 0.2437,
      "step": 341100
    },
    {
      "epoch": 0.4424,
      "grad_norm": 0.2625156342983246,
      "learning_rate": 6.35204e-05,
      "loss": 0.2448,
      "step": 341200
    },
    {
      "epoch": 0.4426,
      "grad_norm": 0.14324608445167542,
      "learning_rate": 6.348040000000001e-05,
      "loss": 0.2398,
      "step": 341300
    },
    {
      "epoch": 0.4428,
      "grad_norm": 0.1656043529510498,
      "learning_rate": 6.34404e-05,
      "loss": 0.2607,
      "step": 341400
    },
    {
      "epoch": 0.443,
      "grad_norm": 0.2322763055562973,
      "learning_rate": 6.340040000000001e-05,
      "loss": 0.2432,
      "step": 341500
    },
    {
      "epoch": 0.4432,
      "grad_norm": 0.2581785321235657,
      "learning_rate": 6.33604e-05,
      "loss": 0.2425,
      "step": 341600
    },
    {
      "epoch": 0.4434,
      "grad_norm": 0.18727286159992218,
      "learning_rate": 6.33204e-05,
      "loss": 0.2418,
      "step": 341700
    },
    {
      "epoch": 0.4436,
      "grad_norm": 0.2105150818824768,
      "learning_rate": 6.32804e-05,
      "loss": 0.2397,
      "step": 341800
    },
    {
      "epoch": 0.4438,
      "grad_norm": 0.17236988246440887,
      "learning_rate": 6.32404e-05,
      "loss": 0.2437,
      "step": 341900
    },
    {
      "epoch": 0.444,
      "grad_norm": 0.28422054648399353,
      "learning_rate": 6.32004e-05,
      "loss": 0.2487,
      "step": 342000
    },
    {
      "epoch": 0.4442,
      "grad_norm": 0.1856316179037094,
      "learning_rate": 6.316040000000001e-05,
      "loss": 0.2402,
      "step": 342100
    },
    {
      "epoch": 0.4444,
      "grad_norm": 0.17778094112873077,
      "learning_rate": 6.31204e-05,
      "loss": 0.2501,
      "step": 342200
    },
    {
      "epoch": 0.4446,
      "grad_norm": 0.2376558780670166,
      "learning_rate": 6.30804e-05,
      "loss": 0.2442,
      "step": 342300
    },
    {
      "epoch": 0.4448,
      "grad_norm": 0.20546750724315643,
      "learning_rate": 6.30404e-05,
      "loss": 0.2476,
      "step": 342400
    },
    {
      "epoch": 0.445,
      "grad_norm": 0.17325349152088165,
      "learning_rate": 6.30004e-05,
      "loss": 0.2413,
      "step": 342500
    },
    {
      "epoch": 0.4452,
      "grad_norm": 0.23911508917808533,
      "learning_rate": 6.296040000000001e-05,
      "loss": 0.2446,
      "step": 342600
    },
    {
      "epoch": 0.4454,
      "grad_norm": 0.22123664617538452,
      "learning_rate": 6.29204e-05,
      "loss": 0.2441,
      "step": 342700
    },
    {
      "epoch": 0.4456,
      "grad_norm": 0.21504764258861542,
      "learning_rate": 6.288040000000001e-05,
      "loss": 0.2502,
      "step": 342800
    },
    {
      "epoch": 0.4458,
      "grad_norm": 0.16267873346805573,
      "learning_rate": 6.28404e-05,
      "loss": 0.243,
      "step": 342900
    },
    {
      "epoch": 0.446,
      "grad_norm": 0.19336003065109253,
      "learning_rate": 6.280040000000001e-05,
      "loss": 0.2497,
      "step": 343000
    },
    {
      "epoch": 0.4462,
      "grad_norm": 0.18519088625907898,
      "learning_rate": 6.27604e-05,
      "loss": 0.2469,
      "step": 343100
    },
    {
      "epoch": 0.4464,
      "grad_norm": 0.18827581405639648,
      "learning_rate": 6.27204e-05,
      "loss": 0.2425,
      "step": 343200
    },
    {
      "epoch": 0.4466,
      "grad_norm": 0.24695651233196259,
      "learning_rate": 6.26804e-05,
      "loss": 0.2406,
      "step": 343300
    },
    {
      "epoch": 0.4468,
      "grad_norm": 0.22703491151332855,
      "learning_rate": 6.26404e-05,
      "loss": 0.2466,
      "step": 343400
    },
    {
      "epoch": 0.447,
      "grad_norm": 0.16678479313850403,
      "learning_rate": 6.26004e-05,
      "loss": 0.2431,
      "step": 343500
    },
    {
      "epoch": 0.4472,
      "grad_norm": 0.16149210929870605,
      "learning_rate": 6.256040000000001e-05,
      "loss": 0.2447,
      "step": 343600
    },
    {
      "epoch": 0.4474,
      "grad_norm": 0.21965859830379486,
      "learning_rate": 6.25204e-05,
      "loss": 0.243,
      "step": 343700
    },
    {
      "epoch": 0.4476,
      "grad_norm": 0.21483761072158813,
      "learning_rate": 6.24804e-05,
      "loss": 0.2437,
      "step": 343800
    },
    {
      "epoch": 0.4478,
      "grad_norm": 0.1887097805738449,
      "learning_rate": 6.244039999999999e-05,
      "loss": 0.2383,
      "step": 343900
    },
    {
      "epoch": 0.448,
      "grad_norm": 0.23063704371452332,
      "learning_rate": 6.24004e-05,
      "loss": 0.2469,
      "step": 344000
    },
    {
      "epoch": 0.4482,
      "grad_norm": 0.18795616924762726,
      "learning_rate": 6.236040000000001e-05,
      "loss": 0.2427,
      "step": 344100
    },
    {
      "epoch": 0.4484,
      "grad_norm": 0.26034119725227356,
      "learning_rate": 6.23204e-05,
      "loss": 0.2457,
      "step": 344200
    },
    {
      "epoch": 0.4486,
      "grad_norm": 0.16719482839107513,
      "learning_rate": 6.228040000000001e-05,
      "loss": 0.2446,
      "step": 344300
    },
    {
      "epoch": 0.4488,
      "grad_norm": 0.2299802452325821,
      "learning_rate": 6.22404e-05,
      "loss": 0.2423,
      "step": 344400
    },
    {
      "epoch": 0.449,
      "grad_norm": 0.15124650299549103,
      "learning_rate": 6.220040000000001e-05,
      "loss": 0.249,
      "step": 344500
    },
    {
      "epoch": 0.4492,
      "grad_norm": 0.27996626496315,
      "learning_rate": 6.21604e-05,
      "loss": 0.2646,
      "step": 344600
    },
    {
      "epoch": 0.4494,
      "grad_norm": 0.18617983162403107,
      "learning_rate": 6.21204e-05,
      "loss": 0.2458,
      "step": 344700
    },
    {
      "epoch": 0.4496,
      "grad_norm": 0.1682778298854828,
      "learning_rate": 6.20804e-05,
      "loss": 0.2454,
      "step": 344800
    },
    {
      "epoch": 0.4498,
      "grad_norm": 0.212728813290596,
      "learning_rate": 6.20404e-05,
      "loss": 0.2453,
      "step": 344900
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.2137337177991867,
      "learning_rate": 6.20004e-05,
      "loss": 0.2421,
      "step": 345000
    },
    {
      "epoch": 0.4502,
      "grad_norm": 0.18748146295547485,
      "learning_rate": 6.196040000000001e-05,
      "loss": 0.2427,
      "step": 345100
    },
    {
      "epoch": 0.4504,
      "grad_norm": 0.19029875099658966,
      "learning_rate": 6.19204e-05,
      "loss": 0.2437,
      "step": 345200
    },
    {
      "epoch": 0.4506,
      "grad_norm": 0.380691260099411,
      "learning_rate": 6.188040000000001e-05,
      "loss": 0.2418,
      "step": 345300
    },
    {
      "epoch": 0.4508,
      "grad_norm": 0.1864905059337616,
      "learning_rate": 6.18404e-05,
      "loss": 0.2442,
      "step": 345400
    },
    {
      "epoch": 0.451,
      "grad_norm": 0.1725115031003952,
      "learning_rate": 6.18004e-05,
      "loss": 0.2473,
      "step": 345500
    },
    {
      "epoch": 0.4512,
      "grad_norm": 0.2126380056142807,
      "learning_rate": 6.17604e-05,
      "loss": 0.246,
      "step": 345600
    },
    {
      "epoch": 0.4514,
      "grad_norm": 0.16333705186843872,
      "learning_rate": 6.17204e-05,
      "loss": 0.2432,
      "step": 345700
    },
    {
      "epoch": 0.4516,
      "grad_norm": 0.20585864782333374,
      "learning_rate": 6.16804e-05,
      "loss": 0.2407,
      "step": 345800
    },
    {
      "epoch": 0.4518,
      "grad_norm": 0.1900789737701416,
      "learning_rate": 6.16404e-05,
      "loss": 0.2418,
      "step": 345900
    },
    {
      "epoch": 0.452,
      "grad_norm": 0.1498749703168869,
      "learning_rate": 6.160040000000001e-05,
      "loss": 0.2478,
      "step": 346000
    },
    {
      "epoch": 0.4522,
      "grad_norm": 0.15842682123184204,
      "learning_rate": 6.15604e-05,
      "loss": 0.2416,
      "step": 346100
    },
    {
      "epoch": 0.4524,
      "grad_norm": 0.24384468793869019,
      "learning_rate": 6.15204e-05,
      "loss": 0.2416,
      "step": 346200
    },
    {
      "epoch": 0.4526,
      "grad_norm": 0.16510146856307983,
      "learning_rate": 6.14804e-05,
      "loss": 0.2414,
      "step": 346300
    },
    {
      "epoch": 0.4528,
      "grad_norm": 0.21207810938358307,
      "learning_rate": 6.14404e-05,
      "loss": 0.2439,
      "step": 346400
    },
    {
      "epoch": 0.453,
      "grad_norm": 0.1559634655714035,
      "learning_rate": 6.14004e-05,
      "loss": 0.2393,
      "step": 346500
    },
    {
      "epoch": 0.4532,
      "grad_norm": 0.16685090959072113,
      "learning_rate": 6.136040000000001e-05,
      "loss": 0.2437,
      "step": 346600
    },
    {
      "epoch": 0.4534,
      "grad_norm": 0.1927775889635086,
      "learning_rate": 6.13204e-05,
      "loss": 0.2431,
      "step": 346700
    },
    {
      "epoch": 0.4536,
      "grad_norm": 0.15297745168209076,
      "learning_rate": 6.128040000000001e-05,
      "loss": 0.2469,
      "step": 346800
    },
    {
      "epoch": 0.4538,
      "grad_norm": 0.20579373836517334,
      "learning_rate": 6.12404e-05,
      "loss": 0.2411,
      "step": 346900
    },
    {
      "epoch": 0.454,
      "grad_norm": 0.1857280731201172,
      "learning_rate": 6.12004e-05,
      "loss": 0.2392,
      "step": 347000
    },
    {
      "epoch": 0.4542,
      "grad_norm": 0.3481803238391876,
      "learning_rate": 6.11604e-05,
      "loss": 0.2442,
      "step": 347100
    },
    {
      "epoch": 0.4544,
      "grad_norm": 0.18435798585414886,
      "learning_rate": 6.11204e-05,
      "loss": 0.2433,
      "step": 347200
    },
    {
      "epoch": 0.4546,
      "grad_norm": 0.18237179517745972,
      "learning_rate": 6.10804e-05,
      "loss": 0.2436,
      "step": 347300
    },
    {
      "epoch": 0.4548,
      "grad_norm": 0.16900283098220825,
      "learning_rate": 6.10404e-05,
      "loss": 0.242,
      "step": 347400
    },
    {
      "epoch": 0.455,
      "grad_norm": 0.19602380692958832,
      "learning_rate": 6.10004e-05,
      "loss": 0.2421,
      "step": 347500
    },
    {
      "epoch": 0.4552,
      "grad_norm": 0.2622341513633728,
      "learning_rate": 6.096040000000001e-05,
      "loss": 0.2415,
      "step": 347600
    },
    {
      "epoch": 0.4554,
      "grad_norm": 0.3507973253726959,
      "learning_rate": 6.09204e-05,
      "loss": 0.2405,
      "step": 347700
    },
    {
      "epoch": 0.4556,
      "grad_norm": 0.2798367738723755,
      "learning_rate": 6.08804e-05,
      "loss": 0.2578,
      "step": 347800
    },
    {
      "epoch": 0.4558,
      "grad_norm": 0.1859496533870697,
      "learning_rate": 6.0840399999999994e-05,
      "loss": 0.2413,
      "step": 347900
    },
    {
      "epoch": 0.456,
      "grad_norm": 0.1584465652704239,
      "learning_rate": 6.08004e-05,
      "loss": 0.2389,
      "step": 348000
    },
    {
      "epoch": 0.4562,
      "grad_norm": 0.18819481134414673,
      "learning_rate": 6.076040000000001e-05,
      "loss": 0.2461,
      "step": 348100
    },
    {
      "epoch": 0.4564,
      "grad_norm": 0.16603697836399078,
      "learning_rate": 6.07204e-05,
      "loss": 0.2412,
      "step": 348200
    },
    {
      "epoch": 0.4566,
      "grad_norm": 0.17016661167144775,
      "learning_rate": 6.06804e-05,
      "loss": 0.2414,
      "step": 348300
    },
    {
      "epoch": 0.4568,
      "grad_norm": 0.21648702025413513,
      "learning_rate": 6.064040000000001e-05,
      "loss": 0.2391,
      "step": 348400
    },
    {
      "epoch": 0.457,
      "grad_norm": 0.1657271832227707,
      "learning_rate": 6.0600400000000003e-05,
      "loss": 0.2437,
      "step": 348500
    },
    {
      "epoch": 0.4572,
      "grad_norm": 0.19239091873168945,
      "learning_rate": 6.0560400000000004e-05,
      "loss": 0.2363,
      "step": 348600
    },
    {
      "epoch": 0.4574,
      "grad_norm": 0.2200000286102295,
      "learning_rate": 6.05204e-05,
      "loss": 0.2412,
      "step": 348700
    },
    {
      "epoch": 0.4576,
      "grad_norm": 0.18818236887454987,
      "learning_rate": 6.0480400000000005e-05,
      "loss": 0.2421,
      "step": 348800
    },
    {
      "epoch": 0.4578,
      "grad_norm": 0.18705151975154877,
      "learning_rate": 6.0440400000000005e-05,
      "loss": 0.2396,
      "step": 348900
    },
    {
      "epoch": 0.458,
      "grad_norm": 0.19457854330539703,
      "learning_rate": 6.04004e-05,
      "loss": 0.2392,
      "step": 349000
    },
    {
      "epoch": 0.4582,
      "grad_norm": 0.1518254578113556,
      "learning_rate": 6.0360400000000006e-05,
      "loss": 0.2497,
      "step": 349100
    },
    {
      "epoch": 0.4584,
      "grad_norm": 0.21858882904052734,
      "learning_rate": 6.03204e-05,
      "loss": 0.2376,
      "step": 349200
    },
    {
      "epoch": 0.4586,
      "grad_norm": 0.24455934762954712,
      "learning_rate": 6.0280400000000006e-05,
      "loss": 0.2364,
      "step": 349300
    },
    {
      "epoch": 0.4588,
      "grad_norm": 0.26377052068710327,
      "learning_rate": 6.024040000000001e-05,
      "loss": 0.2425,
      "step": 349400
    },
    {
      "epoch": 0.459,
      "grad_norm": 0.2012515366077423,
      "learning_rate": 6.02004e-05,
      "loss": 0.2395,
      "step": 349500
    },
    {
      "epoch": 0.4592,
      "grad_norm": 0.1855122596025467,
      "learning_rate": 6.016040000000001e-05,
      "loss": 0.2405,
      "step": 349600
    },
    {
      "epoch": 0.4594,
      "grad_norm": 0.17877699434757233,
      "learning_rate": 6.01204e-05,
      "loss": 0.2417,
      "step": 349700
    },
    {
      "epoch": 0.4596,
      "grad_norm": 0.20735540986061096,
      "learning_rate": 6.00804e-05,
      "loss": 0.244,
      "step": 349800
    },
    {
      "epoch": 0.4598,
      "grad_norm": 0.15910696983337402,
      "learning_rate": 6.004040000000001e-05,
      "loss": 0.2403,
      "step": 349900
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.16688168048858643,
      "learning_rate": 6.00004e-05,
      "loss": 0.241,
      "step": 350000
    },
    {
      "epoch": 0.0002,
      "grad_norm": 0.18989884853363037,
      "learning_rate": 5.99604e-05,
      "loss": 0.2474,
      "step": 350100
    },
    {
      "epoch": 0.0004,
      "grad_norm": 0.39203211665153503,
      "learning_rate": 5.9920399999999996e-05,
      "loss": 0.2589,
      "step": 350200
    },
    {
      "epoch": 0.0006,
      "grad_norm": 0.4958038628101349,
      "learning_rate": 5.9880400000000004e-05,
      "loss": 0.2596,
      "step": 350300
    },
    {
      "epoch": 0.0008,
      "grad_norm": 0.2854653596878052,
      "learning_rate": 5.984040000000001e-05,
      "loss": 0.2638,
      "step": 350400
    },
    {
      "epoch": 0.001,
      "grad_norm": 0.1860920786857605,
      "learning_rate": 5.98004e-05,
      "loss": 0.2509,
      "step": 350500
    },
    {
      "epoch": 0.0012,
      "grad_norm": 0.41365325450897217,
      "learning_rate": 5.9760400000000005e-05,
      "loss": 0.2487,
      "step": 350600
    },
    {
      "epoch": 0.0014,
      "grad_norm": 0.35434776544570923,
      "learning_rate": 5.97204e-05,
      "loss": 0.2528,
      "step": 350700
    },
    {
      "epoch": 0.0016,
      "grad_norm": 0.3473292887210846,
      "learning_rate": 5.9680400000000005e-05,
      "loss": 0.2534,
      "step": 350800
    },
    {
      "epoch": 0.0018,
      "grad_norm": 0.3211774528026581,
      "learning_rate": 5.9640400000000006e-05,
      "loss": 0.2452,
      "step": 350900
    },
    {
      "epoch": 0.002,
      "grad_norm": 0.32706430554389954,
      "learning_rate": 5.96004e-05,
      "loss": 0.2444,
      "step": 351000
    },
    {
      "epoch": 0.0022,
      "grad_norm": 0.3087383210659027,
      "learning_rate": 5.9560400000000007e-05,
      "loss": 0.2609,
      "step": 351100
    },
    {
      "epoch": 0.0024,
      "grad_norm": 0.3712453246116638,
      "learning_rate": 5.95204e-05,
      "loss": 0.2478,
      "step": 351200
    },
    {
      "epoch": 0.0026,
      "grad_norm": 0.5335580706596375,
      "learning_rate": 5.94804e-05,
      "loss": 0.2533,
      "step": 351300
    },
    {
      "epoch": 0.0028,
      "grad_norm": 0.14397287368774414,
      "learning_rate": 5.944040000000001e-05,
      "loss": 0.2397,
      "step": 351400
    },
    {
      "epoch": 0.003,
      "grad_norm": 0.3205547332763672,
      "learning_rate": 5.94004e-05,
      "loss": 0.2501,
      "step": 351500
    },
    {
      "epoch": 0.0032,
      "grad_norm": 0.29998424649238586,
      "learning_rate": 5.93604e-05,
      "loss": 0.2614,
      "step": 351600
    },
    {
      "epoch": 0.0034,
      "grad_norm": 0.8002346754074097,
      "learning_rate": 5.9320399999999995e-05,
      "loss": 0.2461,
      "step": 351700
    },
    {
      "epoch": 0.0036,
      "grad_norm": 0.23687268793582916,
      "learning_rate": 5.92804e-05,
      "loss": 0.2502,
      "step": 351800
    },
    {
      "epoch": 0.0038,
      "grad_norm": 0.15750548243522644,
      "learning_rate": 5.924040000000001e-05,
      "loss": 0.2511,
      "step": 351900
    },
    {
      "epoch": 0.004,
      "grad_norm": 0.3123827576637268,
      "learning_rate": 5.92004e-05,
      "loss": 0.2481,
      "step": 352000
    },
    {
      "epoch": 0.0042,
      "grad_norm": 0.2805723249912262,
      "learning_rate": 5.9160400000000004e-05,
      "loss": 0.2508,
      "step": 352100
    },
    {
      "epoch": 0.0044,
      "grad_norm": 0.1956070512533188,
      "learning_rate": 5.91204e-05,
      "loss": 0.2677,
      "step": 352200
    },
    {
      "epoch": 0.0046,
      "grad_norm": 0.2959635853767395,
      "learning_rate": 5.9080400000000004e-05,
      "loss": 0.2544,
      "step": 352300
    },
    {
      "epoch": 0.0048,
      "grad_norm": 0.23278026282787323,
      "learning_rate": 5.9040400000000005e-05,
      "loss": 0.2426,
      "step": 352400
    },
    {
      "epoch": 0.005,
      "grad_norm": 0.25376245379447937,
      "learning_rate": 5.90004e-05,
      "loss": 0.2397,
      "step": 352500
    },
    {
      "epoch": 0.0052,
      "grad_norm": 0.2448563426733017,
      "learning_rate": 5.8960400000000006e-05,
      "loss": 0.2389,
      "step": 352600
    },
    {
      "epoch": 0.0054,
      "grad_norm": 0.3233645260334015,
      "learning_rate": 5.89204e-05,
      "loss": 0.2447,
      "step": 352700
    },
    {
      "epoch": 0.0056,
      "grad_norm": 0.2815536558628082,
      "learning_rate": 5.88804e-05,
      "loss": 0.2547,
      "step": 352800
    },
    {
      "epoch": 0.0058,
      "grad_norm": 0.30052757263183594,
      "learning_rate": 5.884040000000001e-05,
      "loss": 0.2399,
      "step": 352900
    },
    {
      "epoch": 0.006,
      "grad_norm": 0.26953259110450745,
      "learning_rate": 5.88004e-05,
      "loss": 0.292,
      "step": 353000
    },
    {
      "epoch": 0.0062,
      "grad_norm": 0.17700572311878204,
      "learning_rate": 5.876040000000001e-05,
      "loss": 0.2503,
      "step": 353100
    },
    {
      "epoch": 0.0064,
      "grad_norm": 0.15203849971294403,
      "learning_rate": 5.8720399999999994e-05,
      "loss": 0.2459,
      "step": 353200
    },
    {
      "epoch": 0.0066,
      "grad_norm": 0.6238852143287659,
      "learning_rate": 5.86804e-05,
      "loss": 0.2607,
      "step": 353300
    },
    {
      "epoch": 0.0068,
      "grad_norm": 0.18438929319381714,
      "learning_rate": 5.864040000000001e-05,
      "loss": 0.245,
      "step": 353400
    },
    {
      "epoch": 0.007,
      "grad_norm": 0.18198923766613007,
      "learning_rate": 5.86004e-05,
      "loss": 0.255,
      "step": 353500
    },
    {
      "epoch": 0.0072,
      "grad_norm": 0.1955738663673401,
      "learning_rate": 5.85604e-05,
      "loss": 0.2535,
      "step": 353600
    },
    {
      "epoch": 0.0074,
      "grad_norm": 0.22063769400119781,
      "learning_rate": 5.8520399999999996e-05,
      "loss": 0.2598,
      "step": 353700
    },
    {
      "epoch": 0.0076,
      "grad_norm": 0.15472210943698883,
      "learning_rate": 5.8480400000000003e-05,
      "loss": 0.2614,
      "step": 353800
    },
    {
      "epoch": 0.0078,
      "grad_norm": 0.19167518615722656,
      "learning_rate": 5.8440400000000004e-05,
      "loss": 0.2476,
      "step": 353900
    },
    {
      "epoch": 0.008,
      "grad_norm": 0.2191273272037506,
      "learning_rate": 5.84004e-05,
      "loss": 0.2533,
      "step": 354000
    },
    {
      "epoch": 0.0082,
      "grad_norm": 0.3382338285446167,
      "learning_rate": 5.8360400000000005e-05,
      "loss": 0.2611,
      "step": 354100
    },
    {
      "epoch": 0.0084,
      "grad_norm": 0.559860110282898,
      "learning_rate": 5.83204e-05,
      "loss": 0.2478,
      "step": 354200
    },
    {
      "epoch": 0.0086,
      "grad_norm": 0.1866038739681244,
      "learning_rate": 5.82804e-05,
      "loss": 0.2457,
      "step": 354300
    },
    {
      "epoch": 0.0088,
      "grad_norm": 0.21642564237117767,
      "learning_rate": 5.8240400000000006e-05,
      "loss": 0.246,
      "step": 354400
    },
    {
      "epoch": 0.009,
      "grad_norm": 0.23860223591327667,
      "learning_rate": 5.82004e-05,
      "loss": 0.2595,
      "step": 354500
    },
    {
      "epoch": 0.0092,
      "grad_norm": 0.21850599348545074,
      "learning_rate": 5.8160400000000006e-05,
      "loss": 0.2626,
      "step": 354600
    },
    {
      "epoch": 0.0094,
      "grad_norm": 0.2536315619945526,
      "learning_rate": 5.812040000000001e-05,
      "loss": 0.2551,
      "step": 354700
    },
    {
      "epoch": 0.0096,
      "grad_norm": 0.22218576073646545,
      "learning_rate": 5.80804e-05,
      "loss": 0.2647,
      "step": 354800
    },
    {
      "epoch": 0.0098,
      "grad_norm": 0.1981973648071289,
      "learning_rate": 5.804040000000001e-05,
      "loss": 0.2608,
      "step": 354900
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.23694828152656555,
      "learning_rate": 5.80004e-05,
      "loss": 0.2562,
      "step": 355000
    },
    {
      "epoch": 0.0102,
      "grad_norm": 0.22222231328487396,
      "learning_rate": 5.79604e-05,
      "loss": 0.2607,
      "step": 355100
    },
    {
      "epoch": 0.0104,
      "grad_norm": 0.26125141978263855,
      "learning_rate": 5.792040000000001e-05,
      "loss": 0.2451,
      "step": 355200
    },
    {
      "epoch": 0.0106,
      "grad_norm": 0.2317095398902893,
      "learning_rate": 5.78804e-05,
      "loss": 0.2566,
      "step": 355300
    },
    {
      "epoch": 0.0108,
      "grad_norm": 0.24477675557136536,
      "learning_rate": 5.78404e-05,
      "loss": 0.2598,
      "step": 355400
    },
    {
      "epoch": 0.011,
      "grad_norm": 0.2259080708026886,
      "learning_rate": 5.7800399999999996e-05,
      "loss": 0.2576,
      "step": 355500
    },
    {
      "epoch": 0.0112,
      "grad_norm": 0.20323753356933594,
      "learning_rate": 5.7760400000000004e-05,
      "loss": 0.2531,
      "step": 355600
    },
    {
      "epoch": 0.0114,
      "grad_norm": 0.2604706883430481,
      "learning_rate": 5.772040000000001e-05,
      "loss": 0.2474,
      "step": 355700
    },
    {
      "epoch": 0.0116,
      "grad_norm": 0.2472306191921234,
      "learning_rate": 5.7680400000000004e-05,
      "loss": 0.2574,
      "step": 355800
    },
    {
      "epoch": 0.0118,
      "grad_norm": 0.21734565496444702,
      "learning_rate": 5.7640400000000005e-05,
      "loss": 0.2589,
      "step": 355900
    },
    {
      "epoch": 0.012,
      "grad_norm": 0.1756725013256073,
      "learning_rate": 5.76004e-05,
      "loss": 0.25,
      "step": 356000
    },
    {
      "epoch": 0.0122,
      "grad_norm": 0.2940695881843567,
      "learning_rate": 5.7560400000000005e-05,
      "loss": 0.2797,
      "step": 356100
    },
    {
      "epoch": 0.0124,
      "grad_norm": 0.1878010779619217,
      "learning_rate": 5.7520400000000006e-05,
      "loss": 0.2494,
      "step": 356200
    },
    {
      "epoch": 0.0126,
      "grad_norm": 0.2582753598690033,
      "learning_rate": 5.74804e-05,
      "loss": 0.2584,
      "step": 356300
    },
    {
      "epoch": 0.0128,
      "grad_norm": 0.15801192820072174,
      "learning_rate": 5.7440400000000007e-05,
      "loss": 0.2465,
      "step": 356400
    },
    {
      "epoch": 0.013,
      "grad_norm": 0.26261720061302185,
      "learning_rate": 5.74004e-05,
      "loss": 0.2502,
      "step": 356500
    },
    {
      "epoch": 0.0132,
      "grad_norm": 0.22228971123695374,
      "learning_rate": 5.73604e-05,
      "loss": 0.2445,
      "step": 356600
    },
    {
      "epoch": 0.0134,
      "grad_norm": 0.2221522033214569,
      "learning_rate": 5.732040000000001e-05,
      "loss": 0.2695,
      "step": 356700
    },
    {
      "epoch": 0.0136,
      "grad_norm": 0.24543708562850952,
      "learning_rate": 5.72804e-05,
      "loss": 0.2442,
      "step": 356800
    },
    {
      "epoch": 0.0138,
      "grad_norm": 0.2565612196922302,
      "learning_rate": 5.724040000000001e-05,
      "loss": 0.2623,
      "step": 356900
    },
    {
      "epoch": 0.014,
      "grad_norm": 0.2401857078075409,
      "learning_rate": 5.7200399999999995e-05,
      "loss": 0.2575,
      "step": 357000
    },
    {
      "epoch": 0.0142,
      "grad_norm": 0.18169806897640228,
      "learning_rate": 5.71604e-05,
      "loss": 0.2514,
      "step": 357100
    },
    {
      "epoch": 0.0144,
      "grad_norm": 0.2812741994857788,
      "learning_rate": 5.712040000000001e-05,
      "loss": 0.2513,
      "step": 357200
    },
    {
      "epoch": 0.0146,
      "grad_norm": 0.16843974590301514,
      "learning_rate": 5.70804e-05,
      "loss": 0.2564,
      "step": 357300
    },
    {
      "epoch": 0.0148,
      "grad_norm": 0.31371912360191345,
      "learning_rate": 5.7040400000000004e-05,
      "loss": 0.2476,
      "step": 357400
    },
    {
      "epoch": 0.015,
      "grad_norm": 0.23738493025302887,
      "learning_rate": 5.70004e-05,
      "loss": 0.2573,
      "step": 357500
    },
    {
      "epoch": 0.0152,
      "grad_norm": 0.19511915743350983,
      "learning_rate": 5.6960400000000004e-05,
      "loss": 0.2506,
      "step": 357600
    },
    {
      "epoch": 0.0154,
      "grad_norm": 0.16517697274684906,
      "learning_rate": 5.6920400000000005e-05,
      "loss": 0.2485,
      "step": 357700
    },
    {
      "epoch": 0.0156,
      "grad_norm": 0.19995634257793427,
      "learning_rate": 5.68804e-05,
      "loss": 0.2484,
      "step": 357800
    },
    {
      "epoch": 0.0158,
      "grad_norm": 0.21146655082702637,
      "learning_rate": 5.6840400000000006e-05,
      "loss": 0.2443,
      "step": 357900
    },
    {
      "epoch": 0.016,
      "grad_norm": 0.34297746419906616,
      "learning_rate": 5.68004e-05,
      "loss": 0.2451,
      "step": 358000
    },
    {
      "epoch": 0.0162,
      "grad_norm": 0.20331263542175293,
      "learning_rate": 5.67604e-05,
      "loss": 0.2586,
      "step": 358100
    },
    {
      "epoch": 0.0164,
      "grad_norm": 0.3667886257171631,
      "learning_rate": 5.672040000000001e-05,
      "loss": 0.249,
      "step": 358200
    },
    {
      "epoch": 0.0166,
      "grad_norm": 0.31036943197250366,
      "learning_rate": 5.66804e-05,
      "loss": 0.2457,
      "step": 358300
    },
    {
      "epoch": 0.0168,
      "grad_norm": 0.2320556640625,
      "learning_rate": 5.664040000000001e-05,
      "loss": 0.253,
      "step": 358400
    },
    {
      "epoch": 0.017,
      "grad_norm": 0.23382051289081573,
      "learning_rate": 5.66004e-05,
      "loss": 0.2578,
      "step": 358500
    },
    {
      "epoch": 0.0172,
      "grad_norm": 0.2036983221769333,
      "learning_rate": 5.65604e-05,
      "loss": 0.2636,
      "step": 358600
    },
    {
      "epoch": 0.0174,
      "grad_norm": 0.3016323745250702,
      "learning_rate": 5.652040000000001e-05,
      "loss": 0.2532,
      "step": 358700
    },
    {
      "epoch": 0.0176,
      "grad_norm": 0.1959076225757599,
      "learning_rate": 5.64804e-05,
      "loss": 0.2521,
      "step": 358800
    },
    {
      "epoch": 0.0178,
      "grad_norm": 0.22912850975990295,
      "learning_rate": 5.64404e-05,
      "loss": 0.2401,
      "step": 358900
    },
    {
      "epoch": 0.018,
      "grad_norm": 0.16965632140636444,
      "learning_rate": 5.6400399999999996e-05,
      "loss": 0.2506,
      "step": 359000
    },
    {
      "epoch": 0.0182,
      "grad_norm": 0.17492711544036865,
      "learning_rate": 5.63604e-05,
      "loss": 0.2504,
      "step": 359100
    },
    {
      "epoch": 0.0184,
      "grad_norm": 0.1555829495191574,
      "learning_rate": 5.6320400000000004e-05,
      "loss": 0.254,
      "step": 359200
    },
    {
      "epoch": 0.0186,
      "grad_norm": 0.20478782057762146,
      "learning_rate": 5.62804e-05,
      "loss": 0.245,
      "step": 359300
    },
    {
      "epoch": 0.0188,
      "grad_norm": 0.21451802551746368,
      "learning_rate": 5.6240400000000005e-05,
      "loss": 0.2356,
      "step": 359400
    },
    {
      "epoch": 0.019,
      "grad_norm": 0.4382535517215729,
      "learning_rate": 5.62004e-05,
      "loss": 0.265,
      "step": 359500
    },
    {
      "epoch": 0.0192,
      "grad_norm": 0.20359216630458832,
      "learning_rate": 5.6160400000000005e-05,
      "loss": 0.2408,
      "step": 359600
    },
    {
      "epoch": 0.0194,
      "grad_norm": 0.17969875037670135,
      "learning_rate": 5.6120400000000006e-05,
      "loss": 0.2609,
      "step": 359700
    },
    {
      "epoch": 0.0196,
      "grad_norm": 0.20156468451023102,
      "learning_rate": 5.60804e-05,
      "loss": 0.2464,
      "step": 359800
    },
    {
      "epoch": 0.0198,
      "grad_norm": 0.518825888633728,
      "learning_rate": 5.6040400000000006e-05,
      "loss": 0.255,
      "step": 359900
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.2389851212501526,
      "learning_rate": 5.60004e-05,
      "loss": 0.257,
      "step": 360000
    },
    {
      "epoch": 0.0202,
      "grad_norm": 0.15437252819538116,
      "learning_rate": 5.59604e-05,
      "loss": 0.2511,
      "step": 360100
    },
    {
      "epoch": 0.0204,
      "grad_norm": 0.687829852104187,
      "learning_rate": 5.592040000000001e-05,
      "loss": 0.2643,
      "step": 360200
    },
    {
      "epoch": 0.0206,
      "grad_norm": 0.18927748501300812,
      "learning_rate": 5.58804e-05,
      "loss": 0.2553,
      "step": 360300
    },
    {
      "epoch": 0.0208,
      "grad_norm": 0.1879296749830246,
      "learning_rate": 5.58404e-05,
      "loss": 0.2584,
      "step": 360400
    },
    {
      "epoch": 0.021,
      "grad_norm": 0.2479875683784485,
      "learning_rate": 5.5800399999999995e-05,
      "loss": 0.2374,
      "step": 360500
    },
    {
      "epoch": 0.0212,
      "grad_norm": 0.20449163019657135,
      "learning_rate": 5.57604e-05,
      "loss": 0.2585,
      "step": 360600
    },
    {
      "epoch": 0.0214,
      "grad_norm": 0.18261049687862396,
      "learning_rate": 5.57204e-05,
      "loss": 0.2383,
      "step": 360700
    },
    {
      "epoch": 0.0216,
      "grad_norm": 0.2606537938117981,
      "learning_rate": 5.5680399999999996e-05,
      "loss": 0.2542,
      "step": 360800
    },
    {
      "epoch": 0.0218,
      "grad_norm": 0.18397171795368195,
      "learning_rate": 5.5640400000000003e-05,
      "loss": 0.2565,
      "step": 360900
    },
    {
      "epoch": 0.022,
      "grad_norm": 0.6222195029258728,
      "learning_rate": 5.560040000000001e-05,
      "loss": 0.2505,
      "step": 361000
    },
    {
      "epoch": 0.0222,
      "grad_norm": 0.26597991585731506,
      "learning_rate": 5.5560400000000004e-05,
      "loss": 0.2567,
      "step": 361100
    },
    {
      "epoch": 0.0224,
      "grad_norm": 0.26086699962615967,
      "learning_rate": 5.5520400000000005e-05,
      "loss": 0.2522,
      "step": 361200
    },
    {
      "epoch": 0.0226,
      "grad_norm": 0.3996891379356384,
      "learning_rate": 5.54804e-05,
      "loss": 0.2609,
      "step": 361300
    },
    {
      "epoch": 0.0228,
      "grad_norm": 0.4706418216228485,
      "learning_rate": 5.5440400000000005e-05,
      "loss": 0.2651,
      "step": 361400
    },
    {
      "epoch": 0.023,
      "grad_norm": 0.23967041075229645,
      "learning_rate": 5.5400400000000006e-05,
      "loss": 0.2546,
      "step": 361500
    },
    {
      "epoch": 0.0232,
      "grad_norm": 0.23435647785663605,
      "learning_rate": 5.53604e-05,
      "loss": 0.2494,
      "step": 361600
    },
    {
      "epoch": 0.0234,
      "grad_norm": 0.1719307005405426,
      "learning_rate": 5.5320400000000007e-05,
      "loss": 0.2454,
      "step": 361700
    },
    {
      "epoch": 0.0236,
      "grad_norm": 0.24978755414485931,
      "learning_rate": 5.52804e-05,
      "loss": 0.2624,
      "step": 361800
    },
    {
      "epoch": 0.0238,
      "grad_norm": 0.2097492516040802,
      "learning_rate": 5.52404e-05,
      "loss": 0.2945,
      "step": 361900
    },
    {
      "epoch": 0.024,
      "grad_norm": 0.23128892481327057,
      "learning_rate": 5.520040000000001e-05,
      "loss": 0.2447,
      "step": 362000
    },
    {
      "epoch": 0.0242,
      "grad_norm": 0.2238485962152481,
      "learning_rate": 5.51604e-05,
      "loss": 0.2516,
      "step": 362100
    },
    {
      "epoch": 0.0244,
      "grad_norm": 0.21893225610256195,
      "learning_rate": 5.512040000000001e-05,
      "loss": 0.2568,
      "step": 362200
    },
    {
      "epoch": 0.0246,
      "grad_norm": 0.18265652656555176,
      "learning_rate": 5.5080399999999995e-05,
      "loss": 0.2509,
      "step": 362300
    },
    {
      "epoch": 0.0248,
      "grad_norm": 0.22736512124538422,
      "learning_rate": 5.50404e-05,
      "loss": 0.2554,
      "step": 362400
    },
    {
      "epoch": 0.025,
      "grad_norm": 0.257276326417923,
      "learning_rate": 5.500040000000001e-05,
      "loss": 0.2661,
      "step": 362500
    },
    {
      "epoch": 0.0252,
      "grad_norm": 0.2855454087257385,
      "learning_rate": 5.49604e-05,
      "loss": 0.2708,
      "step": 362600
    },
    {
      "epoch": 0.0254,
      "grad_norm": 0.2235616147518158,
      "learning_rate": 5.4920400000000004e-05,
      "loss": 0.2706,
      "step": 362700
    },
    {
      "epoch": 0.0256,
      "grad_norm": 0.27502375841140747,
      "learning_rate": 5.48804e-05,
      "loss": 0.2533,
      "step": 362800
    },
    {
      "epoch": 0.0258,
      "grad_norm": 0.31357336044311523,
      "learning_rate": 5.4840400000000004e-05,
      "loss": 0.2586,
      "step": 362900
    },
    {
      "epoch": 0.026,
      "grad_norm": 0.22784167528152466,
      "learning_rate": 5.4800400000000005e-05,
      "loss": 0.2924,
      "step": 363000
    },
    {
      "epoch": 0.0262,
      "grad_norm": 0.16127784550189972,
      "learning_rate": 5.47604e-05,
      "loss": 0.2581,
      "step": 363100
    },
    {
      "epoch": 0.0264,
      "grad_norm": 0.23932571709156036,
      "learning_rate": 5.4720400000000006e-05,
      "loss": 0.2612,
      "step": 363200
    },
    {
      "epoch": 0.0266,
      "grad_norm": 0.22151145339012146,
      "learning_rate": 5.46804e-05,
      "loss": 0.2579,
      "step": 363300
    },
    {
      "epoch": 0.0268,
      "grad_norm": 0.2576836347579956,
      "learning_rate": 5.46404e-05,
      "loss": 0.2546,
      "step": 363400
    },
    {
      "epoch": 0.027,
      "grad_norm": 0.2526397407054901,
      "learning_rate": 5.460040000000001e-05,
      "loss": 0.2403,
      "step": 363500
    },
    {
      "epoch": 0.0272,
      "grad_norm": 0.29295867681503296,
      "learning_rate": 5.45604e-05,
      "loss": 0.2351,
      "step": 363600
    },
    {
      "epoch": 0.0274,
      "grad_norm": 0.21946589648723602,
      "learning_rate": 5.452040000000001e-05,
      "loss": 0.2376,
      "step": 363700
    },
    {
      "epoch": 0.0276,
      "grad_norm": 0.34592190384864807,
      "learning_rate": 5.44804e-05,
      "loss": 0.2373,
      "step": 363800
    },
    {
      "epoch": 0.0278,
      "grad_norm": 0.16898639500141144,
      "learning_rate": 5.44404e-05,
      "loss": 0.2348,
      "step": 363900
    },
    {
      "epoch": 0.028,
      "grad_norm": 0.18720662593841553,
      "learning_rate": 5.440040000000001e-05,
      "loss": 0.2453,
      "step": 364000
    },
    {
      "epoch": 0.0282,
      "grad_norm": 0.18894708156585693,
      "learning_rate": 5.43604e-05,
      "loss": 0.2265,
      "step": 364100
    },
    {
      "epoch": 0.0284,
      "grad_norm": 0.24615564942359924,
      "learning_rate": 5.43204e-05,
      "loss": 0.233,
      "step": 364200
    },
    {
      "epoch": 0.0286,
      "grad_norm": 0.18543006479740143,
      "learning_rate": 5.4280399999999996e-05,
      "loss": 0.2448,
      "step": 364300
    },
    {
      "epoch": 0.0288,
      "grad_norm": 0.21889004111289978,
      "learning_rate": 5.42404e-05,
      "loss": 0.2341,
      "step": 364400
    },
    {
      "epoch": 0.029,
      "grad_norm": 0.40268123149871826,
      "learning_rate": 5.4200400000000004e-05,
      "loss": 0.2447,
      "step": 364500
    },
    {
      "epoch": 0.0292,
      "grad_norm": 0.5381551384925842,
      "learning_rate": 5.41604e-05,
      "loss": 0.2338,
      "step": 364600
    },
    {
      "epoch": 0.0294,
      "grad_norm": 0.39877715706825256,
      "learning_rate": 5.4120400000000004e-05,
      "loss": 0.2323,
      "step": 364700
    },
    {
      "epoch": 0.0296,
      "grad_norm": 0.17573441565036774,
      "learning_rate": 5.40804e-05,
      "loss": 0.2382,
      "step": 364800
    },
    {
      "epoch": 0.0298,
      "grad_norm": 0.14901190996170044,
      "learning_rate": 5.4040400000000005e-05,
      "loss": 0.2371,
      "step": 364900
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.22328852117061615,
      "learning_rate": 5.4000400000000006e-05,
      "loss": 0.2417,
      "step": 365000
    },
    {
      "epoch": 0.0302,
      "grad_norm": 0.1952635943889618,
      "learning_rate": 5.39604e-05,
      "loss": 0.2407,
      "step": 365100
    },
    {
      "epoch": 0.0304,
      "grad_norm": 0.1605110615491867,
      "learning_rate": 5.3920400000000006e-05,
      "loss": 0.2352,
      "step": 365200
    },
    {
      "epoch": 0.0306,
      "grad_norm": 0.24076199531555176,
      "learning_rate": 5.38804e-05,
      "loss": 0.2355,
      "step": 365300
    },
    {
      "epoch": 0.0308,
      "grad_norm": 0.269298791885376,
      "learning_rate": 5.38404e-05,
      "loss": 0.2386,
      "step": 365400
    },
    {
      "epoch": 0.031,
      "grad_norm": 0.22119976580142975,
      "learning_rate": 5.380040000000001e-05,
      "loss": 0.2405,
      "step": 365500
    },
    {
      "epoch": 0.0312,
      "grad_norm": 0.21899624168872833,
      "learning_rate": 5.37604e-05,
      "loss": 0.2482,
      "step": 365600
    },
    {
      "epoch": 0.0314,
      "grad_norm": 0.48613792657852173,
      "learning_rate": 5.37204e-05,
      "loss": 0.2388,
      "step": 365700
    },
    {
      "epoch": 0.0316,
      "grad_norm": 0.22591345012187958,
      "learning_rate": 5.3680399999999995e-05,
      "loss": 0.2351,
      "step": 365800
    },
    {
      "epoch": 0.0318,
      "grad_norm": 0.4765823483467102,
      "learning_rate": 5.36404e-05,
      "loss": 0.2387,
      "step": 365900
    },
    {
      "epoch": 0.032,
      "grad_norm": 0.30145007371902466,
      "learning_rate": 5.360040000000001e-05,
      "loss": 0.2419,
      "step": 366000
    },
    {
      "epoch": 0.0322,
      "grad_norm": 0.2979516386985779,
      "learning_rate": 5.3560399999999996e-05,
      "loss": 0.244,
      "step": 366100
    },
    {
      "epoch": 0.0324,
      "grad_norm": 0.23975351452827454,
      "learning_rate": 5.3520400000000003e-05,
      "loss": 0.2393,
      "step": 366200
    },
    {
      "epoch": 0.0326,
      "grad_norm": 0.22233131527900696,
      "learning_rate": 5.34804e-05,
      "loss": 0.2422,
      "step": 366300
    },
    {
      "epoch": 0.0328,
      "grad_norm": 0.1476048082113266,
      "learning_rate": 5.3440400000000004e-05,
      "loss": 0.2493,
      "step": 366400
    },
    {
      "epoch": 0.033,
      "grad_norm": 0.20657120645046234,
      "learning_rate": 5.3400400000000005e-05,
      "loss": 0.2445,
      "step": 366500
    },
    {
      "epoch": 0.0332,
      "grad_norm": 0.5175828337669373,
      "learning_rate": 5.33604e-05,
      "loss": 0.2467,
      "step": 366600
    },
    {
      "epoch": 0.0334,
      "grad_norm": 0.2220238447189331,
      "learning_rate": 5.3320400000000005e-05,
      "loss": 0.319,
      "step": 366700
    },
    {
      "epoch": 0.0336,
      "grad_norm": 0.3628231883049011,
      "learning_rate": 5.3280400000000006e-05,
      "loss": 0.2405,
      "step": 366800
    },
    {
      "epoch": 0.0338,
      "grad_norm": 0.235669806599617,
      "learning_rate": 5.32404e-05,
      "loss": 0.245,
      "step": 366900
    },
    {
      "epoch": 0.034,
      "grad_norm": 0.1623888611793518,
      "learning_rate": 5.3200400000000007e-05,
      "loss": 0.2425,
      "step": 367000
    },
    {
      "epoch": 0.0342,
      "grad_norm": 0.28828221559524536,
      "learning_rate": 5.31604e-05,
      "loss": 0.2411,
      "step": 367100
    },
    {
      "epoch": 0.0344,
      "grad_norm": 0.7484803199768066,
      "learning_rate": 5.31204e-05,
      "loss": 0.2386,
      "step": 367200
    },
    {
      "epoch": 0.0346,
      "grad_norm": 0.39959147572517395,
      "learning_rate": 5.308040000000001e-05,
      "loss": 0.2359,
      "step": 367300
    },
    {
      "epoch": 0.0348,
      "grad_norm": 0.1961216777563095,
      "learning_rate": 5.30404e-05,
      "loss": 0.2398,
      "step": 367400
    },
    {
      "epoch": 0.035,
      "grad_norm": 0.17088356614112854,
      "learning_rate": 5.300040000000001e-05,
      "loss": 0.2363,
      "step": 367500
    },
    {
      "epoch": 0.0352,
      "grad_norm": 0.394522100687027,
      "learning_rate": 5.29604e-05,
      "loss": 0.2385,
      "step": 367600
    },
    {
      "epoch": 0.0354,
      "grad_norm": 0.26079925894737244,
      "learning_rate": 5.29204e-05,
      "loss": 0.2316,
      "step": 367700
    },
    {
      "epoch": 0.0356,
      "grad_norm": 0.3335099220275879,
      "learning_rate": 5.288040000000001e-05,
      "loss": 0.2549,
      "step": 367800
    },
    {
      "epoch": 0.0358,
      "grad_norm": 0.3195158541202545,
      "learning_rate": 5.28404e-05,
      "loss": 0.2463,
      "step": 367900
    },
    {
      "epoch": 0.036,
      "grad_norm": 0.18612483143806458,
      "learning_rate": 5.2800400000000004e-05,
      "loss": 0.2408,
      "step": 368000
    },
    {
      "epoch": 0.0362,
      "grad_norm": 0.4741574823856354,
      "learning_rate": 5.27604e-05,
      "loss": 0.245,
      "step": 368100
    },
    {
      "epoch": 0.0364,
      "grad_norm": 0.26677343249320984,
      "learning_rate": 5.2720400000000004e-05,
      "loss": 0.2402,
      "step": 368200
    },
    {
      "epoch": 0.0366,
      "grad_norm": 0.24383248388767242,
      "learning_rate": 5.2680400000000005e-05,
      "loss": 0.242,
      "step": 368300
    },
    {
      "epoch": 0.0368,
      "grad_norm": 0.18668265640735626,
      "learning_rate": 5.26404e-05,
      "loss": 0.2431,
      "step": 368400
    },
    {
      "epoch": 0.037,
      "grad_norm": 0.18299442529678345,
      "learning_rate": 5.2600400000000005e-05,
      "loss": 0.2464,
      "step": 368500
    },
    {
      "epoch": 0.0372,
      "grad_norm": 0.22154571115970612,
      "learning_rate": 5.25604e-05,
      "loss": 0.2353,
      "step": 368600
    },
    {
      "epoch": 0.0374,
      "grad_norm": 1.3703587055206299,
      "learning_rate": 5.2520400000000006e-05,
      "loss": 0.2474,
      "step": 368700
    },
    {
      "epoch": 0.0376,
      "grad_norm": 0.19284115731716156,
      "learning_rate": 5.248040000000001e-05,
      "loss": 0.2467,
      "step": 368800
    },
    {
      "epoch": 0.0378,
      "grad_norm": 0.2304501235485077,
      "learning_rate": 5.24404e-05,
      "loss": 0.2373,
      "step": 368900
    },
    {
      "epoch": 0.038,
      "grad_norm": 0.1811191737651825,
      "learning_rate": 5.240040000000001e-05,
      "loss": 0.2381,
      "step": 369000
    },
    {
      "epoch": 0.0382,
      "grad_norm": 0.19747400283813477,
      "learning_rate": 5.23604e-05,
      "loss": 0.2437,
      "step": 369100
    },
    {
      "epoch": 0.0384,
      "grad_norm": 0.16430014371871948,
      "learning_rate": 5.23204e-05,
      "loss": 0.2458,
      "step": 369200
    },
    {
      "epoch": 0.0386,
      "grad_norm": 0.27376890182495117,
      "learning_rate": 5.228040000000001e-05,
      "loss": 0.2373,
      "step": 369300
    },
    {
      "epoch": 0.0388,
      "grad_norm": 0.21681749820709229,
      "learning_rate": 5.22404e-05,
      "loss": 0.2479,
      "step": 369400
    },
    {
      "epoch": 0.039,
      "grad_norm": 0.28773918747901917,
      "learning_rate": 5.22004e-05,
      "loss": 0.2521,
      "step": 369500
    },
    {
      "epoch": 0.0392,
      "grad_norm": 0.26816967129707336,
      "learning_rate": 5.2160399999999996e-05,
      "loss": 0.2464,
      "step": 369600
    },
    {
      "epoch": 0.0394,
      "grad_norm": 0.16062785685062408,
      "learning_rate": 5.21204e-05,
      "loss": 0.239,
      "step": 369700
    },
    {
      "epoch": 0.0396,
      "grad_norm": 0.23461496829986572,
      "learning_rate": 5.2080400000000004e-05,
      "loss": 0.2394,
      "step": 369800
    },
    {
      "epoch": 0.0398,
      "grad_norm": 0.19398300349712372,
      "learning_rate": 5.20404e-05,
      "loss": 0.2365,
      "step": 369900
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.16843223571777344,
      "learning_rate": 5.2000400000000004e-05,
      "loss": 0.2425,
      "step": 370000
    },
    {
      "epoch": 0.0402,
      "grad_norm": 0.19908249378204346,
      "learning_rate": 5.19604e-05,
      "loss": 0.2339,
      "step": 370100
    },
    {
      "epoch": 0.0404,
      "grad_norm": 0.22099874913692474,
      "learning_rate": 5.1920400000000005e-05,
      "loss": 0.2519,
      "step": 370200
    },
    {
      "epoch": 0.0406,
      "grad_norm": 0.2163575440645218,
      "learning_rate": 5.1880400000000006e-05,
      "loss": 0.2494,
      "step": 370300
    },
    {
      "epoch": 0.0408,
      "grad_norm": 0.2572519779205322,
      "learning_rate": 5.18404e-05,
      "loss": 0.2431,
      "step": 370400
    },
    {
      "epoch": 0.041,
      "grad_norm": 0.5304775834083557,
      "learning_rate": 5.1800400000000006e-05,
      "loss": 0.2427,
      "step": 370500
    },
    {
      "epoch": 0.0412,
      "grad_norm": 0.3667397201061249,
      "learning_rate": 5.17604e-05,
      "loss": 0.2403,
      "step": 370600
    },
    {
      "epoch": 0.0414,
      "grad_norm": 0.24394138157367706,
      "learning_rate": 5.17204e-05,
      "loss": 0.2401,
      "step": 370700
    },
    {
      "epoch": 0.0416,
      "grad_norm": 0.16485385596752167,
      "learning_rate": 5.168040000000001e-05,
      "loss": 0.2516,
      "step": 370800
    },
    {
      "epoch": 0.0418,
      "grad_norm": 0.19729548692703247,
      "learning_rate": 5.16404e-05,
      "loss": 0.2456,
      "step": 370900
    },
    {
      "epoch": 0.042,
      "grad_norm": 0.3520711064338684,
      "learning_rate": 5.16004e-05,
      "loss": 0.2456,
      "step": 371000
    },
    {
      "epoch": 0.0422,
      "grad_norm": 0.1835898905992508,
      "learning_rate": 5.1560399999999995e-05,
      "loss": 0.2417,
      "step": 371100
    },
    {
      "epoch": 0.0424,
      "grad_norm": 0.502991795539856,
      "learning_rate": 5.15204e-05,
      "loss": 0.2446,
      "step": 371200
    },
    {
      "epoch": 0.0426,
      "grad_norm": 0.2608524560928345,
      "learning_rate": 5.148040000000001e-05,
      "loss": 0.2473,
      "step": 371300
    },
    {
      "epoch": 0.0428,
      "grad_norm": 0.23622851073741913,
      "learning_rate": 5.14404e-05,
      "loss": 0.2451,
      "step": 371400
    },
    {
      "epoch": 0.043,
      "grad_norm": 0.3664970099925995,
      "learning_rate": 5.1400400000000003e-05,
      "loss": 0.2457,
      "step": 371500
    },
    {
      "epoch": 0.0432,
      "grad_norm": 0.16481013596057892,
      "learning_rate": 5.13604e-05,
      "loss": 0.2499,
      "step": 371600
    },
    {
      "epoch": 0.0434,
      "grad_norm": 0.23250015079975128,
      "learning_rate": 5.1320400000000004e-05,
      "loss": 0.2453,
      "step": 371700
    },
    {
      "epoch": 0.0436,
      "grad_norm": 0.262839674949646,
      "learning_rate": 5.1280400000000005e-05,
      "loss": 0.2438,
      "step": 371800
    },
    {
      "epoch": 0.0438,
      "grad_norm": 0.14810775220394135,
      "learning_rate": 5.12404e-05,
      "loss": 0.2445,
      "step": 371900
    },
    {
      "epoch": 0.044,
      "grad_norm": 0.20043599605560303,
      "learning_rate": 5.1200400000000005e-05,
      "loss": 0.2465,
      "step": 372000
    },
    {
      "epoch": 0.0442,
      "grad_norm": 0.21188238263130188,
      "learning_rate": 5.11604e-05,
      "loss": 0.248,
      "step": 372100
    },
    {
      "epoch": 0.0444,
      "grad_norm": 0.19826635718345642,
      "learning_rate": 5.11204e-05,
      "loss": 0.2357,
      "step": 372200
    },
    {
      "epoch": 0.0446,
      "grad_norm": 0.26934173703193665,
      "learning_rate": 5.1080400000000006e-05,
      "loss": 0.2872,
      "step": 372300
    },
    {
      "epoch": 0.0448,
      "grad_norm": 0.21112340688705444,
      "learning_rate": 5.10404e-05,
      "loss": 0.2405,
      "step": 372400
    },
    {
      "epoch": 0.045,
      "grad_norm": 0.17311957478523254,
      "learning_rate": 5.10004e-05,
      "loss": 0.2367,
      "step": 372500
    },
    {
      "epoch": 0.0452,
      "grad_norm": 0.16153335571289062,
      "learning_rate": 5.0960399999999994e-05,
      "loss": 0.2371,
      "step": 372600
    },
    {
      "epoch": 0.0454,
      "grad_norm": 0.19075436890125275,
      "learning_rate": 5.09204e-05,
      "loss": 0.2342,
      "step": 372700
    },
    {
      "epoch": 0.0456,
      "grad_norm": 0.26471832394599915,
      "learning_rate": 5.088040000000001e-05,
      "loss": 0.2336,
      "step": 372800
    },
    {
      "epoch": 0.0458,
      "grad_norm": 0.2320035994052887,
      "learning_rate": 5.08404e-05,
      "loss": 0.238,
      "step": 372900
    },
    {
      "epoch": 0.046,
      "grad_norm": 0.19585338234901428,
      "learning_rate": 5.08004e-05,
      "loss": 0.2374,
      "step": 373000
    },
    {
      "epoch": 0.0462,
      "grad_norm": 0.21926333010196686,
      "learning_rate": 5.076040000000001e-05,
      "loss": 0.2452,
      "step": 373100
    },
    {
      "epoch": 0.0464,
      "grad_norm": 0.2396099716424942,
      "learning_rate": 5.07204e-05,
      "loss": 0.2591,
      "step": 373200
    },
    {
      "epoch": 0.0466,
      "grad_norm": 0.23458154499530792,
      "learning_rate": 5.0680400000000004e-05,
      "loss": 0.2393,
      "step": 373300
    },
    {
      "epoch": 0.0468,
      "grad_norm": 0.23646235466003418,
      "learning_rate": 5.06404e-05,
      "loss": 0.2455,
      "step": 373400
    },
    {
      "epoch": 0.047,
      "grad_norm": 0.16524997353553772,
      "learning_rate": 5.0600400000000004e-05,
      "loss": 0.2423,
      "step": 373500
    },
    {
      "epoch": 0.0472,
      "grad_norm": 0.19128525257110596,
      "learning_rate": 5.0560400000000005e-05,
      "loss": 0.2396,
      "step": 373600
    },
    {
      "epoch": 0.0474,
      "grad_norm": 0.1869351863861084,
      "learning_rate": 5.05204e-05,
      "loss": 0.243,
      "step": 373700
    },
    {
      "epoch": 0.0476,
      "grad_norm": 0.22153295576572418,
      "learning_rate": 5.0480400000000005e-05,
      "loss": 0.2425,
      "step": 373800
    },
    {
      "epoch": 0.0478,
      "grad_norm": 0.20191314816474915,
      "learning_rate": 5.04404e-05,
      "loss": 0.247,
      "step": 373900
    },
    {
      "epoch": 0.048,
      "grad_norm": 0.20306941866874695,
      "learning_rate": 5.0400400000000006e-05,
      "loss": 0.2463,
      "step": 374000
    },
    {
      "epoch": 0.0482,
      "grad_norm": 0.29261788725852966,
      "learning_rate": 5.0360400000000007e-05,
      "loss": 0.2492,
      "step": 374100
    },
    {
      "epoch": 0.0484,
      "grad_norm": 0.3436906635761261,
      "learning_rate": 5.03204e-05,
      "loss": 0.2462,
      "step": 374200
    },
    {
      "epoch": 0.0486,
      "grad_norm": 0.19893062114715576,
      "learning_rate": 5.028040000000001e-05,
      "loss": 0.2487,
      "step": 374300
    },
    {
      "epoch": 0.0488,
      "grad_norm": 0.17730703949928284,
      "learning_rate": 5.02404e-05,
      "loss": 0.2388,
      "step": 374400
    },
    {
      "epoch": 0.049,
      "grad_norm": 0.2693403661251068,
      "learning_rate": 5.02004e-05,
      "loss": 0.2472,
      "step": 374500
    },
    {
      "epoch": 0.0492,
      "grad_norm": 0.20405225455760956,
      "learning_rate": 5.016040000000001e-05,
      "loss": 0.2382,
      "step": 374600
    },
    {
      "epoch": 0.0494,
      "grad_norm": 0.30344367027282715,
      "learning_rate": 5.01204e-05,
      "loss": 0.2484,
      "step": 374700
    },
    {
      "epoch": 0.0496,
      "grad_norm": 0.23984545469284058,
      "learning_rate": 5.00804e-05,
      "loss": 0.2453,
      "step": 374800
    },
    {
      "epoch": 0.0498,
      "grad_norm": 0.20003117620944977,
      "learning_rate": 5.0040399999999996e-05,
      "loss": 0.2434,
      "step": 374900
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.19574657082557678,
      "learning_rate": 5.00004e-05,
      "loss": 0.2461,
      "step": 375000
    },
    {
      "epoch": 0.0502,
      "grad_norm": 0.1617669016122818,
      "learning_rate": 4.9960400000000004e-05,
      "loss": 0.2407,
      "step": 375100
    },
    {
      "epoch": 0.0504,
      "grad_norm": 0.37597665190696716,
      "learning_rate": 4.99204e-05,
      "loss": 0.2579,
      "step": 375200
    },
    {
      "epoch": 0.0506,
      "grad_norm": 0.21448801457881927,
      "learning_rate": 4.9880400000000004e-05,
      "loss": 0.2458,
      "step": 375300
    },
    {
      "epoch": 0.0508,
      "grad_norm": 0.2033076137304306,
      "learning_rate": 4.9840400000000005e-05,
      "loss": 0.246,
      "step": 375400
    },
    {
      "epoch": 0.051,
      "grad_norm": 0.15409351885318756,
      "learning_rate": 4.9800400000000005e-05,
      "loss": 0.2478,
      "step": 375500
    },
    {
      "epoch": 0.0512,
      "grad_norm": 0.2577267289161682,
      "learning_rate": 4.97604e-05,
      "loss": 0.2436,
      "step": 375600
    },
    {
      "epoch": 0.0514,
      "grad_norm": 0.1761707067489624,
      "learning_rate": 4.97204e-05,
      "loss": 0.2505,
      "step": 375700
    },
    {
      "epoch": 0.0516,
      "grad_norm": 0.2014208734035492,
      "learning_rate": 4.9680400000000006e-05,
      "loss": 0.2504,
      "step": 375800
    },
    {
      "epoch": 0.0518,
      "grad_norm": 0.21886250376701355,
      "learning_rate": 4.964040000000001e-05,
      "loss": 0.2437,
      "step": 375900
    },
    {
      "epoch": 0.052,
      "grad_norm": 0.3855169415473938,
      "learning_rate": 4.96004e-05,
      "loss": 0.2589,
      "step": 376000
    },
    {
      "epoch": 0.0522,
      "grad_norm": 0.21187375485897064,
      "learning_rate": 4.95604e-05,
      "loss": 0.2376,
      "step": 376100
    },
    {
      "epoch": 0.0524,
      "grad_norm": 0.16554217040538788,
      "learning_rate": 4.95204e-05,
      "loss": 0.2448,
      "step": 376200
    },
    {
      "epoch": 0.0526,
      "grad_norm": 0.3250042498111725,
      "learning_rate": 4.94804e-05,
      "loss": 0.2583,
      "step": 376300
    },
    {
      "epoch": 0.0528,
      "grad_norm": 0.21343812346458435,
      "learning_rate": 4.94404e-05,
      "loss": 0.2478,
      "step": 376400
    },
    {
      "epoch": 0.053,
      "grad_norm": 0.2855488359928131,
      "learning_rate": 4.94004e-05,
      "loss": 0.2418,
      "step": 376500
    },
    {
      "epoch": 0.0532,
      "grad_norm": 0.21316973865032196,
      "learning_rate": 4.93604e-05,
      "loss": 0.254,
      "step": 376600
    },
    {
      "epoch": 0.0534,
      "grad_norm": 0.16673021018505096,
      "learning_rate": 4.93204e-05,
      "loss": 0.2412,
      "step": 376700
    },
    {
      "epoch": 0.0536,
      "grad_norm": 0.2050374299287796,
      "learning_rate": 4.9280400000000003e-05,
      "loss": 0.2442,
      "step": 376800
    },
    {
      "epoch": 0.0538,
      "grad_norm": 0.17526887357234955,
      "learning_rate": 4.9240400000000004e-05,
      "loss": 0.2456,
      "step": 376900
    },
    {
      "epoch": 0.054,
      "grad_norm": 0.17823076248168945,
      "learning_rate": 4.9200400000000004e-05,
      "loss": 0.2412,
      "step": 377000
    },
    {
      "epoch": 0.0542,
      "grad_norm": 0.19412919878959656,
      "learning_rate": 4.91604e-05,
      "loss": 0.2427,
      "step": 377100
    },
    {
      "epoch": 0.0544,
      "grad_norm": 0.2670155167579651,
      "learning_rate": 4.91204e-05,
      "loss": 0.2478,
      "step": 377200
    },
    {
      "epoch": 0.0546,
      "grad_norm": 0.3044726252555847,
      "learning_rate": 4.9080400000000005e-05,
      "loss": 0.2485,
      "step": 377300
    },
    {
      "epoch": 0.0548,
      "grad_norm": 0.27025651931762695,
      "learning_rate": 4.9040400000000006e-05,
      "loss": 0.2445,
      "step": 377400
    },
    {
      "epoch": 0.055,
      "grad_norm": 0.19618715345859528,
      "learning_rate": 4.90004e-05,
      "loss": 0.2397,
      "step": 377500
    },
    {
      "epoch": 0.0552,
      "grad_norm": 0.22691640257835388,
      "learning_rate": 4.89604e-05,
      "loss": 0.2494,
      "step": 377600
    },
    {
      "epoch": 0.0554,
      "grad_norm": 0.346432089805603,
      "learning_rate": 4.892040000000001e-05,
      "loss": 0.2626,
      "step": 377700
    },
    {
      "epoch": 0.0556,
      "grad_norm": 0.37104085087776184,
      "learning_rate": 4.888040000000001e-05,
      "loss": 0.2472,
      "step": 377800
    },
    {
      "epoch": 0.0558,
      "grad_norm": 0.25854432582855225,
      "learning_rate": 4.88404e-05,
      "loss": 0.2405,
      "step": 377900
    },
    {
      "epoch": 0.056,
      "grad_norm": 0.22568698227405548,
      "learning_rate": 4.88004e-05,
      "loss": 0.2577,
      "step": 378000
    },
    {
      "epoch": 0.0562,
      "grad_norm": 0.19181467592716217,
      "learning_rate": 4.87604e-05,
      "loss": 0.2493,
      "step": 378100
    },
    {
      "epoch": 0.0564,
      "grad_norm": 0.19588135182857513,
      "learning_rate": 4.87204e-05,
      "loss": 0.249,
      "step": 378200
    },
    {
      "epoch": 0.0566,
      "grad_norm": 0.3186732232570648,
      "learning_rate": 4.86804e-05,
      "loss": 0.2346,
      "step": 378300
    },
    {
      "epoch": 0.0568,
      "grad_norm": 0.21397516131401062,
      "learning_rate": 4.86404e-05,
      "loss": 0.254,
      "step": 378400
    },
    {
      "epoch": 0.057,
      "grad_norm": 0.14002354443073273,
      "learning_rate": 4.86004e-05,
      "loss": 0.2503,
      "step": 378500
    },
    {
      "epoch": 0.0572,
      "grad_norm": 0.18620188534259796,
      "learning_rate": 4.8560400000000004e-05,
      "loss": 0.2477,
      "step": 378600
    },
    {
      "epoch": 0.0574,
      "grad_norm": 0.34367236495018005,
      "learning_rate": 4.8520400000000004e-05,
      "loss": 0.2515,
      "step": 378700
    },
    {
      "epoch": 0.0576,
      "grad_norm": 0.20147454738616943,
      "learning_rate": 4.8480400000000004e-05,
      "loss": 0.2509,
      "step": 378800
    },
    {
      "epoch": 0.0578,
      "grad_norm": 0.22999276220798492,
      "learning_rate": 4.8440400000000005e-05,
      "loss": 0.2595,
      "step": 378900
    },
    {
      "epoch": 0.058,
      "grad_norm": 0.1593378633260727,
      "learning_rate": 4.84004e-05,
      "loss": 0.2501,
      "step": 379000
    },
    {
      "epoch": 0.0582,
      "grad_norm": 0.2817172706127167,
      "learning_rate": 4.83604e-05,
      "loss": 0.2411,
      "step": 379100
    },
    {
      "epoch": 0.0584,
      "grad_norm": 0.16755926609039307,
      "learning_rate": 4.8320400000000006e-05,
      "loss": 0.2415,
      "step": 379200
    },
    {
      "epoch": 0.0586,
      "grad_norm": 0.19628214836120605,
      "learning_rate": 4.8280400000000006e-05,
      "loss": 0.2457,
      "step": 379300
    },
    {
      "epoch": 0.0588,
      "grad_norm": 0.16852396726608276,
      "learning_rate": 4.82404e-05,
      "loss": 0.2513,
      "step": 379400
    },
    {
      "epoch": 0.059,
      "grad_norm": 0.3213140666484833,
      "learning_rate": 4.82004e-05,
      "loss": 0.2663,
      "step": 379500
    },
    {
      "epoch": 0.0592,
      "grad_norm": 0.2476736158132553,
      "learning_rate": 4.81604e-05,
      "loss": 0.2671,
      "step": 379600
    },
    {
      "epoch": 0.0594,
      "grad_norm": 0.1939520239830017,
      "learning_rate": 4.812040000000001e-05,
      "loss": 0.2486,
      "step": 379700
    },
    {
      "epoch": 0.0596,
      "grad_norm": 0.15622283518314362,
      "learning_rate": 4.80804e-05,
      "loss": 0.2358,
      "step": 379800
    },
    {
      "epoch": 0.0598,
      "grad_norm": 0.18065965175628662,
      "learning_rate": 4.80404e-05,
      "loss": 0.2286,
      "step": 379900
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.16677778959274292,
      "learning_rate": 4.80004e-05,
      "loss": 0.2283,
      "step": 380000
    },
    {
      "epoch": 0.0602,
      "grad_norm": 0.2198459357023239,
      "learning_rate": 4.79604e-05,
      "loss": 0.2241,
      "step": 380100
    },
    {
      "epoch": 0.0604,
      "grad_norm": 0.18271780014038086,
      "learning_rate": 4.79204e-05,
      "loss": 0.2241,
      "step": 380200
    },
    {
      "epoch": 0.0606,
      "grad_norm": 0.16740696132183075,
      "learning_rate": 4.78804e-05,
      "loss": 0.2233,
      "step": 380300
    },
    {
      "epoch": 0.0608,
      "grad_norm": 0.1739855706691742,
      "learning_rate": 4.7840400000000004e-05,
      "loss": 0.2212,
      "step": 380400
    },
    {
      "epoch": 0.061,
      "grad_norm": 0.20133738219738007,
      "learning_rate": 4.7800400000000004e-05,
      "loss": 0.2228,
      "step": 380500
    },
    {
      "epoch": 0.0612,
      "grad_norm": 0.15135647356510162,
      "learning_rate": 4.77604e-05,
      "loss": 0.2212,
      "step": 380600
    },
    {
      "epoch": 0.0614,
      "grad_norm": 0.19083605706691742,
      "learning_rate": 4.7720400000000005e-05,
      "loss": 0.2218,
      "step": 380700
    },
    {
      "epoch": 0.0616,
      "grad_norm": 0.16143475472927094,
      "learning_rate": 4.7680400000000005e-05,
      "loss": 0.2249,
      "step": 380800
    },
    {
      "epoch": 0.0618,
      "grad_norm": 0.23794889450073242,
      "learning_rate": 4.76404e-05,
      "loss": 0.2298,
      "step": 380900
    },
    {
      "epoch": 0.062,
      "grad_norm": 0.19995640218257904,
      "learning_rate": 4.76004e-05,
      "loss": 0.2209,
      "step": 381000
    },
    {
      "epoch": 0.0622,
      "grad_norm": 0.1754998117685318,
      "learning_rate": 4.7560400000000006e-05,
      "loss": 0.2258,
      "step": 381100
    },
    {
      "epoch": 0.0624,
      "grad_norm": 0.18169523775577545,
      "learning_rate": 4.752040000000001e-05,
      "loss": 0.2254,
      "step": 381200
    },
    {
      "epoch": 0.0626,
      "grad_norm": 0.18677940964698792,
      "learning_rate": 4.74804e-05,
      "loss": 0.2254,
      "step": 381300
    },
    {
      "epoch": 0.0628,
      "grad_norm": 0.1730923056602478,
      "learning_rate": 4.74404e-05,
      "loss": 0.2246,
      "step": 381400
    },
    {
      "epoch": 0.063,
      "grad_norm": 0.20200613141059875,
      "learning_rate": 4.74004e-05,
      "loss": 0.2234,
      "step": 381500
    },
    {
      "epoch": 0.0632,
      "grad_norm": 0.33600297570228577,
      "learning_rate": 4.73604e-05,
      "loss": 0.2258,
      "step": 381600
    },
    {
      "epoch": 0.0634,
      "grad_norm": 0.16173823177814484,
      "learning_rate": 4.73204e-05,
      "loss": 0.2248,
      "step": 381700
    },
    {
      "epoch": 0.0636,
      "grad_norm": 0.16113696992397308,
      "learning_rate": 4.72804e-05,
      "loss": 0.2205,
      "step": 381800
    },
    {
      "epoch": 0.0638,
      "grad_norm": 0.17868146300315857,
      "learning_rate": 4.72404e-05,
      "loss": 0.2205,
      "step": 381900
    },
    {
      "epoch": 0.064,
      "grad_norm": 0.17537720501422882,
      "learning_rate": 4.72004e-05,
      "loss": 0.2271,
      "step": 382000
    },
    {
      "epoch": 0.0642,
      "grad_norm": 0.17318007349967957,
      "learning_rate": 4.71604e-05,
      "loss": 0.2282,
      "step": 382100
    },
    {
      "epoch": 0.0644,
      "grad_norm": 0.14845995604991913,
      "learning_rate": 4.7120400000000004e-05,
      "loss": 0.2247,
      "step": 382200
    },
    {
      "epoch": 0.0646,
      "grad_norm": 0.17553631961345673,
      "learning_rate": 4.7080400000000004e-05,
      "loss": 0.2244,
      "step": 382300
    },
    {
      "epoch": 0.0648,
      "grad_norm": 0.20734809339046478,
      "learning_rate": 4.7040400000000005e-05,
      "loss": 0.2221,
      "step": 382400
    },
    {
      "epoch": 0.065,
      "grad_norm": 0.1688157171010971,
      "learning_rate": 4.70004e-05,
      "loss": 0.2225,
      "step": 382500
    },
    {
      "epoch": 0.0652,
      "grad_norm": 0.17154380679130554,
      "learning_rate": 4.6960400000000005e-05,
      "loss": 0.2267,
      "step": 382600
    },
    {
      "epoch": 0.0654,
      "grad_norm": 0.19217541813850403,
      "learning_rate": 4.6920400000000006e-05,
      "loss": 0.2263,
      "step": 382700
    },
    {
      "epoch": 0.0656,
      "grad_norm": 0.1954587697982788,
      "learning_rate": 4.68804e-05,
      "loss": 0.2242,
      "step": 382800
    },
    {
      "epoch": 0.0658,
      "grad_norm": 0.18474943935871124,
      "learning_rate": 4.68404e-05,
      "loss": 0.2257,
      "step": 382900
    },
    {
      "epoch": 0.066,
      "grad_norm": 0.1891472041606903,
      "learning_rate": 4.68004e-05,
      "loss": 0.2259,
      "step": 383000
    },
    {
      "epoch": 0.0662,
      "grad_norm": 0.16297364234924316,
      "learning_rate": 4.676040000000001e-05,
      "loss": 0.2239,
      "step": 383100
    },
    {
      "epoch": 0.0664,
      "grad_norm": 0.26741859316825867,
      "learning_rate": 4.67204e-05,
      "loss": 0.2265,
      "step": 383200
    },
    {
      "epoch": 0.0666,
      "grad_norm": 0.22171659767627716,
      "learning_rate": 4.66804e-05,
      "loss": 0.2272,
      "step": 383300
    },
    {
      "epoch": 0.0668,
      "grad_norm": 0.16574308276176453,
      "learning_rate": 4.66404e-05,
      "loss": 0.2272,
      "step": 383400
    },
    {
      "epoch": 0.067,
      "grad_norm": 0.15950843691825867,
      "learning_rate": 4.66004e-05,
      "loss": 0.2259,
      "step": 383500
    },
    {
      "epoch": 0.0672,
      "grad_norm": 0.1801784336566925,
      "learning_rate": 4.65604e-05,
      "loss": 0.2277,
      "step": 383600
    },
    {
      "epoch": 0.0674,
      "grad_norm": 0.14734512567520142,
      "learning_rate": 4.65204e-05,
      "loss": 0.224,
      "step": 383700
    },
    {
      "epoch": 0.0676,
      "grad_norm": 0.20324218273162842,
      "learning_rate": 4.64804e-05,
      "loss": 0.2257,
      "step": 383800
    },
    {
      "epoch": 0.0678,
      "grad_norm": 0.190351203083992,
      "learning_rate": 4.6440400000000003e-05,
      "loss": 0.2238,
      "step": 383900
    },
    {
      "epoch": 0.068,
      "grad_norm": 0.17872920632362366,
      "learning_rate": 4.6400400000000004e-05,
      "loss": 0.2253,
      "step": 384000
    },
    {
      "epoch": 0.0682,
      "grad_norm": 0.1632169634103775,
      "learning_rate": 4.6360400000000004e-05,
      "loss": 0.2296,
      "step": 384100
    },
    {
      "epoch": 0.0684,
      "grad_norm": 0.219464972615242,
      "learning_rate": 4.6320400000000005e-05,
      "loss": 0.2267,
      "step": 384200
    },
    {
      "epoch": 0.0686,
      "grad_norm": 0.19538037478923798,
      "learning_rate": 4.62804e-05,
      "loss": 0.2216,
      "step": 384300
    },
    {
      "epoch": 0.0688,
      "grad_norm": 0.16144002974033356,
      "learning_rate": 4.62404e-05,
      "loss": 0.2254,
      "step": 384400
    },
    {
      "epoch": 0.069,
      "grad_norm": 0.254450261592865,
      "learning_rate": 4.6200400000000006e-05,
      "loss": 0.2295,
      "step": 384500
    },
    {
      "epoch": 0.0692,
      "grad_norm": 0.17138288915157318,
      "learning_rate": 4.6160400000000006e-05,
      "loss": 0.2254,
      "step": 384600
    },
    {
      "epoch": 0.0694,
      "grad_norm": 0.22051529586315155,
      "learning_rate": 4.61204e-05,
      "loss": 0.2305,
      "step": 384700
    },
    {
      "epoch": 0.0696,
      "grad_norm": 0.17428357899188995,
      "learning_rate": 4.60804e-05,
      "loss": 0.2237,
      "step": 384800
    },
    {
      "epoch": 0.0698,
      "grad_norm": 0.16001330316066742,
      "learning_rate": 4.60404e-05,
      "loss": 0.2286,
      "step": 384900
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.16556522250175476,
      "learning_rate": 4.600040000000001e-05,
      "loss": 0.2301,
      "step": 385000
    },
    {
      "epoch": 0.0702,
      "grad_norm": NaN,
      "learning_rate": 4.59604e-05,
      "loss": 0.2256,
      "step": 385100
    },
    {
      "epoch": 0.0704,
      "grad_norm": 0.272354394197464,
      "learning_rate": 4.59204e-05,
      "loss": 0.229,
      "step": 385200
    },
    {
      "epoch": 0.0706,
      "grad_norm": 0.28128334879875183,
      "learning_rate": 4.58804e-05,
      "loss": 0.2292,
      "step": 385300
    },
    {
      "epoch": 0.0708,
      "grad_norm": 0.1840551495552063,
      "learning_rate": 4.58404e-05,
      "loss": 0.2297,
      "step": 385400
    },
    {
      "epoch": 0.071,
      "grad_norm": 0.13684575259685516,
      "learning_rate": 4.58004e-05,
      "loss": 0.2253,
      "step": 385500
    },
    {
      "epoch": 0.0712,
      "grad_norm": 0.17495335638523102,
      "learning_rate": 4.57604e-05,
      "loss": 0.2241,
      "step": 385600
    },
    {
      "epoch": 0.0714,
      "grad_norm": 0.2242329716682434,
      "learning_rate": 4.5720400000000004e-05,
      "loss": 0.2238,
      "step": 385700
    },
    {
      "epoch": 0.0716,
      "grad_norm": 0.20640377700328827,
      "learning_rate": 4.5680400000000004e-05,
      "loss": 0.2274,
      "step": 385800
    },
    {
      "epoch": 0.0718,
      "grad_norm": 0.19061022996902466,
      "learning_rate": 4.56404e-05,
      "loss": 0.2252,
      "step": 385900
    },
    {
      "epoch": 0.072,
      "grad_norm": 0.3011106550693512,
      "learning_rate": 4.5600400000000005e-05,
      "loss": 0.2245,
      "step": 386000
    },
    {
      "epoch": 0.0722,
      "grad_norm": 0.19182322919368744,
      "learning_rate": 4.5560400000000005e-05,
      "loss": 0.2285,
      "step": 386100
    },
    {
      "epoch": 0.0724,
      "grad_norm": 0.15975511074066162,
      "learning_rate": 4.55204e-05,
      "loss": 0.2298,
      "step": 386200
    },
    {
      "epoch": 0.0726,
      "grad_norm": 0.16059580445289612,
      "learning_rate": 4.54804e-05,
      "loss": 0.2273,
      "step": 386300
    },
    {
      "epoch": 0.0728,
      "grad_norm": 0.1402633935213089,
      "learning_rate": 4.54404e-05,
      "loss": 0.2276,
      "step": 386400
    },
    {
      "epoch": 0.073,
      "grad_norm": 0.3752526044845581,
      "learning_rate": 4.540040000000001e-05,
      "loss": 0.2246,
      "step": 386500
    },
    {
      "epoch": 0.0732,
      "grad_norm": 0.15643823146820068,
      "learning_rate": 4.53604e-05,
      "loss": 0.2244,
      "step": 386600
    },
    {
      "epoch": 0.0734,
      "grad_norm": 0.16925235092639923,
      "learning_rate": 4.53204e-05,
      "loss": 0.2266,
      "step": 386700
    },
    {
      "epoch": 0.0736,
      "grad_norm": 0.27427610754966736,
      "learning_rate": 4.52804e-05,
      "loss": 0.226,
      "step": 386800
    },
    {
      "epoch": 0.0738,
      "grad_norm": 0.175080806016922,
      "learning_rate": 4.52404e-05,
      "loss": 0.2286,
      "step": 386900
    },
    {
      "epoch": 0.074,
      "grad_norm": 0.23322616517543793,
      "learning_rate": 4.52004e-05,
      "loss": 0.2275,
      "step": 387000
    },
    {
      "epoch": 0.0742,
      "grad_norm": 0.17752708494663239,
      "learning_rate": 4.51604e-05,
      "loss": 0.2277,
      "step": 387100
    },
    {
      "epoch": 0.0744,
      "grad_norm": 0.18653720617294312,
      "learning_rate": 4.51204e-05,
      "loss": 0.2278,
      "step": 387200
    },
    {
      "epoch": 0.0746,
      "grad_norm": 0.24310079216957092,
      "learning_rate": 4.50804e-05,
      "loss": 0.2267,
      "step": 387300
    },
    {
      "epoch": 0.0748,
      "grad_norm": 0.16927888989448547,
      "learning_rate": 4.50404e-05,
      "loss": 0.2263,
      "step": 387400
    },
    {
      "epoch": 0.075,
      "grad_norm": 0.21127398312091827,
      "learning_rate": 4.5000400000000004e-05,
      "loss": 0.2318,
      "step": 387500
    },
    {
      "epoch": 0.0752,
      "grad_norm": 0.1537827104330063,
      "learning_rate": 4.4960400000000004e-05,
      "loss": 0.228,
      "step": 387600
    },
    {
      "epoch": 0.0754,
      "grad_norm": 0.2062017023563385,
      "learning_rate": 4.4920400000000004e-05,
      "loss": 0.2266,
      "step": 387700
    },
    {
      "epoch": 0.0756,
      "grad_norm": 0.1648017019033432,
      "learning_rate": 4.48804e-05,
      "loss": 0.2262,
      "step": 387800
    },
    {
      "epoch": 0.0758,
      "grad_norm": 0.17811764776706696,
      "learning_rate": 4.4840400000000005e-05,
      "loss": 0.2265,
      "step": 387900
    },
    {
      "epoch": 0.076,
      "grad_norm": 0.25383955240249634,
      "learning_rate": 4.4800400000000006e-05,
      "loss": 0.2307,
      "step": 388000
    },
    {
      "epoch": 0.0762,
      "grad_norm": 0.16029523313045502,
      "learning_rate": 4.47604e-05,
      "loss": 0.2271,
      "step": 388100
    },
    {
      "epoch": 0.0764,
      "grad_norm": 0.3880062699317932,
      "learning_rate": 4.47204e-05,
      "loss": 0.2219,
      "step": 388200
    },
    {
      "epoch": 0.0766,
      "grad_norm": 0.14046251773834229,
      "learning_rate": 4.46804e-05,
      "loss": 0.2272,
      "step": 388300
    },
    {
      "epoch": 0.0768,
      "grad_norm": 0.20129117369651794,
      "learning_rate": 4.464040000000001e-05,
      "loss": 0.2276,
      "step": 388400
    },
    {
      "epoch": 0.077,
      "grad_norm": 0.1550118327140808,
      "learning_rate": 4.46004e-05,
      "loss": 0.2253,
      "step": 388500
    },
    {
      "epoch": 0.0772,
      "grad_norm": 0.196579709649086,
      "learning_rate": 4.45604e-05,
      "loss": 0.2239,
      "step": 388600
    },
    {
      "epoch": 0.0774,
      "grad_norm": 0.23339785635471344,
      "learning_rate": 4.45204e-05,
      "loss": 0.2279,
      "step": 388700
    },
    {
      "epoch": 0.0776,
      "grad_norm": 0.20383737981319427,
      "learning_rate": 4.44804e-05,
      "loss": 0.2268,
      "step": 388800
    },
    {
      "epoch": 0.0778,
      "grad_norm": 0.15796469151973724,
      "learning_rate": 4.44404e-05,
      "loss": 0.2258,
      "step": 388900
    },
    {
      "epoch": 0.078,
      "grad_norm": 0.1844768226146698,
      "learning_rate": 4.44004e-05,
      "loss": 0.2325,
      "step": 389000
    },
    {
      "epoch": 0.0782,
      "grad_norm": 0.15635459125041962,
      "learning_rate": 4.43604e-05,
      "loss": 0.2297,
      "step": 389100
    },
    {
      "epoch": 0.0784,
      "grad_norm": 0.1638452708721161,
      "learning_rate": 4.4320400000000003e-05,
      "loss": 0.2229,
      "step": 389200
    },
    {
      "epoch": 0.0786,
      "grad_norm": 0.2581493854522705,
      "learning_rate": 4.42804e-05,
      "loss": 0.2257,
      "step": 389300
    },
    {
      "epoch": 0.0788,
      "grad_norm": 0.2030574083328247,
      "learning_rate": 4.4240400000000004e-05,
      "loss": 0.2282,
      "step": 389400
    },
    {
      "epoch": 0.079,
      "grad_norm": 0.14893430471420288,
      "learning_rate": 4.4200400000000005e-05,
      "loss": 0.2285,
      "step": 389500
    },
    {
      "epoch": 0.0792,
      "grad_norm": 0.15095509588718414,
      "learning_rate": 4.4160400000000005e-05,
      "loss": 0.2303,
      "step": 389600
    },
    {
      "epoch": 0.0794,
      "grad_norm": 0.20052938163280487,
      "learning_rate": 4.41204e-05,
      "loss": 0.2293,
      "step": 389700
    },
    {
      "epoch": 0.0796,
      "grad_norm": 0.2110019326210022,
      "learning_rate": 4.40804e-05,
      "loss": 0.2284,
      "step": 389800
    },
    {
      "epoch": 0.0798,
      "grad_norm": 0.18157057464122772,
      "learning_rate": 4.4040400000000006e-05,
      "loss": 0.2306,
      "step": 389900
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.21049971878528595,
      "learning_rate": 4.40004e-05,
      "loss": 0.2289,
      "step": 390000
    },
    {
      "epoch": 0.0802,
      "grad_norm": 0.18213237822055817,
      "learning_rate": 4.39604e-05,
      "loss": 0.2295,
      "step": 390100
    },
    {
      "epoch": 0.0804,
      "grad_norm": 0.18442107737064362,
      "learning_rate": 4.39204e-05,
      "loss": 0.2284,
      "step": 390200
    },
    {
      "epoch": 0.0806,
      "grad_norm": 0.18602213263511658,
      "learning_rate": 4.388040000000001e-05,
      "loss": 0.2363,
      "step": 390300
    },
    {
      "epoch": 0.0808,
      "grad_norm": 0.17077329754829407,
      "learning_rate": 4.38404e-05,
      "loss": 0.2298,
      "step": 390400
    },
    {
      "epoch": 0.081,
      "grad_norm": 0.20275497436523438,
      "learning_rate": 4.38004e-05,
      "loss": 0.2323,
      "step": 390500
    },
    {
      "epoch": 0.0812,
      "grad_norm": 0.17397406697273254,
      "learning_rate": 4.37604e-05,
      "loss": 0.2261,
      "step": 390600
    },
    {
      "epoch": 0.0814,
      "grad_norm": 0.14256907999515533,
      "learning_rate": 4.37204e-05,
      "loss": 0.2278,
      "step": 390700
    },
    {
      "epoch": 0.0816,
      "grad_norm": 0.1629251092672348,
      "learning_rate": 4.36804e-05,
      "loss": 0.2259,
      "step": 390800
    },
    {
      "epoch": 0.0818,
      "grad_norm": 0.2366534024477005,
      "learning_rate": 4.36404e-05,
      "loss": 0.2315,
      "step": 390900
    },
    {
      "epoch": 0.082,
      "grad_norm": 0.16596932709217072,
      "learning_rate": 4.3600400000000004e-05,
      "loss": 0.2259,
      "step": 391000
    },
    {
      "epoch": 0.0822,
      "grad_norm": 0.1851380616426468,
      "learning_rate": 4.3560400000000004e-05,
      "loss": 0.2209,
      "step": 391100
    },
    {
      "epoch": 0.0824,
      "grad_norm": 0.1714228391647339,
      "learning_rate": 4.35204e-05,
      "loss": 0.2256,
      "step": 391200
    },
    {
      "epoch": 0.0826,
      "grad_norm": 0.14281560480594635,
      "learning_rate": 4.3480400000000005e-05,
      "loss": 0.2262,
      "step": 391300
    },
    {
      "epoch": 0.0828,
      "grad_norm": 0.17006812989711761,
      "learning_rate": 4.3440400000000005e-05,
      "loss": 0.2278,
      "step": 391400
    },
    {
      "epoch": 0.083,
      "grad_norm": 0.2721594572067261,
      "learning_rate": 4.3400400000000005e-05,
      "loss": 0.2244,
      "step": 391500
    },
    {
      "epoch": 0.0832,
      "grad_norm": 0.15255306661128998,
      "learning_rate": 4.33604e-05,
      "loss": 0.2265,
      "step": 391600
    },
    {
      "epoch": 0.0834,
      "grad_norm": 0.1884760856628418,
      "learning_rate": 4.33204e-05,
      "loss": 0.2287,
      "step": 391700
    },
    {
      "epoch": 0.0836,
      "grad_norm": 0.3142029941082001,
      "learning_rate": 4.328040000000001e-05,
      "loss": 0.227,
      "step": 391800
    },
    {
      "epoch": 0.0838,
      "grad_norm": 0.19981245696544647,
      "learning_rate": 4.32404e-05,
      "loss": 0.2277,
      "step": 391900
    },
    {
      "epoch": 0.084,
      "grad_norm": 0.206933856010437,
      "learning_rate": 4.32004e-05,
      "loss": 0.2298,
      "step": 392000
    },
    {
      "epoch": 0.0842,
      "grad_norm": 0.17843426764011383,
      "learning_rate": 4.31604e-05,
      "loss": 0.2258,
      "step": 392100
    },
    {
      "epoch": 0.0844,
      "grad_norm": 0.18036629259586334,
      "learning_rate": 4.31204e-05,
      "loss": 0.2289,
      "step": 392200
    },
    {
      "epoch": 0.0846,
      "grad_norm": 0.17636817693710327,
      "learning_rate": 4.30804e-05,
      "loss": 0.226,
      "step": 392300
    },
    {
      "epoch": 0.0848,
      "grad_norm": 0.1490882784128189,
      "learning_rate": 4.30404e-05,
      "loss": 0.2267,
      "step": 392400
    },
    {
      "epoch": 0.085,
      "grad_norm": 0.16680997610092163,
      "learning_rate": 4.30004e-05,
      "loss": 0.226,
      "step": 392500
    },
    {
      "epoch": 0.0852,
      "grad_norm": 0.1657077521085739,
      "learning_rate": 4.29604e-05,
      "loss": 0.2267,
      "step": 392600
    },
    {
      "epoch": 0.0854,
      "grad_norm": 0.2529284656047821,
      "learning_rate": 4.2920399999999997e-05,
      "loss": 0.2281,
      "step": 392700
    },
    {
      "epoch": 0.0856,
      "grad_norm": 0.16982971131801605,
      "learning_rate": 4.2880400000000004e-05,
      "loss": 0.2256,
      "step": 392800
    },
    {
      "epoch": 0.0858,
      "grad_norm": 0.14875228703022003,
      "learning_rate": 4.2840400000000004e-05,
      "loss": 0.2272,
      "step": 392900
    },
    {
      "epoch": 0.086,
      "grad_norm": 0.2149345576763153,
      "learning_rate": 4.2800400000000004e-05,
      "loss": 0.2302,
      "step": 393000
    },
    {
      "epoch": 0.0862,
      "grad_norm": 0.17671360075473785,
      "learning_rate": 4.27604e-05,
      "loss": 0.2254,
      "step": 393100
    },
    {
      "epoch": 0.0864,
      "grad_norm": 0.17670981585979462,
      "learning_rate": 4.27204e-05,
      "loss": 0.2294,
      "step": 393200
    },
    {
      "epoch": 0.0866,
      "grad_norm": 0.16914287209510803,
      "learning_rate": 4.2680400000000006e-05,
      "loss": 0.2436,
      "step": 393300
    },
    {
      "epoch": 0.0868,
      "grad_norm": 0.17037755250930786,
      "learning_rate": 4.26404e-05,
      "loss": 0.2279,
      "step": 393400
    },
    {
      "epoch": 0.087,
      "grad_norm": 0.1500585824251175,
      "learning_rate": 4.26004e-05,
      "loss": 0.2252,
      "step": 393500
    },
    {
      "epoch": 0.0872,
      "grad_norm": 0.1869862973690033,
      "learning_rate": 4.25604e-05,
      "loss": 0.2271,
      "step": 393600
    },
    {
      "epoch": 0.0874,
      "grad_norm": 0.16152675449848175,
      "learning_rate": 4.252040000000001e-05,
      "loss": 0.229,
      "step": 393700
    },
    {
      "epoch": 0.0876,
      "grad_norm": 0.19350165128707886,
      "learning_rate": 4.24804e-05,
      "loss": 0.2271,
      "step": 393800
    },
    {
      "epoch": 0.0878,
      "grad_norm": 0.16843397915363312,
      "learning_rate": 4.24404e-05,
      "loss": 0.2272,
      "step": 393900
    },
    {
      "epoch": 0.088,
      "grad_norm": 0.17295992374420166,
      "learning_rate": 4.24004e-05,
      "loss": 0.2252,
      "step": 394000
    },
    {
      "epoch": 0.0882,
      "grad_norm": 0.2227911800146103,
      "learning_rate": 4.23604e-05,
      "loss": 0.2327,
      "step": 394100
    },
    {
      "epoch": 0.0884,
      "grad_norm": 0.22953267395496368,
      "learning_rate": 4.23204e-05,
      "loss": 0.2279,
      "step": 394200
    },
    {
      "epoch": 0.0886,
      "grad_norm": 0.22172044217586517,
      "learning_rate": 4.22804e-05,
      "loss": 0.2307,
      "step": 394300
    },
    {
      "epoch": 0.0888,
      "grad_norm": 0.1599377691745758,
      "learning_rate": 4.22404e-05,
      "loss": 0.2273,
      "step": 394400
    },
    {
      "epoch": 0.089,
      "grad_norm": 0.1549975574016571,
      "learning_rate": 4.2200400000000003e-05,
      "loss": 0.2267,
      "step": 394500
    },
    {
      "epoch": 0.0892,
      "grad_norm": 0.1666010618209839,
      "learning_rate": 4.21604e-05,
      "loss": 0.226,
      "step": 394600
    },
    {
      "epoch": 0.0894,
      "grad_norm": 0.1673179715871811,
      "learning_rate": 4.2120400000000004e-05,
      "loss": 0.2289,
      "step": 394700
    },
    {
      "epoch": 0.0896,
      "grad_norm": 0.16998769342899323,
      "learning_rate": 4.2080400000000005e-05,
      "loss": 0.2295,
      "step": 394800
    },
    {
      "epoch": 0.0898,
      "grad_norm": 0.14935901761054993,
      "learning_rate": 4.2040400000000005e-05,
      "loss": 0.2252,
      "step": 394900
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.14684082567691803,
      "learning_rate": 4.20004e-05,
      "loss": 0.2277,
      "step": 395000
    },
    {
      "epoch": 0.0902,
      "grad_norm": 0.2135782241821289,
      "learning_rate": 4.19604e-05,
      "loss": 0.2292,
      "step": 395100
    },
    {
      "epoch": 0.0904,
      "grad_norm": 0.19524498283863068,
      "learning_rate": 4.1920400000000006e-05,
      "loss": 0.2283,
      "step": 395200
    },
    {
      "epoch": 0.0906,
      "grad_norm": 0.19192557036876678,
      "learning_rate": 4.18804e-05,
      "loss": 0.2271,
      "step": 395300
    },
    {
      "epoch": 0.0908,
      "grad_norm": 0.1547107994556427,
      "learning_rate": 4.18404e-05,
      "loss": 0.2292,
      "step": 395400
    },
    {
      "epoch": 0.091,
      "grad_norm": 0.15521804988384247,
      "learning_rate": 4.18004e-05,
      "loss": 0.2262,
      "step": 395500
    },
    {
      "epoch": 0.0912,
      "grad_norm": 0.17244058847427368,
      "learning_rate": 4.17604e-05,
      "loss": 0.226,
      "step": 395600
    },
    {
      "epoch": 0.0914,
      "grad_norm": 0.17803426086902618,
      "learning_rate": 4.17204e-05,
      "loss": 0.2264,
      "step": 395700
    },
    {
      "epoch": 0.0916,
      "grad_norm": 0.17432081699371338,
      "learning_rate": 4.16804e-05,
      "loss": 0.2299,
      "step": 395800
    },
    {
      "epoch": 0.0918,
      "grad_norm": 0.23937202990055084,
      "learning_rate": 4.16404e-05,
      "loss": 0.2252,
      "step": 395900
    },
    {
      "epoch": 0.092,
      "grad_norm": 0.17643794417381287,
      "learning_rate": 4.16004e-05,
      "loss": 0.2253,
      "step": 396000
    },
    {
      "epoch": 0.0922,
      "grad_norm": 0.1799926906824112,
      "learning_rate": 4.1560399999999996e-05,
      "loss": 0.2311,
      "step": 396100
    },
    {
      "epoch": 0.0924,
      "grad_norm": 0.14338967204093933,
      "learning_rate": 4.15204e-05,
      "loss": 0.2262,
      "step": 396200
    },
    {
      "epoch": 0.0926,
      "grad_norm": 0.21923920512199402,
      "learning_rate": 4.1480400000000004e-05,
      "loss": 0.2252,
      "step": 396300
    },
    {
      "epoch": 0.0928,
      "grad_norm": 0.1880873441696167,
      "learning_rate": 4.1440400000000004e-05,
      "loss": 0.2258,
      "step": 396400
    },
    {
      "epoch": 0.093,
      "grad_norm": 0.15461334586143494,
      "learning_rate": 4.14004e-05,
      "loss": 0.2256,
      "step": 396500
    },
    {
      "epoch": 0.0932,
      "grad_norm": 0.15690234303474426,
      "learning_rate": 4.1360400000000005e-05,
      "loss": 0.2257,
      "step": 396600
    },
    {
      "epoch": 0.0934,
      "grad_norm": 0.15058964490890503,
      "learning_rate": 4.1320400000000005e-05,
      "loss": 0.2286,
      "step": 396700
    },
    {
      "epoch": 0.0936,
      "grad_norm": 0.17528659105300903,
      "learning_rate": 4.1280400000000005e-05,
      "loss": 0.2278,
      "step": 396800
    },
    {
      "epoch": 0.0938,
      "grad_norm": 0.30279967188835144,
      "learning_rate": 4.12404e-05,
      "loss": 0.2242,
      "step": 396900
    },
    {
      "epoch": 0.094,
      "grad_norm": 0.19836954772472382,
      "learning_rate": 4.12004e-05,
      "loss": 0.2264,
      "step": 397000
    },
    {
      "epoch": 0.0942,
      "grad_norm": 0.1697409451007843,
      "learning_rate": 4.1160400000000007e-05,
      "loss": 0.2273,
      "step": 397100
    },
    {
      "epoch": 0.0944,
      "grad_norm": 0.1903415024280548,
      "learning_rate": 4.11204e-05,
      "loss": 0.2318,
      "step": 397200
    },
    {
      "epoch": 0.0946,
      "grad_norm": 0.16050055623054504,
      "learning_rate": 4.10804e-05,
      "loss": 0.229,
      "step": 397300
    },
    {
      "epoch": 0.0948,
      "grad_norm": 0.1549975872039795,
      "learning_rate": 4.10404e-05,
      "loss": 0.2235,
      "step": 397400
    },
    {
      "epoch": 0.095,
      "grad_norm": 0.22184526920318604,
      "learning_rate": 4.10004e-05,
      "loss": 0.2278,
      "step": 397500
    },
    {
      "epoch": 0.0952,
      "grad_norm": 0.16259317100048065,
      "learning_rate": 4.09604e-05,
      "loss": 0.2276,
      "step": 397600
    },
    {
      "epoch": 0.0954,
      "grad_norm": 0.20985530316829681,
      "learning_rate": 4.09204e-05,
      "loss": 0.227,
      "step": 397700
    },
    {
      "epoch": 0.0956,
      "grad_norm": 0.1478949785232544,
      "learning_rate": 4.08804e-05,
      "loss": 0.231,
      "step": 397800
    },
    {
      "epoch": 0.0958,
      "grad_norm": 0.20481759309768677,
      "learning_rate": 4.08404e-05,
      "loss": 0.2252,
      "step": 397900
    },
    {
      "epoch": 0.096,
      "grad_norm": 0.16993680596351624,
      "learning_rate": 4.0800399999999996e-05,
      "loss": 0.2241,
      "step": 398000
    },
    {
      "epoch": 0.0962,
      "grad_norm": 0.2794969379901886,
      "learning_rate": 4.0760400000000004e-05,
      "loss": 0.2303,
      "step": 398100
    },
    {
      "epoch": 0.0964,
      "grad_norm": 0.1929108202457428,
      "learning_rate": 4.0720400000000004e-05,
      "loss": 0.2284,
      "step": 398200
    },
    {
      "epoch": 0.0966,
      "grad_norm": 0.16150234639644623,
      "learning_rate": 4.0680400000000004e-05,
      "loss": 0.2217,
      "step": 398300
    },
    {
      "epoch": 0.0968,
      "grad_norm": 0.15625301003456116,
      "learning_rate": 4.06404e-05,
      "loss": 0.2255,
      "step": 398400
    },
    {
      "epoch": 0.097,
      "grad_norm": 0.16894198954105377,
      "learning_rate": 4.06004e-05,
      "loss": 0.2228,
      "step": 398500
    },
    {
      "epoch": 0.0972,
      "grad_norm": 0.17462077736854553,
      "learning_rate": 4.0560400000000006e-05,
      "loss": 0.2255,
      "step": 398600
    },
    {
      "epoch": 0.0974,
      "grad_norm": 0.2863595187664032,
      "learning_rate": 4.0520400000000006e-05,
      "loss": 0.2272,
      "step": 398700
    },
    {
      "epoch": 0.0976,
      "grad_norm": 0.16492559015750885,
      "learning_rate": 4.04804e-05,
      "loss": 0.2275,
      "step": 398800
    },
    {
      "epoch": 0.0978,
      "grad_norm": 0.14487357437610626,
      "learning_rate": 4.04404e-05,
      "loss": 0.2248,
      "step": 398900
    },
    {
      "epoch": 0.098,
      "grad_norm": 0.22204236686229706,
      "learning_rate": 4.04004e-05,
      "loss": 0.2293,
      "step": 399000
    },
    {
      "epoch": 0.0982,
      "grad_norm": 0.16108718514442444,
      "learning_rate": 4.03604e-05,
      "loss": 0.2224,
      "step": 399100
    },
    {
      "epoch": 0.0984,
      "grad_norm": 0.1740008294582367,
      "learning_rate": 4.03204e-05,
      "loss": 0.2246,
      "step": 399200
    },
    {
      "epoch": 0.0986,
      "grad_norm": 0.1709638088941574,
      "learning_rate": 4.02804e-05,
      "loss": 0.2249,
      "step": 399300
    },
    {
      "epoch": 0.0988,
      "grad_norm": 0.2050946056842804,
      "learning_rate": 4.02404e-05,
      "loss": 0.2158,
      "step": 399400
    },
    {
      "epoch": 0.099,
      "grad_norm": 0.1883661299943924,
      "learning_rate": 4.02004e-05,
      "loss": 0.226,
      "step": 399500
    },
    {
      "epoch": 0.0992,
      "grad_norm": 0.17286445200443268,
      "learning_rate": 4.01604e-05,
      "loss": 0.224,
      "step": 399600
    },
    {
      "epoch": 0.0994,
      "grad_norm": 0.13540473580360413,
      "learning_rate": 4.01204e-05,
      "loss": 0.2284,
      "step": 399700
    },
    {
      "epoch": 0.0996,
      "grad_norm": 0.1601712554693222,
      "learning_rate": 4.00804e-05,
      "loss": 0.2252,
      "step": 399800
    },
    {
      "epoch": 0.0998,
      "grad_norm": 0.2265101969242096,
      "learning_rate": 4.00404e-05,
      "loss": 0.2211,
      "step": 399900
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.18272989988327026,
      "learning_rate": 4.0000400000000004e-05,
      "loss": 0.2229,
      "step": 400000
    },
    {
      "epoch": 0.1002,
      "grad_norm": 0.18946868181228638,
      "learning_rate": 3.9960400000000005e-05,
      "loss": 0.2258,
      "step": 400100
    },
    {
      "epoch": 0.1004,
      "grad_norm": 0.1850133091211319,
      "learning_rate": 3.9920400000000005e-05,
      "loss": 0.2277,
      "step": 400200
    },
    {
      "epoch": 0.1006,
      "grad_norm": 0.18626666069030762,
      "learning_rate": 3.98804e-05,
      "loss": 0.2251,
      "step": 400300
    },
    {
      "epoch": 0.1008,
      "grad_norm": 0.18756605684757233,
      "learning_rate": 3.98404e-05,
      "loss": 0.2252,
      "step": 400400
    },
    {
      "epoch": 0.101,
      "grad_norm": 0.17669208347797394,
      "learning_rate": 3.9800400000000006e-05,
      "loss": 0.2292,
      "step": 400500
    },
    {
      "epoch": 0.1012,
      "grad_norm": 0.1939968466758728,
      "learning_rate": 3.9760400000000006e-05,
      "loss": 0.2247,
      "step": 400600
    },
    {
      "epoch": 0.1014,
      "grad_norm": 0.18509630858898163,
      "learning_rate": 3.97204e-05,
      "loss": 0.226,
      "step": 400700
    },
    {
      "epoch": 0.1016,
      "grad_norm": 0.20480674505233765,
      "learning_rate": 3.96804e-05,
      "loss": 0.2238,
      "step": 400800
    },
    {
      "epoch": 0.1018,
      "grad_norm": 0.21943143010139465,
      "learning_rate": 3.96404e-05,
      "loss": 0.2237,
      "step": 400900
    },
    {
      "epoch": 0.102,
      "grad_norm": 0.15782029926776886,
      "learning_rate": 3.96004e-05,
      "loss": 0.2272,
      "step": 401000
    },
    {
      "epoch": 0.1022,
      "grad_norm": 0.19730308651924133,
      "learning_rate": 3.95604e-05,
      "loss": 0.2259,
      "step": 401100
    },
    {
      "epoch": 0.1024,
      "grad_norm": 0.18070414662361145,
      "learning_rate": 3.95204e-05,
      "loss": 0.2232,
      "step": 401200
    },
    {
      "epoch": 0.1026,
      "grad_norm": 0.14294645190238953,
      "learning_rate": 3.94804e-05,
      "loss": 0.2255,
      "step": 401300
    },
    {
      "epoch": 0.1028,
      "grad_norm": 0.19923175871372223,
      "learning_rate": 3.94404e-05,
      "loss": 0.2247,
      "step": 401400
    },
    {
      "epoch": 0.103,
      "grad_norm": 0.15311995148658752,
      "learning_rate": 3.94004e-05,
      "loss": 0.2227,
      "step": 401500
    },
    {
      "epoch": 0.1032,
      "grad_norm": 0.181657612323761,
      "learning_rate": 3.9360400000000004e-05,
      "loss": 0.2258,
      "step": 401600
    },
    {
      "epoch": 0.1034,
      "grad_norm": 0.2274169921875,
      "learning_rate": 3.9320400000000004e-05,
      "loss": 0.2317,
      "step": 401700
    },
    {
      "epoch": 0.1036,
      "grad_norm": 0.2559068202972412,
      "learning_rate": 3.92804e-05,
      "loss": 0.2246,
      "step": 401800
    },
    {
      "epoch": 0.1038,
      "grad_norm": 0.1480039358139038,
      "learning_rate": 3.92404e-05,
      "loss": 0.2339,
      "step": 401900
    },
    {
      "epoch": 0.104,
      "grad_norm": 0.1887112855911255,
      "learning_rate": 3.9200400000000005e-05,
      "loss": 0.2268,
      "step": 402000
    },
    {
      "epoch": 0.1042,
      "grad_norm": 0.1991635113954544,
      "learning_rate": 3.9160400000000005e-05,
      "loss": 0.24,
      "step": 402100
    },
    {
      "epoch": 0.1044,
      "grad_norm": 0.19259975850582123,
      "learning_rate": 3.91204e-05,
      "loss": 0.2319,
      "step": 402200
    },
    {
      "epoch": 0.1046,
      "grad_norm": 0.16919143497943878,
      "learning_rate": 3.90804e-05,
      "loss": 0.2268,
      "step": 402300
    },
    {
      "epoch": 0.1048,
      "grad_norm": 0.24544304609298706,
      "learning_rate": 3.90404e-05,
      "loss": 0.2299,
      "step": 402400
    },
    {
      "epoch": 0.105,
      "grad_norm": 0.21261242032051086,
      "learning_rate": 3.900040000000001e-05,
      "loss": 0.2239,
      "step": 402500
    },
    {
      "epoch": 0.1052,
      "grad_norm": 0.19854722917079926,
      "learning_rate": 3.89604e-05,
      "loss": 0.2317,
      "step": 402600
    },
    {
      "epoch": 0.1054,
      "grad_norm": 0.29832783341407776,
      "learning_rate": 3.89204e-05,
      "loss": 0.2272,
      "step": 402700
    },
    {
      "epoch": 0.1056,
      "grad_norm": 0.18028579652309418,
      "learning_rate": 3.88804e-05,
      "loss": 0.2314,
      "step": 402800
    },
    {
      "epoch": 0.1058,
      "grad_norm": 0.18919360637664795,
      "learning_rate": 3.88404e-05,
      "loss": 0.224,
      "step": 402900
    },
    {
      "epoch": 0.106,
      "grad_norm": 0.21304070949554443,
      "learning_rate": 3.88004e-05,
      "loss": 0.2299,
      "step": 403000
    },
    {
      "epoch": 0.1062,
      "grad_norm": 0.18480710685253143,
      "learning_rate": 3.87604e-05,
      "loss": 0.2302,
      "step": 403100
    },
    {
      "epoch": 0.1064,
      "grad_norm": 0.38507312536239624,
      "learning_rate": 3.87204e-05,
      "loss": 0.2338,
      "step": 403200
    },
    {
      "epoch": 0.1066,
      "grad_norm": 0.15473273396492004,
      "learning_rate": 3.86804e-05,
      "loss": 0.226,
      "step": 403300
    },
    {
      "epoch": 0.1068,
      "grad_norm": 0.22138014435768127,
      "learning_rate": 3.8640400000000004e-05,
      "loss": 0.2265,
      "step": 403400
    },
    {
      "epoch": 0.107,
      "grad_norm": 0.22224438190460205,
      "learning_rate": 3.8600400000000004e-05,
      "loss": 0.2282,
      "step": 403500
    },
    {
      "epoch": 0.1072,
      "grad_norm": 0.16137254238128662,
      "learning_rate": 3.8560400000000004e-05,
      "loss": 0.2282,
      "step": 403600
    },
    {
      "epoch": 0.1074,
      "grad_norm": 0.18113075196743011,
      "learning_rate": 3.85204e-05,
      "loss": 0.2261,
      "step": 403700
    },
    {
      "epoch": 0.1076,
      "grad_norm": 0.17297004163265228,
      "learning_rate": 3.84804e-05,
      "loss": 0.2268,
      "step": 403800
    },
    {
      "epoch": 0.1078,
      "grad_norm": 0.20817789435386658,
      "learning_rate": 3.8440400000000006e-05,
      "loss": 0.2258,
      "step": 403900
    },
    {
      "epoch": 0.108,
      "grad_norm": 0.20474325120449066,
      "learning_rate": 3.8400400000000006e-05,
      "loss": 0.2257,
      "step": 404000
    },
    {
      "epoch": 0.1082,
      "grad_norm": 0.15835532546043396,
      "learning_rate": 3.83604e-05,
      "loss": 0.227,
      "step": 404100
    },
    {
      "epoch": 0.1084,
      "grad_norm": 0.17614057660102844,
      "learning_rate": 3.83204e-05,
      "loss": 0.226,
      "step": 404200
    },
    {
      "epoch": 0.1086,
      "grad_norm": 0.17774057388305664,
      "learning_rate": 3.82804e-05,
      "loss": 0.225,
      "step": 404300
    },
    {
      "epoch": 0.1088,
      "grad_norm": 0.1751686930656433,
      "learning_rate": 3.82404e-05,
      "loss": 0.2274,
      "step": 404400
    },
    {
      "epoch": 0.109,
      "grad_norm": 0.20452751219272614,
      "learning_rate": 3.82004e-05,
      "loss": 0.2278,
      "step": 404500
    },
    {
      "epoch": 0.1092,
      "grad_norm": 0.1584174484014511,
      "learning_rate": 3.81604e-05,
      "loss": 0.2296,
      "step": 404600
    },
    {
      "epoch": 0.1094,
      "grad_norm": 0.23247212171554565,
      "learning_rate": 3.81204e-05,
      "loss": 0.2265,
      "step": 404700
    },
    {
      "epoch": 0.1096,
      "grad_norm": 0.19274774193763733,
      "learning_rate": 3.80804e-05,
      "loss": 0.2277,
      "step": 404800
    },
    {
      "epoch": 0.1098,
      "grad_norm": 0.19633884727954865,
      "learning_rate": 3.80404e-05,
      "loss": 0.227,
      "step": 404900
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.18933461606502533,
      "learning_rate": 3.80004e-05,
      "loss": 0.2295,
      "step": 405000
    },
    {
      "epoch": 0.1102,
      "grad_norm": 0.20844264328479767,
      "learning_rate": 3.79604e-05,
      "loss": 0.2282,
      "step": 405100
    },
    {
      "epoch": 0.1104,
      "grad_norm": 0.2164926677942276,
      "learning_rate": 3.7920400000000004e-05,
      "loss": 0.2304,
      "step": 405200
    },
    {
      "epoch": 0.1106,
      "grad_norm": 0.16100728511810303,
      "learning_rate": 3.78804e-05,
      "loss": 0.2245,
      "step": 405300
    },
    {
      "epoch": 0.1108,
      "grad_norm": 0.18901631236076355,
      "learning_rate": 3.7840400000000005e-05,
      "loss": 0.2268,
      "step": 405400
    },
    {
      "epoch": 0.111,
      "grad_norm": 0.18958155810832977,
      "learning_rate": 3.7800400000000005e-05,
      "loss": 0.2286,
      "step": 405500
    },
    {
      "epoch": 0.1112,
      "grad_norm": 0.1504233032464981,
      "learning_rate": 3.77604e-05,
      "loss": 0.2325,
      "step": 405600
    },
    {
      "epoch": 0.1114,
      "grad_norm": 0.18149876594543457,
      "learning_rate": 3.77204e-05,
      "loss": 0.2237,
      "step": 405700
    },
    {
      "epoch": 0.1116,
      "grad_norm": 0.16026699542999268,
      "learning_rate": 3.76804e-05,
      "loss": 0.2266,
      "step": 405800
    },
    {
      "epoch": 0.1118,
      "grad_norm": 0.17877326905727386,
      "learning_rate": 3.7640400000000006e-05,
      "loss": 0.2239,
      "step": 405900
    },
    {
      "epoch": 0.112,
      "grad_norm": 0.20097322762012482,
      "learning_rate": 3.76004e-05,
      "loss": 0.2278,
      "step": 406000
    },
    {
      "epoch": 0.1122,
      "grad_norm": 0.21661312878131866,
      "learning_rate": 3.75604e-05,
      "loss": 0.2278,
      "step": 406100
    },
    {
      "epoch": 0.1124,
      "grad_norm": 0.1761494129896164,
      "learning_rate": 3.75204e-05,
      "loss": 0.2249,
      "step": 406200
    },
    {
      "epoch": 0.1126,
      "grad_norm": 0.6409944295883179,
      "learning_rate": 3.74804e-05,
      "loss": 0.2274,
      "step": 406300
    },
    {
      "epoch": 0.1128,
      "grad_norm": 0.14468325674533844,
      "learning_rate": 3.74404e-05,
      "loss": 0.2252,
      "step": 406400
    },
    {
      "epoch": 0.113,
      "grad_norm": 0.1829795092344284,
      "learning_rate": 3.74004e-05,
      "loss": 0.2213,
      "step": 406500
    },
    {
      "epoch": 0.1132,
      "grad_norm": 0.18367330729961395,
      "learning_rate": 3.73604e-05,
      "loss": 0.2284,
      "step": 406600
    },
    {
      "epoch": 0.1134,
      "grad_norm": 0.16368336975574493,
      "learning_rate": 3.73204e-05,
      "loss": 0.2256,
      "step": 406700
    },
    {
      "epoch": 0.1136,
      "grad_norm": 0.15827390551567078,
      "learning_rate": 3.72804e-05,
      "loss": 0.2226,
      "step": 406800
    },
    {
      "epoch": 0.1138,
      "grad_norm": 0.23074659705162048,
      "learning_rate": 3.7240400000000003e-05,
      "loss": 0.2257,
      "step": 406900
    },
    {
      "epoch": 0.114,
      "grad_norm": 0.15967126190662384,
      "learning_rate": 3.7200400000000004e-05,
      "loss": 0.2229,
      "step": 407000
    },
    {
      "epoch": 0.1142,
      "grad_norm": 0.20047785341739655,
      "learning_rate": 3.71604e-05,
      "loss": 0.2253,
      "step": 407100
    },
    {
      "epoch": 0.1144,
      "grad_norm": 0.2328701764345169,
      "learning_rate": 3.71204e-05,
      "loss": 0.2216,
      "step": 407200
    },
    {
      "epoch": 0.1146,
      "grad_norm": 0.2147849202156067,
      "learning_rate": 3.7080400000000005e-05,
      "loss": 0.2309,
      "step": 407300
    },
    {
      "epoch": 0.1148,
      "grad_norm": 0.2192113846540451,
      "learning_rate": 3.7040400000000005e-05,
      "loss": 0.2279,
      "step": 407400
    },
    {
      "epoch": 0.115,
      "grad_norm": 0.1873077005147934,
      "learning_rate": 3.70004e-05,
      "loss": 0.2245,
      "step": 407500
    },
    {
      "epoch": 0.1152,
      "grad_norm": 0.16018076241016388,
      "learning_rate": 3.69604e-05,
      "loss": 0.2375,
      "step": 407600
    },
    {
      "epoch": 0.1154,
      "grad_norm": 0.3920210897922516,
      "learning_rate": 3.69204e-05,
      "loss": 0.224,
      "step": 407700
    },
    {
      "epoch": 0.1156,
      "grad_norm": 0.19828300178050995,
      "learning_rate": 3.688040000000001e-05,
      "loss": 0.232,
      "step": 407800
    },
    {
      "epoch": 0.1158,
      "grad_norm": 0.17930054664611816,
      "learning_rate": 3.68404e-05,
      "loss": 0.2226,
      "step": 407900
    },
    {
      "epoch": 0.116,
      "grad_norm": 0.16561190783977509,
      "learning_rate": 3.68004e-05,
      "loss": 0.2236,
      "step": 408000
    },
    {
      "epoch": 0.1162,
      "grad_norm": 0.20231448113918304,
      "learning_rate": 3.67604e-05,
      "loss": 0.2238,
      "step": 408100
    },
    {
      "epoch": 0.1164,
      "grad_norm": 0.1568211019039154,
      "learning_rate": 3.67204e-05,
      "loss": 0.2274,
      "step": 408200
    },
    {
      "epoch": 0.1166,
      "grad_norm": 0.2083185911178589,
      "learning_rate": 3.66804e-05,
      "loss": 0.2217,
      "step": 408300
    },
    {
      "epoch": 0.1168,
      "grad_norm": 0.20765464007854462,
      "learning_rate": 3.66404e-05,
      "loss": 0.2261,
      "step": 408400
    },
    {
      "epoch": 0.117,
      "grad_norm": 0.15366986393928528,
      "learning_rate": 3.66004e-05,
      "loss": 0.2243,
      "step": 408500
    },
    {
      "epoch": 0.1172,
      "grad_norm": 0.1394629180431366,
      "learning_rate": 3.65604e-05,
      "loss": 0.2254,
      "step": 408600
    },
    {
      "epoch": 0.1174,
      "grad_norm": 0.19688405096530914,
      "learning_rate": 3.65204e-05,
      "loss": 0.2244,
      "step": 408700
    },
    {
      "epoch": 0.1176,
      "grad_norm": 0.2001643031835556,
      "learning_rate": 3.6480400000000004e-05,
      "loss": 0.2238,
      "step": 408800
    },
    {
      "epoch": 0.1178,
      "grad_norm": 0.17548991739749908,
      "learning_rate": 3.6440400000000004e-05,
      "loss": 0.2296,
      "step": 408900
    },
    {
      "epoch": 0.118,
      "grad_norm": 0.1478511244058609,
      "learning_rate": 3.64004e-05,
      "loss": 0.2272,
      "step": 409000
    },
    {
      "epoch": 0.1182,
      "grad_norm": 0.17043019831180573,
      "learning_rate": 3.63604e-05,
      "loss": 0.2227,
      "step": 409100
    },
    {
      "epoch": 0.1184,
      "grad_norm": 0.16290804743766785,
      "learning_rate": 3.6320400000000006e-05,
      "loss": 0.2258,
      "step": 409200
    },
    {
      "epoch": 0.1186,
      "grad_norm": 0.17892128229141235,
      "learning_rate": 3.6280400000000006e-05,
      "loss": 0.2222,
      "step": 409300
    },
    {
      "epoch": 0.1188,
      "grad_norm": 0.14661169052124023,
      "learning_rate": 3.62404e-05,
      "loss": 0.2259,
      "step": 409400
    },
    {
      "epoch": 0.119,
      "grad_norm": 0.21006256341934204,
      "learning_rate": 3.62004e-05,
      "loss": 0.2244,
      "step": 409500
    },
    {
      "epoch": 0.1192,
      "grad_norm": 0.2831527590751648,
      "learning_rate": 3.61604e-05,
      "loss": 0.2251,
      "step": 409600
    },
    {
      "epoch": 0.1194,
      "grad_norm": 0.1787988245487213,
      "learning_rate": 3.612040000000001e-05,
      "loss": 0.2231,
      "step": 409700
    },
    {
      "epoch": 0.1196,
      "grad_norm": 0.17835716903209686,
      "learning_rate": 3.60804e-05,
      "loss": 0.2278,
      "step": 409800
    },
    {
      "epoch": 0.1198,
      "grad_norm": 0.1802866905927658,
      "learning_rate": 3.60404e-05,
      "loss": 0.225,
      "step": 409900
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.15969882905483246,
      "learning_rate": 3.60004e-05,
      "loss": 0.2233,
      "step": 410000
    },
    {
      "epoch": 0.1202,
      "grad_norm": 0.24545249342918396,
      "learning_rate": 3.59604e-05,
      "loss": 0.2259,
      "step": 410100
    },
    {
      "epoch": 0.1204,
      "grad_norm": 0.163286954164505,
      "learning_rate": 3.59204e-05,
      "loss": 0.2271,
      "step": 410200
    },
    {
      "epoch": 0.1206,
      "grad_norm": 0.16609875857830048,
      "learning_rate": 3.58804e-05,
      "loss": 0.2328,
      "step": 410300
    },
    {
      "epoch": 0.1208,
      "grad_norm": 0.1794077455997467,
      "learning_rate": 3.58404e-05,
      "loss": 0.2259,
      "step": 410400
    },
    {
      "epoch": 0.121,
      "grad_norm": 0.15229471027851105,
      "learning_rate": 3.5800400000000004e-05,
      "loss": 0.2235,
      "step": 410500
    },
    {
      "epoch": 0.1212,
      "grad_norm": 0.2652372419834137,
      "learning_rate": 3.57604e-05,
      "loss": 0.224,
      "step": 410600
    },
    {
      "epoch": 0.1214,
      "grad_norm": 0.1757022589445114,
      "learning_rate": 3.5720400000000004e-05,
      "loss": 0.2271,
      "step": 410700
    },
    {
      "epoch": 0.1216,
      "grad_norm": 0.1457105576992035,
      "learning_rate": 3.5680400000000005e-05,
      "loss": 0.2265,
      "step": 410800
    },
    {
      "epoch": 0.1218,
      "grad_norm": 0.15136027336120605,
      "learning_rate": 3.56404e-05,
      "loss": 0.2276,
      "step": 410900
    },
    {
      "epoch": 0.122,
      "grad_norm": 0.1632482409477234,
      "learning_rate": 3.56004e-05,
      "loss": 0.2273,
      "step": 411000
    },
    {
      "epoch": 0.1222,
      "grad_norm": 0.16236582398414612,
      "learning_rate": 3.55604e-05,
      "loss": 0.2215,
      "step": 411100
    },
    {
      "epoch": 0.1224,
      "grad_norm": 0.16312521696090698,
      "learning_rate": 3.5520400000000006e-05,
      "loss": 0.2289,
      "step": 411200
    },
    {
      "epoch": 0.1226,
      "grad_norm": 0.1699019968509674,
      "learning_rate": 3.54804e-05,
      "loss": 0.2303,
      "step": 411300
    },
    {
      "epoch": 0.1228,
      "grad_norm": 0.17259664833545685,
      "learning_rate": 3.54404e-05,
      "loss": 0.2262,
      "step": 411400
    },
    {
      "epoch": 0.123,
      "grad_norm": 0.18414495885372162,
      "learning_rate": 3.54004e-05,
      "loss": 0.2232,
      "step": 411500
    },
    {
      "epoch": 0.1232,
      "grad_norm": 0.17581482231616974,
      "learning_rate": 3.53604e-05,
      "loss": 0.2287,
      "step": 411600
    },
    {
      "epoch": 0.1234,
      "grad_norm": 0.18369857966899872,
      "learning_rate": 3.53204e-05,
      "loss": 0.2326,
      "step": 411700
    },
    {
      "epoch": 0.1236,
      "grad_norm": 0.22637276351451874,
      "learning_rate": 3.52804e-05,
      "loss": 0.2278,
      "step": 411800
    },
    {
      "epoch": 0.1238,
      "grad_norm": 0.14915701746940613,
      "learning_rate": 3.52404e-05,
      "loss": 0.2222,
      "step": 411900
    },
    {
      "epoch": 0.124,
      "grad_norm": 0.16789127886295319,
      "learning_rate": 3.52004e-05,
      "loss": 0.225,
      "step": 412000
    },
    {
      "epoch": 0.1242,
      "grad_norm": 0.16109804809093475,
      "learning_rate": 3.51604e-05,
      "loss": 0.2284,
      "step": 412100
    },
    {
      "epoch": 0.1244,
      "grad_norm": 0.19975192844867706,
      "learning_rate": 3.5120400000000003e-05,
      "loss": 0.2221,
      "step": 412200
    },
    {
      "epoch": 0.1246,
      "grad_norm": 0.18988633155822754,
      "learning_rate": 3.5080400000000004e-05,
      "loss": 0.227,
      "step": 412300
    },
    {
      "epoch": 0.1248,
      "grad_norm": 0.14888371527194977,
      "learning_rate": 3.5040400000000004e-05,
      "loss": 0.225,
      "step": 412400
    },
    {
      "epoch": 0.125,
      "grad_norm": 0.20460769534111023,
      "learning_rate": 3.50004e-05,
      "loss": 0.2202,
      "step": 412500
    },
    {
      "epoch": 0.1252,
      "grad_norm": 0.18129372596740723,
      "learning_rate": 3.4960400000000005e-05,
      "loss": 0.2261,
      "step": 412600
    },
    {
      "epoch": 0.1254,
      "grad_norm": 0.25600898265838623,
      "learning_rate": 3.4920400000000005e-05,
      "loss": 0.2264,
      "step": 412700
    },
    {
      "epoch": 0.1256,
      "grad_norm": 0.16887041926383972,
      "learning_rate": 3.48804e-05,
      "loss": 0.2247,
      "step": 412800
    },
    {
      "epoch": 0.1258,
      "grad_norm": 0.1853538155555725,
      "learning_rate": 3.48404e-05,
      "loss": 0.2283,
      "step": 412900
    },
    {
      "epoch": 0.126,
      "grad_norm": 0.17123030126094818,
      "learning_rate": 3.48004e-05,
      "loss": 0.2257,
      "step": 413000
    },
    {
      "epoch": 0.1262,
      "grad_norm": 0.1886633038520813,
      "learning_rate": 3.476040000000001e-05,
      "loss": 0.2263,
      "step": 413100
    },
    {
      "epoch": 0.1264,
      "grad_norm": 0.16311076283454895,
      "learning_rate": 3.47204e-05,
      "loss": 0.2247,
      "step": 413200
    },
    {
      "epoch": 0.1266,
      "grad_norm": 0.18546321988105774,
      "learning_rate": 3.46804e-05,
      "loss": 0.2287,
      "step": 413300
    },
    {
      "epoch": 0.1268,
      "grad_norm": 0.20453518629074097,
      "learning_rate": 3.46404e-05,
      "loss": 0.2248,
      "step": 413400
    },
    {
      "epoch": 0.127,
      "grad_norm": 0.16402287781238556,
      "learning_rate": 3.46004e-05,
      "loss": 0.2272,
      "step": 413500
    },
    {
      "epoch": 0.1272,
      "grad_norm": 0.31062790751457214,
      "learning_rate": 3.45604e-05,
      "loss": 0.2248,
      "step": 413600
    },
    {
      "epoch": 0.1274,
      "grad_norm": 0.1984662562608719,
      "learning_rate": 3.45204e-05,
      "loss": 0.2262,
      "step": 413700
    },
    {
      "epoch": 0.1276,
      "grad_norm": 0.3165430426597595,
      "learning_rate": 3.44804e-05,
      "loss": 0.2261,
      "step": 413800
    },
    {
      "epoch": 0.1278,
      "grad_norm": 0.1811901032924652,
      "learning_rate": 3.44404e-05,
      "loss": 0.227,
      "step": 413900
    },
    {
      "epoch": 0.128,
      "grad_norm": 0.15821123123168945,
      "learning_rate": 3.44004e-05,
      "loss": 0.2252,
      "step": 414000
    },
    {
      "epoch": 0.1282,
      "grad_norm": 0.15300792455673218,
      "learning_rate": 3.4360400000000004e-05,
      "loss": 0.2242,
      "step": 414100
    },
    {
      "epoch": 0.1284,
      "grad_norm": 0.2539973556995392,
      "learning_rate": 3.4320400000000004e-05,
      "loss": 0.2287,
      "step": 414200
    },
    {
      "epoch": 0.1286,
      "grad_norm": 0.1613328754901886,
      "learning_rate": 3.4280400000000005e-05,
      "loss": 0.2258,
      "step": 414300
    },
    {
      "epoch": 0.1288,
      "grad_norm": 0.27399247884750366,
      "learning_rate": 3.42404e-05,
      "loss": 0.2233,
      "step": 414400
    },
    {
      "epoch": 0.129,
      "grad_norm": 0.17201277613639832,
      "learning_rate": 3.42004e-05,
      "loss": 0.2243,
      "step": 414500
    },
    {
      "epoch": 0.1292,
      "grad_norm": 0.21430495381355286,
      "learning_rate": 3.4160400000000006e-05,
      "loss": 0.2295,
      "step": 414600
    },
    {
      "epoch": 0.1294,
      "grad_norm": 0.1703038066625595,
      "learning_rate": 3.41204e-05,
      "loss": 0.2274,
      "step": 414700
    },
    {
      "epoch": 0.1296,
      "grad_norm": 0.21057896316051483,
      "learning_rate": 3.40804e-05,
      "loss": 0.2275,
      "step": 414800
    },
    {
      "epoch": 0.1298,
      "grad_norm": 0.17118048667907715,
      "learning_rate": 3.40404e-05,
      "loss": 0.2261,
      "step": 414900
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.15217669308185577,
      "learning_rate": 3.40004e-05,
      "loss": 0.2286,
      "step": 415000
    },
    {
      "epoch": 0.1302,
      "grad_norm": 0.1708402931690216,
      "learning_rate": 3.39604e-05,
      "loss": 0.2276,
      "step": 415100
    },
    {
      "epoch": 0.1304,
      "grad_norm": 0.15635427832603455,
      "learning_rate": 3.39204e-05,
      "loss": 0.2271,
      "step": 415200
    },
    {
      "epoch": 0.1306,
      "grad_norm": 0.17114631831645966,
      "learning_rate": 3.38804e-05,
      "loss": 0.2289,
      "step": 415300
    },
    {
      "epoch": 0.1308,
      "grad_norm": 0.1455591470003128,
      "learning_rate": 3.38404e-05,
      "loss": 0.2314,
      "step": 415400
    },
    {
      "epoch": 0.131,
      "grad_norm": 0.19031578302383423,
      "learning_rate": 3.38004e-05,
      "loss": 0.2252,
      "step": 415500
    },
    {
      "epoch": 0.1312,
      "grad_norm": 0.19295114278793335,
      "learning_rate": 3.37604e-05,
      "loss": 0.2259,
      "step": 415600
    },
    {
      "epoch": 0.1314,
      "grad_norm": 0.18876852095127106,
      "learning_rate": 3.37204e-05,
      "loss": 0.2278,
      "step": 415700
    },
    {
      "epoch": 0.1316,
      "grad_norm": 0.20941956341266632,
      "learning_rate": 3.3680400000000004e-05,
      "loss": 0.2277,
      "step": 415800
    },
    {
      "epoch": 0.1318,
      "grad_norm": 0.16733889281749725,
      "learning_rate": 3.36404e-05,
      "loss": 0.2257,
      "step": 415900
    },
    {
      "epoch": 0.132,
      "grad_norm": 0.15858490765094757,
      "learning_rate": 3.3600400000000004e-05,
      "loss": 0.2228,
      "step": 416000
    },
    {
      "epoch": 0.1322,
      "grad_norm": 0.18554091453552246,
      "learning_rate": 3.3560400000000005e-05,
      "loss": 0.2359,
      "step": 416100
    },
    {
      "epoch": 0.1324,
      "grad_norm": 0.15781132876873016,
      "learning_rate": 3.35204e-05,
      "loss": 0.2273,
      "step": 416200
    },
    {
      "epoch": 0.1326,
      "grad_norm": 0.18581081926822662,
      "learning_rate": 3.34804e-05,
      "loss": 0.2252,
      "step": 416300
    },
    {
      "epoch": 0.1328,
      "grad_norm": 0.19088365137577057,
      "learning_rate": 3.34404e-05,
      "loss": 0.2281,
      "step": 416400
    },
    {
      "epoch": 0.133,
      "grad_norm": 0.1562078595161438,
      "learning_rate": 3.3400400000000006e-05,
      "loss": 0.2262,
      "step": 416500
    },
    {
      "epoch": 0.1332,
      "grad_norm": 0.18324877321720123,
      "learning_rate": 3.33604e-05,
      "loss": 0.2264,
      "step": 416600
    },
    {
      "epoch": 0.1334,
      "grad_norm": 0.1622045785188675,
      "learning_rate": 3.33204e-05,
      "loss": 0.2271,
      "step": 416700
    },
    {
      "epoch": 0.1336,
      "grad_norm": 0.16861198842525482,
      "learning_rate": 3.32804e-05,
      "loss": 0.2245,
      "step": 416800
    },
    {
      "epoch": 0.1338,
      "grad_norm": 0.20437948405742645,
      "learning_rate": 3.32404e-05,
      "loss": 0.2264,
      "step": 416900
    },
    {
      "epoch": 0.134,
      "grad_norm": 0.2063809335231781,
      "learning_rate": 3.32004e-05,
      "loss": 0.2249,
      "step": 417000
    },
    {
      "epoch": 0.1342,
      "grad_norm": 0.18307951092720032,
      "learning_rate": 3.31604e-05,
      "loss": 0.2324,
      "step": 417100
    },
    {
      "epoch": 0.1344,
      "grad_norm": 0.1892336905002594,
      "learning_rate": 3.31204e-05,
      "loss": 0.2248,
      "step": 417200
    },
    {
      "epoch": 0.1346,
      "grad_norm": 0.2223246991634369,
      "learning_rate": 3.30804e-05,
      "loss": 0.225,
      "step": 417300
    },
    {
      "epoch": 0.1348,
      "grad_norm": 0.15627120435237885,
      "learning_rate": 3.3040399999999996e-05,
      "loss": 0.2228,
      "step": 417400
    },
    {
      "epoch": 0.135,
      "grad_norm": 0.16463282704353333,
      "learning_rate": 3.3000400000000003e-05,
      "loss": 0.2257,
      "step": 417500
    },
    {
      "epoch": 0.1352,
      "grad_norm": 0.17951765656471252,
      "learning_rate": 3.2960400000000004e-05,
      "loss": 0.2234,
      "step": 417600
    },
    {
      "epoch": 0.1354,
      "grad_norm": 0.17068003118038177,
      "learning_rate": 3.2920400000000004e-05,
      "loss": 0.2244,
      "step": 417700
    },
    {
      "epoch": 0.1356,
      "grad_norm": 0.17062413692474365,
      "learning_rate": 3.28804e-05,
      "loss": 0.2224,
      "step": 417800
    },
    {
      "epoch": 0.1358,
      "grad_norm": 0.27390560507774353,
      "learning_rate": 3.28404e-05,
      "loss": 0.2274,
      "step": 417900
    },
    {
      "epoch": 0.136,
      "grad_norm": 0.15134724974632263,
      "learning_rate": 3.2800400000000005e-05,
      "loss": 0.2296,
      "step": 418000
    },
    {
      "epoch": 0.1362,
      "grad_norm": 0.161467045545578,
      "learning_rate": 3.27604e-05,
      "loss": 0.2216,
      "step": 418100
    },
    {
      "epoch": 0.1364,
      "grad_norm": 0.3517807126045227,
      "learning_rate": 3.27204e-05,
      "loss": 0.2276,
      "step": 418200
    },
    {
      "epoch": 0.1366,
      "grad_norm": 0.19215285778045654,
      "learning_rate": 3.26804e-05,
      "loss": 0.2307,
      "step": 418300
    },
    {
      "epoch": 0.1368,
      "grad_norm": 0.18861427903175354,
      "learning_rate": 3.264040000000001e-05,
      "loss": 0.2275,
      "step": 418400
    },
    {
      "epoch": 0.137,
      "grad_norm": 0.2437637597322464,
      "learning_rate": 3.26004e-05,
      "loss": 0.2244,
      "step": 418500
    },
    {
      "epoch": 0.1372,
      "grad_norm": 0.156779944896698,
      "learning_rate": 3.25604e-05,
      "loss": 0.227,
      "step": 418600
    },
    {
      "epoch": 0.1374,
      "grad_norm": 0.14431990683078766,
      "learning_rate": 3.25204e-05,
      "loss": 0.2256,
      "step": 418700
    },
    {
      "epoch": 0.1376,
      "grad_norm": 0.2503659427165985,
      "learning_rate": 3.24804e-05,
      "loss": 0.2234,
      "step": 418800
    },
    {
      "epoch": 0.1378,
      "grad_norm": 0.20400291681289673,
      "learning_rate": 3.24404e-05,
      "loss": 0.2221,
      "step": 418900
    },
    {
      "epoch": 0.138,
      "grad_norm": 0.16242775321006775,
      "learning_rate": 3.24004e-05,
      "loss": 0.2234,
      "step": 419000
    },
    {
      "epoch": 0.1382,
      "grad_norm": 0.18688417971134186,
      "learning_rate": 3.23604e-05,
      "loss": 0.2302,
      "step": 419100
    },
    {
      "epoch": 0.1384,
      "grad_norm": 0.18453970551490784,
      "learning_rate": 3.23204e-05,
      "loss": 0.2234,
      "step": 419200
    },
    {
      "epoch": 0.1386,
      "grad_norm": 0.1676274985074997,
      "learning_rate": 3.22804e-05,
      "loss": 0.2288,
      "step": 419300
    },
    {
      "epoch": 0.1388,
      "grad_norm": 0.21078824996948242,
      "learning_rate": 3.2240400000000004e-05,
      "loss": 0.2244,
      "step": 419400
    },
    {
      "epoch": 0.139,
      "grad_norm": 0.24954631924629211,
      "learning_rate": 3.2200400000000004e-05,
      "loss": 0.2242,
      "step": 419500
    },
    {
      "epoch": 0.1392,
      "grad_norm": 0.2090296894311905,
      "learning_rate": 3.2160400000000005e-05,
      "loss": 0.2219,
      "step": 419600
    },
    {
      "epoch": 0.1394,
      "grad_norm": 0.15903155505657196,
      "learning_rate": 3.21204e-05,
      "loss": 0.2287,
      "step": 419700
    },
    {
      "epoch": 0.1396,
      "grad_norm": 0.1801076978445053,
      "learning_rate": 3.20804e-05,
      "loss": 0.224,
      "step": 419800
    },
    {
      "epoch": 0.1398,
      "grad_norm": 0.14464084804058075,
      "learning_rate": 3.2040400000000006e-05,
      "loss": 0.2256,
      "step": 419900
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.1716356873512268,
      "learning_rate": 3.20004e-05,
      "loss": 0.2269,
      "step": 420000
    },
    {
      "epoch": 0.1402,
      "grad_norm": 0.2161165177822113,
      "learning_rate": 3.19604e-05,
      "loss": 0.2243,
      "step": 420100
    },
    {
      "epoch": 0.1404,
      "grad_norm": 0.2271643877029419,
      "learning_rate": 3.19204e-05,
      "loss": 0.2227,
      "step": 420200
    },
    {
      "epoch": 0.1406,
      "grad_norm": 0.18350888788700104,
      "learning_rate": 3.18804e-05,
      "loss": 0.2281,
      "step": 420300
    },
    {
      "epoch": 0.1408,
      "grad_norm": 0.17086632549762726,
      "learning_rate": 3.18404e-05,
      "loss": 0.2256,
      "step": 420400
    },
    {
      "epoch": 0.141,
      "grad_norm": 0.20264512300491333,
      "learning_rate": 3.18004e-05,
      "loss": 0.226,
      "step": 420500
    },
    {
      "epoch": 0.1412,
      "grad_norm": 0.1515614241361618,
      "learning_rate": 3.17604e-05,
      "loss": 0.2271,
      "step": 420600
    },
    {
      "epoch": 0.1414,
      "grad_norm": 0.19251267611980438,
      "learning_rate": 3.17204e-05,
      "loss": 0.2291,
      "step": 420700
    },
    {
      "epoch": 0.1416,
      "grad_norm": 0.1888919323682785,
      "learning_rate": 3.1680399999999996e-05,
      "loss": 0.2233,
      "step": 420800
    },
    {
      "epoch": 0.1418,
      "grad_norm": 0.18011757731437683,
      "learning_rate": 3.16404e-05,
      "loss": 0.2205,
      "step": 420900
    },
    {
      "epoch": 0.142,
      "grad_norm": 0.1989680528640747,
      "learning_rate": 3.16004e-05,
      "loss": 0.2256,
      "step": 421000
    },
    {
      "epoch": 0.1422,
      "grad_norm": 0.19918207824230194,
      "learning_rate": 3.1560400000000004e-05,
      "loss": 0.2251,
      "step": 421100
    },
    {
      "epoch": 0.1424,
      "grad_norm": 0.1635255366563797,
      "learning_rate": 3.15204e-05,
      "loss": 0.227,
      "step": 421200
    },
    {
      "epoch": 0.1426,
      "grad_norm": 0.1558133214712143,
      "learning_rate": 3.14804e-05,
      "loss": 0.2225,
      "step": 421300
    },
    {
      "epoch": 0.1428,
      "grad_norm": 0.16875283420085907,
      "learning_rate": 3.1440400000000005e-05,
      "loss": 0.2254,
      "step": 421400
    },
    {
      "epoch": 0.143,
      "grad_norm": 0.16321547329425812,
      "learning_rate": 3.1400400000000005e-05,
      "loss": 0.2272,
      "step": 421500
    },
    {
      "epoch": 0.1432,
      "grad_norm": 0.1466277837753296,
      "learning_rate": 3.13604e-05,
      "loss": 0.2234,
      "step": 421600
    },
    {
      "epoch": 0.1434,
      "grad_norm": 0.1759670078754425,
      "learning_rate": 3.13204e-05,
      "loss": 0.2229,
      "step": 421700
    },
    {
      "epoch": 0.1436,
      "grad_norm": 0.190858855843544,
      "learning_rate": 3.1280400000000006e-05,
      "loss": 0.2303,
      "step": 421800
    },
    {
      "epoch": 0.1438,
      "grad_norm": 0.2101619392633438,
      "learning_rate": 3.12404e-05,
      "loss": 0.223,
      "step": 421900
    },
    {
      "epoch": 0.144,
      "grad_norm": 0.19018958508968353,
      "learning_rate": 3.12004e-05,
      "loss": 0.2224,
      "step": 422000
    },
    {
      "epoch": 0.1442,
      "grad_norm": 0.14888699352741241,
      "learning_rate": 3.11604e-05,
      "loss": 0.2252,
      "step": 422100
    },
    {
      "epoch": 0.1444,
      "grad_norm": 0.20979218184947968,
      "learning_rate": 3.11204e-05,
      "loss": 0.2288,
      "step": 422200
    },
    {
      "epoch": 0.1446,
      "grad_norm": 0.1412057876586914,
      "learning_rate": 3.10804e-05,
      "loss": 0.2256,
      "step": 422300
    },
    {
      "epoch": 0.1448,
      "grad_norm": 0.19975054264068604,
      "learning_rate": 3.10404e-05,
      "loss": 0.2279,
      "step": 422400
    },
    {
      "epoch": 0.145,
      "grad_norm": 0.15945744514465332,
      "learning_rate": 3.10004e-05,
      "loss": 0.2254,
      "step": 422500
    },
    {
      "epoch": 0.1452,
      "grad_norm": 0.19987906515598297,
      "learning_rate": 3.09604e-05,
      "loss": 0.2219,
      "step": 422600
    },
    {
      "epoch": 0.1454,
      "grad_norm": 0.19047895073890686,
      "learning_rate": 3.0920399999999996e-05,
      "loss": 0.2206,
      "step": 422700
    },
    {
      "epoch": 0.1456,
      "grad_norm": 0.17094969749450684,
      "learning_rate": 3.08804e-05,
      "loss": 0.2283,
      "step": 422800
    },
    {
      "epoch": 0.1458,
      "grad_norm": 0.18042507767677307,
      "learning_rate": 3.0840400000000004e-05,
      "loss": 0.2217,
      "step": 422900
    },
    {
      "epoch": 0.146,
      "grad_norm": 0.17472079396247864,
      "learning_rate": 3.0800400000000004e-05,
      "loss": 0.2275,
      "step": 423000
    },
    {
      "epoch": 0.1462,
      "grad_norm": 0.18705900013446808,
      "learning_rate": 3.07604e-05,
      "loss": 0.2217,
      "step": 423100
    },
    {
      "epoch": 0.1464,
      "grad_norm": 0.19462667405605316,
      "learning_rate": 3.07204e-05,
      "loss": 0.2257,
      "step": 423200
    },
    {
      "epoch": 0.1466,
      "grad_norm": 0.18578322231769562,
      "learning_rate": 3.0680400000000005e-05,
      "loss": 0.2241,
      "step": 423300
    },
    {
      "epoch": 0.1468,
      "grad_norm": 0.2105281949043274,
      "learning_rate": 3.0640400000000006e-05,
      "loss": 0.234,
      "step": 423400
    },
    {
      "epoch": 0.147,
      "grad_norm": 0.1453637033700943,
      "learning_rate": 3.06004e-05,
      "loss": 0.2235,
      "step": 423500
    },
    {
      "epoch": 0.1472,
      "grad_norm": 0.17777366936206818,
      "learning_rate": 3.05604e-05,
      "loss": 0.2241,
      "step": 423600
    },
    {
      "epoch": 0.1474,
      "grad_norm": 0.16608352959156036,
      "learning_rate": 3.05204e-05,
      "loss": 0.2234,
      "step": 423700
    },
    {
      "epoch": 0.1476,
      "grad_norm": 0.16040940582752228,
      "learning_rate": 3.0480400000000004e-05,
      "loss": 0.2228,
      "step": 423800
    },
    {
      "epoch": 0.1478,
      "grad_norm": 0.15881648659706116,
      "learning_rate": 3.04404e-05,
      "loss": 0.2241,
      "step": 423900
    },
    {
      "epoch": 0.148,
      "grad_norm": 0.18990524113178253,
      "learning_rate": 3.04004e-05,
      "loss": 0.2263,
      "step": 424000
    },
    {
      "epoch": 0.1482,
      "grad_norm": 0.16839918494224548,
      "learning_rate": 3.03604e-05,
      "loss": 0.2238,
      "step": 424100
    },
    {
      "epoch": 0.1484,
      "grad_norm": 0.14922964572906494,
      "learning_rate": 3.03204e-05,
      "loss": 0.2238,
      "step": 424200
    },
    {
      "epoch": 0.1486,
      "grad_norm": 0.14740827679634094,
      "learning_rate": 3.0280400000000002e-05,
      "loss": 0.2266,
      "step": 424300
    },
    {
      "epoch": 0.1488,
      "grad_norm": 0.18283480405807495,
      "learning_rate": 3.0240400000000003e-05,
      "loss": 0.2267,
      "step": 424400
    },
    {
      "epoch": 0.149,
      "grad_norm": 0.22374579310417175,
      "learning_rate": 3.02004e-05,
      "loss": 0.2242,
      "step": 424500
    },
    {
      "epoch": 0.1492,
      "grad_norm": 0.21242839097976685,
      "learning_rate": 3.01604e-05,
      "loss": 0.2261,
      "step": 424600
    },
    {
      "epoch": 0.1494,
      "grad_norm": 0.17830756306648254,
      "learning_rate": 3.0120400000000004e-05,
      "loss": 0.227,
      "step": 424700
    },
    {
      "epoch": 0.1496,
      "grad_norm": 0.2056087851524353,
      "learning_rate": 3.0080400000000004e-05,
      "loss": 0.2271,
      "step": 424800
    },
    {
      "epoch": 0.1498,
      "grad_norm": 0.15750348567962646,
      "learning_rate": 3.00404e-05,
      "loss": 0.2258,
      "step": 424900
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.1605195552110672,
      "learning_rate": 3.00004e-05,
      "loss": 0.2269,
      "step": 425000
    },
    {
      "epoch": 0.1502,
      "grad_norm": 0.17202390730381012,
      "learning_rate": 2.99604e-05,
      "loss": 0.2278,
      "step": 425100
    },
    {
      "epoch": 0.1504,
      "grad_norm": 0.18993693590164185,
      "learning_rate": 2.9920400000000002e-05,
      "loss": 0.2296,
      "step": 425200
    },
    {
      "epoch": 0.1506,
      "grad_norm": 0.16901157796382904,
      "learning_rate": 2.9880400000000003e-05,
      "loss": 0.2262,
      "step": 425300
    },
    {
      "epoch": 0.1508,
      "grad_norm": 0.18246543407440186,
      "learning_rate": 2.9840400000000003e-05,
      "loss": 0.2246,
      "step": 425400
    },
    {
      "epoch": 0.151,
      "grad_norm": 0.17869111895561218,
      "learning_rate": 2.98004e-05,
      "loss": 0.2241,
      "step": 425500
    },
    {
      "epoch": 0.1512,
      "grad_norm": 0.1899707317352295,
      "learning_rate": 2.97604e-05,
      "loss": 0.2259,
      "step": 425600
    },
    {
      "epoch": 0.1514,
      "grad_norm": 0.18976940214633942,
      "learning_rate": 2.9720400000000004e-05,
      "loss": 0.2248,
      "step": 425700
    },
    {
      "epoch": 0.1516,
      "grad_norm": 0.19041858613491058,
      "learning_rate": 2.96804e-05,
      "loss": 0.2265,
      "step": 425800
    },
    {
      "epoch": 0.1518,
      "grad_norm": 0.19394929707050323,
      "learning_rate": 2.9640400000000002e-05,
      "loss": 0.2302,
      "step": 425900
    },
    {
      "epoch": 0.152,
      "grad_norm": 0.16645781695842743,
      "learning_rate": 2.96004e-05,
      "loss": 0.227,
      "step": 426000
    },
    {
      "epoch": 0.1522,
      "grad_norm": 0.15843971073627472,
      "learning_rate": 2.95604e-05,
      "loss": 0.2269,
      "step": 426100
    },
    {
      "epoch": 0.1524,
      "grad_norm": 0.18866930902004242,
      "learning_rate": 2.9520400000000003e-05,
      "loss": 0.2269,
      "step": 426200
    },
    {
      "epoch": 0.1526,
      "grad_norm": 0.18919654190540314,
      "learning_rate": 2.9480400000000003e-05,
      "loss": 0.2248,
      "step": 426300
    },
    {
      "epoch": 0.1528,
      "grad_norm": 0.22273516654968262,
      "learning_rate": 2.94404e-05,
      "loss": 0.2214,
      "step": 426400
    },
    {
      "epoch": 0.153,
      "grad_norm": 0.20623090863227844,
      "learning_rate": 2.94004e-05,
      "loss": 0.2245,
      "step": 426500
    },
    {
      "epoch": 0.1532,
      "grad_norm": 0.16490277647972107,
      "learning_rate": 2.9360399999999998e-05,
      "loss": 0.2234,
      "step": 426600
    },
    {
      "epoch": 0.1534,
      "grad_norm": 0.30756646394729614,
      "learning_rate": 2.9320400000000005e-05,
      "loss": 0.2289,
      "step": 426700
    },
    {
      "epoch": 0.1536,
      "grad_norm": 0.18680694699287415,
      "learning_rate": 2.9280400000000002e-05,
      "loss": 0.2269,
      "step": 426800
    },
    {
      "epoch": 0.1538,
      "grad_norm": 0.16538353264331818,
      "learning_rate": 2.9240400000000002e-05,
      "loss": 0.2249,
      "step": 426900
    },
    {
      "epoch": 0.154,
      "grad_norm": 0.21780794858932495,
      "learning_rate": 2.92004e-05,
      "loss": 0.2263,
      "step": 427000
    },
    {
      "epoch": 0.1542,
      "grad_norm": 0.1775357723236084,
      "learning_rate": 2.91604e-05,
      "loss": 0.2231,
      "step": 427100
    },
    {
      "epoch": 0.1544,
      "grad_norm": 0.17265263199806213,
      "learning_rate": 2.9120400000000003e-05,
      "loss": 0.2299,
      "step": 427200
    },
    {
      "epoch": 0.1546,
      "grad_norm": 0.19637222588062286,
      "learning_rate": 2.9080400000000004e-05,
      "loss": 0.223,
      "step": 427300
    },
    {
      "epoch": 0.1548,
      "grad_norm": 0.2579210102558136,
      "learning_rate": 2.90404e-05,
      "loss": 0.2263,
      "step": 427400
    },
    {
      "epoch": 0.155,
      "grad_norm": 0.17191685736179352,
      "learning_rate": 2.90004e-05,
      "loss": 0.2241,
      "step": 427500
    },
    {
      "epoch": 0.1552,
      "grad_norm": 0.19325803220272064,
      "learning_rate": 2.8960399999999998e-05,
      "loss": 0.2216,
      "step": 427600
    },
    {
      "epoch": 0.1554,
      "grad_norm": 0.16005942225456238,
      "learning_rate": 2.8920400000000002e-05,
      "loss": 0.2276,
      "step": 427700
    },
    {
      "epoch": 0.1556,
      "grad_norm": 0.17553672194480896,
      "learning_rate": 2.8880400000000002e-05,
      "loss": 0.227,
      "step": 427800
    },
    {
      "epoch": 0.1558,
      "grad_norm": 0.24208134412765503,
      "learning_rate": 2.88404e-05,
      "loss": 0.2234,
      "step": 427900
    },
    {
      "epoch": 0.156,
      "grad_norm": 0.25327280163764954,
      "learning_rate": 2.88004e-05,
      "loss": 0.2258,
      "step": 428000
    },
    {
      "epoch": 0.1562,
      "grad_norm": 0.19016684591770172,
      "learning_rate": 2.8760400000000003e-05,
      "loss": 0.2251,
      "step": 428100
    },
    {
      "epoch": 0.1564,
      "grad_norm": 0.16044831275939941,
      "learning_rate": 2.8720400000000004e-05,
      "loss": 0.2227,
      "step": 428200
    },
    {
      "epoch": 0.1566,
      "grad_norm": 0.2034226357936859,
      "learning_rate": 2.86804e-05,
      "loss": 0.2269,
      "step": 428300
    },
    {
      "epoch": 0.1568,
      "grad_norm": 0.16030117869377136,
      "learning_rate": 2.86404e-05,
      "loss": 0.2299,
      "step": 428400
    },
    {
      "epoch": 0.157,
      "grad_norm": 0.1629854291677475,
      "learning_rate": 2.8600399999999998e-05,
      "loss": 0.2237,
      "step": 428500
    },
    {
      "epoch": 0.1572,
      "grad_norm": 0.1795048862695694,
      "learning_rate": 2.8560400000000005e-05,
      "loss": 0.2246,
      "step": 428600
    },
    {
      "epoch": 0.1574,
      "grad_norm": 0.19448979198932648,
      "learning_rate": 2.8520400000000002e-05,
      "loss": 0.2243,
      "step": 428700
    },
    {
      "epoch": 0.1576,
      "grad_norm": 0.16780062019824982,
      "learning_rate": 2.8480400000000003e-05,
      "loss": 0.2251,
      "step": 428800
    },
    {
      "epoch": 0.1578,
      "grad_norm": 0.18648077547550201,
      "learning_rate": 2.84404e-05,
      "loss": 0.2269,
      "step": 428900
    },
    {
      "epoch": 0.158,
      "grad_norm": 0.15443378686904907,
      "learning_rate": 2.84004e-05,
      "loss": 0.2243,
      "step": 429000
    },
    {
      "epoch": 0.1582,
      "grad_norm": 0.18094438314437866,
      "learning_rate": 2.8360400000000004e-05,
      "loss": 0.2293,
      "step": 429100
    },
    {
      "epoch": 0.1584,
      "grad_norm": 0.15523366630077362,
      "learning_rate": 2.8320400000000004e-05,
      "loss": 0.2278,
      "step": 429200
    },
    {
      "epoch": 0.1586,
      "grad_norm": 0.18622247874736786,
      "learning_rate": 2.82804e-05,
      "loss": 0.2242,
      "step": 429300
    },
    {
      "epoch": 0.1588,
      "grad_norm": 0.20111188292503357,
      "learning_rate": 2.82404e-05,
      "loss": 0.2248,
      "step": 429400
    },
    {
      "epoch": 0.159,
      "grad_norm": 0.23439711332321167,
      "learning_rate": 2.82004e-05,
      "loss": 0.2294,
      "step": 429500
    },
    {
      "epoch": 0.1592,
      "grad_norm": 0.20886704325675964,
      "learning_rate": 2.8160400000000002e-05,
      "loss": 0.2288,
      "step": 429600
    },
    {
      "epoch": 0.1594,
      "grad_norm": 0.1680990606546402,
      "learning_rate": 2.8120400000000003e-05,
      "loss": 0.2275,
      "step": 429700
    },
    {
      "epoch": 0.1596,
      "grad_norm": 0.1492314487695694,
      "learning_rate": 2.80804e-05,
      "loss": 0.2248,
      "step": 429800
    },
    {
      "epoch": 0.1598,
      "grad_norm": 0.2123488038778305,
      "learning_rate": 2.80404e-05,
      "loss": 0.2235,
      "step": 429900
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.17558607459068298,
      "learning_rate": 2.80004e-05,
      "loss": 0.2247,
      "step": 430000
    },
    {
      "epoch": 0.1602,
      "grad_norm": 0.26399195194244385,
      "learning_rate": 2.7960400000000004e-05,
      "loss": 0.2248,
      "step": 430100
    },
    {
      "epoch": 0.1604,
      "grad_norm": 0.16792334616184235,
      "learning_rate": 2.79204e-05,
      "loss": 0.2251,
      "step": 430200
    },
    {
      "epoch": 0.1606,
      "grad_norm": 0.16660743951797485,
      "learning_rate": 2.78804e-05,
      "loss": 0.2265,
      "step": 430300
    },
    {
      "epoch": 0.1608,
      "grad_norm": 0.2021878957748413,
      "learning_rate": 2.78404e-05,
      "loss": 0.2231,
      "step": 430400
    },
    {
      "epoch": 0.161,
      "grad_norm": 0.16516359150409698,
      "learning_rate": 2.78004e-05,
      "loss": 0.2317,
      "step": 430500
    },
    {
      "epoch": 0.1612,
      "grad_norm": 0.16823023557662964,
      "learning_rate": 2.7760400000000003e-05,
      "loss": 0.2274,
      "step": 430600
    },
    {
      "epoch": 0.1614,
      "grad_norm": 0.15308980643749237,
      "learning_rate": 2.7720400000000003e-05,
      "loss": 0.2278,
      "step": 430700
    },
    {
      "epoch": 0.1616,
      "grad_norm": 0.17121495306491852,
      "learning_rate": 2.76804e-05,
      "loss": 0.2291,
      "step": 430800
    },
    {
      "epoch": 0.1618,
      "grad_norm": 0.17018477618694305,
      "learning_rate": 2.76404e-05,
      "loss": 0.2214,
      "step": 430900
    },
    {
      "epoch": 0.162,
      "grad_norm": 0.18164673447608948,
      "learning_rate": 2.7600400000000004e-05,
      "loss": 0.2264,
      "step": 431000
    },
    {
      "epoch": 0.1622,
      "grad_norm": 0.2391088306903839,
      "learning_rate": 2.75604e-05,
      "loss": 0.2256,
      "step": 431100
    },
    {
      "epoch": 0.1624,
      "grad_norm": 0.1486031413078308,
      "learning_rate": 2.75204e-05,
      "loss": 0.2305,
      "step": 431200
    },
    {
      "epoch": 0.1626,
      "grad_norm": 0.3389119803905487,
      "learning_rate": 2.7480400000000002e-05,
      "loss": 0.2268,
      "step": 431300
    },
    {
      "epoch": 0.1628,
      "grad_norm": 0.203375905752182,
      "learning_rate": 2.74404e-05,
      "loss": 0.2235,
      "step": 431400
    },
    {
      "epoch": 0.163,
      "grad_norm": 0.16444042325019836,
      "learning_rate": 2.7400400000000003e-05,
      "loss": 0.2258,
      "step": 431500
    },
    {
      "epoch": 0.1632,
      "grad_norm": 0.26179274916648865,
      "learning_rate": 2.7360400000000003e-05,
      "loss": 0.2255,
      "step": 431600
    },
    {
      "epoch": 0.1634,
      "grad_norm": 0.20581679046154022,
      "learning_rate": 2.73204e-05,
      "loss": 0.2228,
      "step": 431700
    },
    {
      "epoch": 0.1636,
      "grad_norm": 0.19416281580924988,
      "learning_rate": 2.72804e-05,
      "loss": 0.2289,
      "step": 431800
    },
    {
      "epoch": 0.1638,
      "grad_norm": 0.17209891974925995,
      "learning_rate": 2.7240399999999998e-05,
      "loss": 0.2248,
      "step": 431900
    },
    {
      "epoch": 0.164,
      "grad_norm": 0.16744202375411987,
      "learning_rate": 2.7200400000000005e-05,
      "loss": 0.2291,
      "step": 432000
    },
    {
      "epoch": 0.1642,
      "grad_norm": 0.18142908811569214,
      "learning_rate": 2.7160400000000002e-05,
      "loss": 0.2228,
      "step": 432100
    },
    {
      "epoch": 0.1644,
      "grad_norm": 0.2096833437681198,
      "learning_rate": 2.7120400000000002e-05,
      "loss": 0.2295,
      "step": 432200
    },
    {
      "epoch": 0.1646,
      "grad_norm": 0.15951569378376007,
      "learning_rate": 2.70804e-05,
      "loss": 0.225,
      "step": 432300
    },
    {
      "epoch": 0.1648,
      "grad_norm": 0.16083496809005737,
      "learning_rate": 2.70404e-05,
      "loss": 0.226,
      "step": 432400
    },
    {
      "epoch": 0.165,
      "grad_norm": 0.1795596182346344,
      "learning_rate": 2.7000400000000003e-05,
      "loss": 0.2243,
      "step": 432500
    },
    {
      "epoch": 0.1652,
      "grad_norm": 0.18583637475967407,
      "learning_rate": 2.6960400000000004e-05,
      "loss": 0.2226,
      "step": 432600
    },
    {
      "epoch": 0.1654,
      "grad_norm": 0.146294966340065,
      "learning_rate": 2.69204e-05,
      "loss": 0.224,
      "step": 432700
    },
    {
      "epoch": 0.1656,
      "grad_norm": 0.17165237665176392,
      "learning_rate": 2.68804e-05,
      "loss": 0.2225,
      "step": 432800
    },
    {
      "epoch": 0.1658,
      "grad_norm": 0.17422151565551758,
      "learning_rate": 2.6840399999999998e-05,
      "loss": 0.2269,
      "step": 432900
    },
    {
      "epoch": 0.166,
      "grad_norm": 0.17443683743476868,
      "learning_rate": 2.6800400000000002e-05,
      "loss": 0.2253,
      "step": 433000
    },
    {
      "epoch": 0.1662,
      "grad_norm": 0.18008099496364594,
      "learning_rate": 2.6760400000000002e-05,
      "loss": 0.2232,
      "step": 433100
    },
    {
      "epoch": 0.1664,
      "grad_norm": 0.17084477841854095,
      "learning_rate": 2.6720400000000003e-05,
      "loss": 0.2333,
      "step": 433200
    },
    {
      "epoch": 0.1666,
      "grad_norm": 0.17184089124202728,
      "learning_rate": 2.66804e-05,
      "loss": 0.2236,
      "step": 433300
    },
    {
      "epoch": 0.1668,
      "grad_norm": 0.17407119274139404,
      "learning_rate": 2.66404e-05,
      "loss": 0.2247,
      "step": 433400
    },
    {
      "epoch": 0.167,
      "grad_norm": 0.155440554022789,
      "learning_rate": 2.6600400000000004e-05,
      "loss": 0.2283,
      "step": 433500
    },
    {
      "epoch": 0.1672,
      "grad_norm": 0.15436717867851257,
      "learning_rate": 2.65604e-05,
      "loss": 0.2259,
      "step": 433600
    },
    {
      "epoch": 0.1674,
      "grad_norm": 0.20875059068202972,
      "learning_rate": 2.65204e-05,
      "loss": 0.2306,
      "step": 433700
    },
    {
      "epoch": 0.1676,
      "grad_norm": 0.38795992732048035,
      "learning_rate": 2.6480399999999998e-05,
      "loss": 0.23,
      "step": 433800
    },
    {
      "epoch": 0.1678,
      "grad_norm": 0.29372575879096985,
      "learning_rate": 2.64404e-05,
      "loss": 0.2278,
      "step": 433900
    },
    {
      "epoch": 0.168,
      "grad_norm": 0.1667247861623764,
      "learning_rate": 2.6400400000000002e-05,
      "loss": 0.2275,
      "step": 434000
    },
    {
      "epoch": 0.1682,
      "grad_norm": 0.1645084172487259,
      "learning_rate": 2.6360400000000003e-05,
      "loss": 0.2242,
      "step": 434100
    },
    {
      "epoch": 0.1684,
      "grad_norm": 0.20222832262516022,
      "learning_rate": 2.63204e-05,
      "loss": 0.2271,
      "step": 434200
    },
    {
      "epoch": 0.1686,
      "grad_norm": 0.3428006172180176,
      "learning_rate": 2.62804e-05,
      "loss": 0.2247,
      "step": 434300
    },
    {
      "epoch": 0.1688,
      "grad_norm": 0.1995532214641571,
      "learning_rate": 2.6240400000000004e-05,
      "loss": 0.2257,
      "step": 434400
    },
    {
      "epoch": 0.169,
      "grad_norm": 0.18206648528575897,
      "learning_rate": 2.6200400000000004e-05,
      "loss": 0.2242,
      "step": 434500
    },
    {
      "epoch": 0.1692,
      "grad_norm": 0.15445922315120697,
      "learning_rate": 2.61604e-05,
      "loss": 0.2251,
      "step": 434600
    },
    {
      "epoch": 0.1694,
      "grad_norm": 0.1566971242427826,
      "learning_rate": 2.61204e-05,
      "loss": 0.2229,
      "step": 434700
    },
    {
      "epoch": 0.1696,
      "grad_norm": 0.16852755844593048,
      "learning_rate": 2.60804e-05,
      "loss": 0.2287,
      "step": 434800
    },
    {
      "epoch": 0.1698,
      "grad_norm": 0.17215055227279663,
      "learning_rate": 2.6040400000000002e-05,
      "loss": 0.2224,
      "step": 434900
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.22290776669979095,
      "learning_rate": 2.6000400000000003e-05,
      "loss": 0.2338,
      "step": 435000
    },
    {
      "epoch": 0.1702,
      "grad_norm": 0.16391290724277496,
      "learning_rate": 2.5960400000000003e-05,
      "loss": 0.2263,
      "step": 435100
    },
    {
      "epoch": 0.1704,
      "grad_norm": 0.17789016664028168,
      "learning_rate": 2.59204e-05,
      "loss": 0.2276,
      "step": 435200
    },
    {
      "epoch": 0.1706,
      "grad_norm": 0.17922519147396088,
      "learning_rate": 2.58804e-05,
      "loss": 0.2208,
      "step": 435300
    },
    {
      "epoch": 0.1708,
      "grad_norm": 0.16985663771629333,
      "learning_rate": 2.5840400000000004e-05,
      "loss": 0.2267,
      "step": 435400
    },
    {
      "epoch": 0.171,
      "grad_norm": 0.19357039034366608,
      "learning_rate": 2.58004e-05,
      "loss": 0.2289,
      "step": 435500
    },
    {
      "epoch": 0.1712,
      "grad_norm": 0.1737697869539261,
      "learning_rate": 2.57604e-05,
      "loss": 0.2238,
      "step": 435600
    },
    {
      "epoch": 0.1714,
      "grad_norm": 0.17874495685100555,
      "learning_rate": 2.57204e-05,
      "loss": 0.2233,
      "step": 435700
    },
    {
      "epoch": 0.1716,
      "grad_norm": 0.17967748641967773,
      "learning_rate": 2.56804e-05,
      "loss": 0.223,
      "step": 435800
    },
    {
      "epoch": 0.1718,
      "grad_norm": 0.1678665578365326,
      "learning_rate": 2.5640400000000003e-05,
      "loss": 0.2282,
      "step": 435900
    },
    {
      "epoch": 0.172,
      "grad_norm": 0.1909477561712265,
      "learning_rate": 2.5600400000000003e-05,
      "loss": 0.2272,
      "step": 436000
    },
    {
      "epoch": 0.1722,
      "grad_norm": 0.22589877247810364,
      "learning_rate": 2.55604e-05,
      "loss": 0.231,
      "step": 436100
    },
    {
      "epoch": 0.1724,
      "grad_norm": 0.16562041640281677,
      "learning_rate": 2.55204e-05,
      "loss": 0.2381,
      "step": 436200
    },
    {
      "epoch": 0.1726,
      "grad_norm": 0.19467677175998688,
      "learning_rate": 2.5480399999999997e-05,
      "loss": 0.2286,
      "step": 436300
    },
    {
      "epoch": 0.1728,
      "grad_norm": 0.7384561896324158,
      "learning_rate": 2.5440400000000005e-05,
      "loss": 0.2238,
      "step": 436400
    },
    {
      "epoch": 0.173,
      "grad_norm": 0.1401069164276123,
      "learning_rate": 2.54004e-05,
      "loss": 0.2256,
      "step": 436500
    },
    {
      "epoch": 0.1732,
      "grad_norm": 0.20685148239135742,
      "learning_rate": 2.5360400000000002e-05,
      "loss": 0.2265,
      "step": 436600
    },
    {
      "epoch": 0.1734,
      "grad_norm": 0.19295437633991241,
      "learning_rate": 2.53204e-05,
      "loss": 0.2262,
      "step": 436700
    },
    {
      "epoch": 0.1736,
      "grad_norm": 0.23843425512313843,
      "learning_rate": 2.52804e-05,
      "loss": 0.2305,
      "step": 436800
    },
    {
      "epoch": 0.1738,
      "grad_norm": 0.16619522869586945,
      "learning_rate": 2.5240400000000003e-05,
      "loss": 0.2272,
      "step": 436900
    },
    {
      "epoch": 0.174,
      "grad_norm": 0.17892450094223022,
      "learning_rate": 2.52004e-05,
      "loss": 0.2283,
      "step": 437000
    },
    {
      "epoch": 0.1742,
      "grad_norm": 0.15790720283985138,
      "learning_rate": 2.51604e-05,
      "loss": 0.2273,
      "step": 437100
    },
    {
      "epoch": 0.1744,
      "grad_norm": 0.1899292767047882,
      "learning_rate": 2.51204e-05,
      "loss": 0.2229,
      "step": 437200
    },
    {
      "epoch": 0.1746,
      "grad_norm": 0.14178262650966644,
      "learning_rate": 2.5080400000000005e-05,
      "loss": 0.2284,
      "step": 437300
    },
    {
      "epoch": 0.1748,
      "grad_norm": 0.16377106308937073,
      "learning_rate": 2.50404e-05,
      "loss": 0.2259,
      "step": 437400
    },
    {
      "epoch": 0.175,
      "grad_norm": 0.21757832169532776,
      "learning_rate": 2.5000400000000002e-05,
      "loss": 0.2281,
      "step": 437500
    },
    {
      "epoch": 0.1752,
      "grad_norm": 0.16363129019737244,
      "learning_rate": 2.49604e-05,
      "loss": 0.2266,
      "step": 437600
    },
    {
      "epoch": 0.1754,
      "grad_norm": 0.1644134521484375,
      "learning_rate": 2.4920400000000003e-05,
      "loss": 0.2226,
      "step": 437700
    },
    {
      "epoch": 0.1756,
      "grad_norm": 0.14680400490760803,
      "learning_rate": 2.48804e-05,
      "loss": 0.2245,
      "step": 437800
    },
    {
      "epoch": 0.1758,
      "grad_norm": 0.14819519221782684,
      "learning_rate": 2.4840400000000004e-05,
      "loss": 0.2251,
      "step": 437900
    },
    {
      "epoch": 0.176,
      "grad_norm": 0.18625690042972565,
      "learning_rate": 2.48004e-05,
      "loss": 0.2286,
      "step": 438000
    },
    {
      "epoch": 0.1762,
      "grad_norm": 0.15373022854328156,
      "learning_rate": 2.47604e-05,
      "loss": 0.2275,
      "step": 438100
    },
    {
      "epoch": 0.1764,
      "grad_norm": 0.1993611454963684,
      "learning_rate": 2.47204e-05,
      "loss": 0.2278,
      "step": 438200
    },
    {
      "epoch": 0.1766,
      "grad_norm": 0.24995622038841248,
      "learning_rate": 2.4680400000000002e-05,
      "loss": 0.2302,
      "step": 438300
    },
    {
      "epoch": 0.1768,
      "grad_norm": 0.17456863820552826,
      "learning_rate": 2.4640400000000002e-05,
      "loss": 0.2309,
      "step": 438400
    },
    {
      "epoch": 0.177,
      "grad_norm": 0.16433963179588318,
      "learning_rate": 2.4600400000000003e-05,
      "loss": 0.2263,
      "step": 438500
    },
    {
      "epoch": 0.1772,
      "grad_norm": 0.16630324721336365,
      "learning_rate": 2.45604e-05,
      "loss": 0.2269,
      "step": 438600
    },
    {
      "epoch": 0.1774,
      "grad_norm": 0.19276133179664612,
      "learning_rate": 2.4520400000000003e-05,
      "loss": 0.2263,
      "step": 438700
    },
    {
      "epoch": 0.1776,
      "grad_norm": 0.19641803205013275,
      "learning_rate": 2.44804e-05,
      "loss": 0.2256,
      "step": 438800
    },
    {
      "epoch": 0.1778,
      "grad_norm": 0.164507657289505,
      "learning_rate": 2.44404e-05,
      "loss": 0.224,
      "step": 438900
    },
    {
      "epoch": 0.178,
      "grad_norm": 0.15548419952392578,
      "learning_rate": 2.44004e-05,
      "loss": 0.2268,
      "step": 439000
    },
    {
      "epoch": 0.1782,
      "grad_norm": 0.15844307839870453,
      "learning_rate": 2.43604e-05,
      "loss": 0.2254,
      "step": 439100
    },
    {
      "epoch": 0.1784,
      "grad_norm": 0.18595917522907257,
      "learning_rate": 2.4320400000000002e-05,
      "loss": 0.2252,
      "step": 439200
    },
    {
      "epoch": 0.1786,
      "grad_norm": 0.27682188153266907,
      "learning_rate": 2.42804e-05,
      "loss": 0.2253,
      "step": 439300
    },
    {
      "epoch": 0.1788,
      "grad_norm": 0.1652635633945465,
      "learning_rate": 2.4240400000000003e-05,
      "loss": 0.2291,
      "step": 439400
    },
    {
      "epoch": 0.179,
      "grad_norm": 0.3612424433231354,
      "learning_rate": 2.42004e-05,
      "loss": 0.2258,
      "step": 439500
    },
    {
      "epoch": 0.1792,
      "grad_norm": 0.17087934911251068,
      "learning_rate": 2.4160400000000003e-05,
      "loss": 0.2233,
      "step": 439600
    },
    {
      "epoch": 0.1794,
      "grad_norm": 0.15505070984363556,
      "learning_rate": 2.41204e-05,
      "loss": 0.2257,
      "step": 439700
    },
    {
      "epoch": 0.1796,
      "grad_norm": 0.1622672826051712,
      "learning_rate": 2.40804e-05,
      "loss": 0.2252,
      "step": 439800
    },
    {
      "epoch": 0.1798,
      "grad_norm": 0.19575311243534088,
      "learning_rate": 2.40404e-05,
      "loss": 0.2238,
      "step": 439900
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.1737244427204132,
      "learning_rate": 2.40004e-05,
      "loss": 0.2249,
      "step": 440000
    },
    {
      "epoch": 0.1802,
      "grad_norm": 0.24295039474964142,
      "learning_rate": 2.3960400000000002e-05,
      "loss": 0.2219,
      "step": 440100
    },
    {
      "epoch": 0.1804,
      "grad_norm": 0.17787252366542816,
      "learning_rate": 2.3920400000000002e-05,
      "loss": 0.2262,
      "step": 440200
    },
    {
      "epoch": 0.1806,
      "grad_norm": 0.15430960059165955,
      "learning_rate": 2.38804e-05,
      "loss": 0.2237,
      "step": 440300
    },
    {
      "epoch": 0.1808,
      "grad_norm": 0.17729860544204712,
      "learning_rate": 2.3840400000000003e-05,
      "loss": 0.2292,
      "step": 440400
    },
    {
      "epoch": 0.181,
      "grad_norm": 0.19913239777088165,
      "learning_rate": 2.38004e-05,
      "loss": 0.2273,
      "step": 440500
    },
    {
      "epoch": 0.1812,
      "grad_norm": 0.15231093764305115,
      "learning_rate": 2.3760400000000004e-05,
      "loss": 0.2206,
      "step": 440600
    },
    {
      "epoch": 0.1814,
      "grad_norm": 0.18984483182430267,
      "learning_rate": 2.37204e-05,
      "loss": 0.2152,
      "step": 440700
    },
    {
      "epoch": 0.1816,
      "grad_norm": 0.16935545206069946,
      "learning_rate": 2.36804e-05,
      "loss": 0.2228,
      "step": 440800
    },
    {
      "epoch": 0.1818,
      "grad_norm": 0.20050957798957825,
      "learning_rate": 2.36404e-05,
      "loss": 0.2235,
      "step": 440900
    },
    {
      "epoch": 0.182,
      "grad_norm": 0.19681406021118164,
      "learning_rate": 2.3600400000000002e-05,
      "loss": 0.2278,
      "step": 441000
    },
    {
      "epoch": 0.1822,
      "grad_norm": 0.14469000697135925,
      "learning_rate": 2.3560400000000002e-05,
      "loss": 0.2282,
      "step": 441100
    },
    {
      "epoch": 0.1824,
      "grad_norm": 0.22131530940532684,
      "learning_rate": 2.35204e-05,
      "loss": 0.2255,
      "step": 441200
    },
    {
      "epoch": 0.1826,
      "grad_norm": 0.23927749693393707,
      "learning_rate": 2.3480400000000003e-05,
      "loss": 0.2306,
      "step": 441300
    },
    {
      "epoch": 0.1828,
      "grad_norm": 0.22400574386119843,
      "learning_rate": 2.34404e-05,
      "loss": 0.2295,
      "step": 441400
    },
    {
      "epoch": 0.183,
      "grad_norm": 0.21024823188781738,
      "learning_rate": 2.34004e-05,
      "loss": 0.2315,
      "step": 441500
    },
    {
      "epoch": 0.1832,
      "grad_norm": 0.293183296918869,
      "learning_rate": 2.33604e-05,
      "loss": 0.2217,
      "step": 441600
    },
    {
      "epoch": 0.1834,
      "grad_norm": 0.17141690850257874,
      "learning_rate": 2.33204e-05,
      "loss": 0.2252,
      "step": 441700
    },
    {
      "epoch": 0.1836,
      "grad_norm": 0.19596771895885468,
      "learning_rate": 2.32804e-05,
      "loss": 0.2261,
      "step": 441800
    },
    {
      "epoch": 0.1838,
      "grad_norm": 0.1949038952589035,
      "learning_rate": 2.3240400000000002e-05,
      "loss": 0.2252,
      "step": 441900
    },
    {
      "epoch": 0.184,
      "grad_norm": 0.22144322097301483,
      "learning_rate": 2.32004e-05,
      "loss": 0.2218,
      "step": 442000
    },
    {
      "epoch": 0.1842,
      "grad_norm": 0.17387978732585907,
      "learning_rate": 2.3160400000000003e-05,
      "loss": 0.2226,
      "step": 442100
    },
    {
      "epoch": 0.1844,
      "grad_norm": 0.1720968782901764,
      "learning_rate": 2.31204e-05,
      "loss": 0.2263,
      "step": 442200
    },
    {
      "epoch": 0.1846,
      "grad_norm": 0.2119131088256836,
      "learning_rate": 2.3080400000000004e-05,
      "loss": 0.2296,
      "step": 442300
    },
    {
      "epoch": 0.1848,
      "grad_norm": 0.14969196915626526,
      "learning_rate": 2.30404e-05,
      "loss": 0.2229,
      "step": 442400
    },
    {
      "epoch": 0.185,
      "grad_norm": 0.18958836793899536,
      "learning_rate": 2.30004e-05,
      "loss": 0.222,
      "step": 442500
    },
    {
      "epoch": 0.1852,
      "grad_norm": 0.21705825626850128,
      "learning_rate": 2.29604e-05,
      "loss": 0.2251,
      "step": 442600
    },
    {
      "epoch": 0.1854,
      "grad_norm": 0.19531495869159698,
      "learning_rate": 2.29204e-05,
      "loss": 0.2252,
      "step": 442700
    },
    {
      "epoch": 0.1856,
      "grad_norm": 0.1760839819908142,
      "learning_rate": 2.2880400000000002e-05,
      "loss": 0.2248,
      "step": 442800
    },
    {
      "epoch": 0.1858,
      "grad_norm": 0.18036484718322754,
      "learning_rate": 2.28404e-05,
      "loss": 0.222,
      "step": 442900
    },
    {
      "epoch": 0.186,
      "grad_norm": 0.200086772441864,
      "learning_rate": 2.2800400000000003e-05,
      "loss": 0.2259,
      "step": 443000
    },
    {
      "epoch": 0.1862,
      "grad_norm": 0.5199419856071472,
      "learning_rate": 2.27604e-05,
      "loss": 0.2236,
      "step": 443100
    },
    {
      "epoch": 0.1864,
      "grad_norm": 0.1505269855260849,
      "learning_rate": 2.27204e-05,
      "loss": 0.2243,
      "step": 443200
    },
    {
      "epoch": 0.1866,
      "grad_norm": 0.21658334136009216,
      "learning_rate": 2.26804e-05,
      "loss": 0.2277,
      "step": 443300
    },
    {
      "epoch": 0.1868,
      "grad_norm": 0.1422983855009079,
      "learning_rate": 2.26404e-05,
      "loss": 0.2214,
      "step": 443400
    },
    {
      "epoch": 0.187,
      "grad_norm": 0.18516691029071808,
      "learning_rate": 2.26004e-05,
      "loss": 0.2241,
      "step": 443500
    },
    {
      "epoch": 0.1872,
      "grad_norm": 0.13827116787433624,
      "learning_rate": 2.2560400000000002e-05,
      "loss": 0.2277,
      "step": 443600
    },
    {
      "epoch": 0.1874,
      "grad_norm": 0.33428114652633667,
      "learning_rate": 2.25204e-05,
      "loss": 0.2243,
      "step": 443700
    },
    {
      "epoch": 0.1876,
      "grad_norm": 0.2030119001865387,
      "learning_rate": 2.2480400000000002e-05,
      "loss": 0.2277,
      "step": 443800
    },
    {
      "epoch": 0.1878,
      "grad_norm": 0.16090461611747742,
      "learning_rate": 2.24404e-05,
      "loss": 0.2297,
      "step": 443900
    },
    {
      "epoch": 0.188,
      "grad_norm": 0.2067965716123581,
      "learning_rate": 2.2400400000000003e-05,
      "loss": 0.2228,
      "step": 444000
    },
    {
      "epoch": 0.1882,
      "grad_norm": 0.15703140199184418,
      "learning_rate": 2.23604e-05,
      "loss": 0.2282,
      "step": 444100
    },
    {
      "epoch": 0.1884,
      "grad_norm": 0.1881639063358307,
      "learning_rate": 2.2320400000000004e-05,
      "loss": 0.2245,
      "step": 444200
    },
    {
      "epoch": 0.1886,
      "grad_norm": 0.23924614489078522,
      "learning_rate": 2.22804e-05,
      "loss": 0.2267,
      "step": 444300
    },
    {
      "epoch": 0.1888,
      "grad_norm": 0.17333418130874634,
      "learning_rate": 2.22404e-05,
      "loss": 0.2229,
      "step": 444400
    },
    {
      "epoch": 0.189,
      "grad_norm": 0.17022064328193665,
      "learning_rate": 2.2200400000000002e-05,
      "loss": 0.225,
      "step": 444500
    },
    {
      "epoch": 0.1892,
      "grad_norm": 0.17431145906448364,
      "learning_rate": 2.2160400000000002e-05,
      "loss": 0.226,
      "step": 444600
    },
    {
      "epoch": 0.1894,
      "grad_norm": 0.20178723335266113,
      "learning_rate": 2.2120400000000003e-05,
      "loss": 0.226,
      "step": 444700
    },
    {
      "epoch": 0.1896,
      "grad_norm": 0.18924662470817566,
      "learning_rate": 2.20804e-05,
      "loss": 0.2254,
      "step": 444800
    },
    {
      "epoch": 0.1898,
      "grad_norm": 0.23246605694293976,
      "learning_rate": 2.20404e-05,
      "loss": 0.2246,
      "step": 444900
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.15815980732440948,
      "learning_rate": 2.20004e-05,
      "loss": 0.2249,
      "step": 445000
    },
    {
      "epoch": 0.1902,
      "grad_norm": 0.19790607690811157,
      "learning_rate": 2.19604e-05,
      "loss": 0.2302,
      "step": 445100
    },
    {
      "epoch": 0.1904,
      "grad_norm": 0.19628171622753143,
      "learning_rate": 2.19204e-05,
      "loss": 0.2264,
      "step": 445200
    },
    {
      "epoch": 0.1906,
      "grad_norm": 0.16025413572788239,
      "learning_rate": 2.18804e-05,
      "loss": 0.2202,
      "step": 445300
    },
    {
      "epoch": 0.1908,
      "grad_norm": 0.2109999656677246,
      "learning_rate": 2.18404e-05,
      "loss": 0.224,
      "step": 445400
    },
    {
      "epoch": 0.191,
      "grad_norm": 0.17746329307556152,
      "learning_rate": 2.1800400000000002e-05,
      "loss": 0.2253,
      "step": 445500
    },
    {
      "epoch": 0.1912,
      "grad_norm": 0.18444675207138062,
      "learning_rate": 2.17604e-05,
      "loss": 0.2252,
      "step": 445600
    },
    {
      "epoch": 0.1914,
      "grad_norm": 0.21549223363399506,
      "learning_rate": 2.1720400000000003e-05,
      "loss": 0.2265,
      "step": 445700
    },
    {
      "epoch": 0.1916,
      "grad_norm": 0.16874556243419647,
      "learning_rate": 2.16804e-05,
      "loss": 0.2263,
      "step": 445800
    },
    {
      "epoch": 0.1918,
      "grad_norm": 0.14884252846240997,
      "learning_rate": 2.1640400000000004e-05,
      "loss": 0.2252,
      "step": 445900
    },
    {
      "epoch": 0.192,
      "grad_norm": 0.2405053973197937,
      "learning_rate": 2.16004e-05,
      "loss": 0.2267,
      "step": 446000
    },
    {
      "epoch": 0.1922,
      "grad_norm": 0.1512579768896103,
      "learning_rate": 2.15604e-05,
      "loss": 0.2243,
      "step": 446100
    },
    {
      "epoch": 0.1924,
      "grad_norm": 0.15043461322784424,
      "learning_rate": 2.15204e-05,
      "loss": 0.226,
      "step": 446200
    },
    {
      "epoch": 0.1926,
      "grad_norm": 0.14837458729743958,
      "learning_rate": 2.1480400000000002e-05,
      "loss": 0.2251,
      "step": 446300
    },
    {
      "epoch": 0.1928,
      "grad_norm": 0.1467418372631073,
      "learning_rate": 2.1440400000000002e-05,
      "loss": 0.2244,
      "step": 446400
    },
    {
      "epoch": 0.193,
      "grad_norm": 0.16356542706489563,
      "learning_rate": 2.1400400000000003e-05,
      "loss": 0.2255,
      "step": 446500
    },
    {
      "epoch": 0.1932,
      "grad_norm": 0.19749680161476135,
      "learning_rate": 2.13604e-05,
      "loss": 0.2255,
      "step": 446600
    },
    {
      "epoch": 0.1934,
      "grad_norm": 0.20441758632659912,
      "learning_rate": 2.13204e-05,
      "loss": 0.2232,
      "step": 446700
    },
    {
      "epoch": 0.1936,
      "grad_norm": 0.1879335343837738,
      "learning_rate": 2.12804e-05,
      "loss": 0.2306,
      "step": 446800
    },
    {
      "epoch": 0.1938,
      "grad_norm": 0.1568703055381775,
      "learning_rate": 2.12404e-05,
      "loss": 0.2233,
      "step": 446900
    },
    {
      "epoch": 0.194,
      "grad_norm": 0.1516915112733841,
      "learning_rate": 2.12004e-05,
      "loss": 0.228,
      "step": 447000
    },
    {
      "epoch": 0.1942,
      "grad_norm": 0.1923481822013855,
      "learning_rate": 2.1160399999999998e-05,
      "loss": 0.2304,
      "step": 447100
    },
    {
      "epoch": 0.1944,
      "grad_norm": 0.19211551547050476,
      "learning_rate": 2.1120400000000002e-05,
      "loss": 0.2253,
      "step": 447200
    },
    {
      "epoch": 0.1946,
      "grad_norm": 0.2306138575077057,
      "learning_rate": 2.10804e-05,
      "loss": 0.2228,
      "step": 447300
    },
    {
      "epoch": 0.1948,
      "grad_norm": 0.18642979860305786,
      "learning_rate": 2.1040400000000003e-05,
      "loss": 0.2235,
      "step": 447400
    },
    {
      "epoch": 0.195,
      "grad_norm": 0.1702134609222412,
      "learning_rate": 2.10004e-05,
      "loss": 0.2297,
      "step": 447500
    },
    {
      "epoch": 0.1952,
      "grad_norm": 0.187478706240654,
      "learning_rate": 2.0960400000000003e-05,
      "loss": 0.2291,
      "step": 447600
    },
    {
      "epoch": 0.1954,
      "grad_norm": 0.1675885021686554,
      "learning_rate": 2.09204e-05,
      "loss": 0.2257,
      "step": 447700
    },
    {
      "epoch": 0.1956,
      "grad_norm": 0.20350529253482819,
      "learning_rate": 2.08804e-05,
      "loss": 0.225,
      "step": 447800
    },
    {
      "epoch": 0.1958,
      "grad_norm": 0.1661386489868164,
      "learning_rate": 2.08404e-05,
      "loss": 0.2274,
      "step": 447900
    },
    {
      "epoch": 0.196,
      "grad_norm": 0.15800422430038452,
      "learning_rate": 2.08004e-05,
      "loss": 0.2311,
      "step": 448000
    },
    {
      "epoch": 0.1962,
      "grad_norm": 0.14770348370075226,
      "learning_rate": 2.0760400000000002e-05,
      "loss": 0.2255,
      "step": 448100
    },
    {
      "epoch": 0.1964,
      "grad_norm": 0.22858108580112457,
      "learning_rate": 2.0720400000000002e-05,
      "loss": 0.225,
      "step": 448200
    },
    {
      "epoch": 0.1966,
      "grad_norm": 0.1924201101064682,
      "learning_rate": 2.06804e-05,
      "loss": 0.222,
      "step": 448300
    },
    {
      "epoch": 0.1968,
      "grad_norm": 0.2220558375120163,
      "learning_rate": 2.06404e-05,
      "loss": 0.2284,
      "step": 448400
    },
    {
      "epoch": 0.197,
      "grad_norm": 0.17530234158039093,
      "learning_rate": 2.06004e-05,
      "loss": 0.2256,
      "step": 448500
    },
    {
      "epoch": 0.1972,
      "grad_norm": 0.19386076927185059,
      "learning_rate": 2.05604e-05,
      "loss": 0.2267,
      "step": 448600
    },
    {
      "epoch": 0.1974,
      "grad_norm": 0.38340792059898376,
      "learning_rate": 2.05204e-05,
      "loss": 0.2245,
      "step": 448700
    },
    {
      "epoch": 0.1976,
      "grad_norm": 0.2019384652376175,
      "learning_rate": 2.04804e-05,
      "loss": 0.225,
      "step": 448800
    },
    {
      "epoch": 0.1978,
      "grad_norm": 0.16011062264442444,
      "learning_rate": 2.04404e-05,
      "loss": 0.228,
      "step": 448900
    },
    {
      "epoch": 0.198,
      "grad_norm": 0.2042662352323532,
      "learning_rate": 2.04004e-05,
      "loss": 0.2257,
      "step": 449000
    },
    {
      "epoch": 0.1982,
      "grad_norm": 0.15106824040412903,
      "learning_rate": 2.0360400000000002e-05,
      "loss": 0.2236,
      "step": 449100
    },
    {
      "epoch": 0.1984,
      "grad_norm": 0.15230531990528107,
      "learning_rate": 2.03204e-05,
      "loss": 0.2222,
      "step": 449200
    },
    {
      "epoch": 0.1986,
      "grad_norm": 0.1945761889219284,
      "learning_rate": 2.0280400000000003e-05,
      "loss": 0.2256,
      "step": 449300
    },
    {
      "epoch": 0.1988,
      "grad_norm": 0.2893814444541931,
      "learning_rate": 2.02404e-05,
      "loss": 0.2289,
      "step": 449400
    },
    {
      "epoch": 0.199,
      "grad_norm": 0.16829931735992432,
      "learning_rate": 2.02004e-05,
      "loss": 0.2241,
      "step": 449500
    },
    {
      "epoch": 0.1992,
      "grad_norm": 0.29933756589889526,
      "learning_rate": 2.01604e-05,
      "loss": 0.2259,
      "step": 449600
    },
    {
      "epoch": 0.1994,
      "grad_norm": 0.151699960231781,
      "learning_rate": 2.01204e-05,
      "loss": 0.2214,
      "step": 449700
    },
    {
      "epoch": 0.1996,
      "grad_norm": 0.2555668354034424,
      "learning_rate": 2.0080400000000002e-05,
      "loss": 0.2268,
      "step": 449800
    },
    {
      "epoch": 0.1998,
      "grad_norm": 0.18724453449249268,
      "learning_rate": 2.0040400000000002e-05,
      "loss": 0.2262,
      "step": 449900
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.21397201716899872,
      "learning_rate": 2.00004e-05,
      "loss": 0.2222,
      "step": 450000
    },
    {
      "epoch": 0.2002,
      "grad_norm": 0.31661924719810486,
      "learning_rate": 1.9960400000000003e-05,
      "loss": 0.2524,
      "step": 450100
    },
    {
      "epoch": 0.2004,
      "grad_norm": 0.28846636414527893,
      "learning_rate": 1.99204e-05,
      "loss": 0.2639,
      "step": 450200
    },
    {
      "epoch": 0.2006,
      "grad_norm": 0.2062097191810608,
      "learning_rate": 1.98804e-05,
      "loss": 0.2605,
      "step": 450300
    },
    {
      "epoch": 0.2008,
      "grad_norm": 0.37499383091926575,
      "learning_rate": 1.98404e-05,
      "loss": 0.2526,
      "step": 450400
    },
    {
      "epoch": 0.201,
      "grad_norm": 0.2817152142524719,
      "learning_rate": 1.98004e-05,
      "loss": 0.255,
      "step": 450500
    },
    {
      "epoch": 0.2012,
      "grad_norm": 0.2787519097328186,
      "learning_rate": 1.97604e-05,
      "loss": 0.2779,
      "step": 450600
    },
    {
      "epoch": 0.2014,
      "grad_norm": 0.3081032931804657,
      "learning_rate": 1.97204e-05,
      "loss": 0.2528,
      "step": 450700
    },
    {
      "epoch": 0.2016,
      "grad_norm": 0.45167797803878784,
      "learning_rate": 1.9680400000000002e-05,
      "loss": 0.2531,
      "step": 450800
    },
    {
      "epoch": 0.2018,
      "grad_norm": 0.2237127274274826,
      "learning_rate": 1.96404e-05,
      "loss": 0.2742,
      "step": 450900
    },
    {
      "epoch": 0.202,
      "grad_norm": 0.18586333096027374,
      "learning_rate": 1.9600400000000003e-05,
      "loss": 0.2499,
      "step": 451000
    },
    {
      "epoch": 0.2022,
      "grad_norm": 0.2039756029844284,
      "learning_rate": 1.95604e-05,
      "loss": 0.2538,
      "step": 451100
    },
    {
      "epoch": 0.2024,
      "grad_norm": 0.2960463762283325,
      "learning_rate": 1.95204e-05,
      "loss": 0.2555,
      "step": 451200
    },
    {
      "epoch": 0.2026,
      "grad_norm": 0.5285388827323914,
      "learning_rate": 1.94804e-05,
      "loss": 0.2882,
      "step": 451300
    },
    {
      "epoch": 0.2028,
      "grad_norm": 0.20539702475070953,
      "learning_rate": 1.94404e-05,
      "loss": 0.2681,
      "step": 451400
    },
    {
      "epoch": 0.203,
      "grad_norm": 0.40104198455810547,
      "learning_rate": 1.94004e-05,
      "loss": 0.2579,
      "step": 451500
    },
    {
      "epoch": 0.2032,
      "grad_norm": 0.24579091370105743,
      "learning_rate": 1.9360400000000002e-05,
      "loss": 0.2561,
      "step": 451600
    },
    {
      "epoch": 0.2034,
      "grad_norm": 0.24217070639133453,
      "learning_rate": 1.93204e-05,
      "loss": 0.251,
      "step": 451700
    },
    {
      "epoch": 0.2036,
      "grad_norm": 0.18442483246326447,
      "learning_rate": 1.9280400000000003e-05,
      "loss": 0.2504,
      "step": 451800
    },
    {
      "epoch": 0.2038,
      "grad_norm": 0.24910864233970642,
      "learning_rate": 1.92404e-05,
      "loss": 0.3131,
      "step": 451900
    },
    {
      "epoch": 0.204,
      "grad_norm": 0.2014876753091812,
      "learning_rate": 1.9200400000000003e-05,
      "loss": 0.2991,
      "step": 452000
    },
    {
      "epoch": 0.2042,
      "grad_norm": 0.1724986732006073,
      "learning_rate": 1.91604e-05,
      "loss": 0.2537,
      "step": 452100
    },
    {
      "epoch": 0.2044,
      "grad_norm": 0.16175585985183716,
      "learning_rate": 1.91204e-05,
      "loss": 0.2581,
      "step": 452200
    },
    {
      "epoch": 0.2046,
      "grad_norm": 0.29984617233276367,
      "learning_rate": 1.90804e-05,
      "loss": 0.2608,
      "step": 452300
    },
    {
      "epoch": 0.2048,
      "grad_norm": 0.32500120997428894,
      "learning_rate": 1.90404e-05,
      "loss": 0.2811,
      "step": 452400
    },
    {
      "epoch": 0.205,
      "grad_norm": 0.14806152880191803,
      "learning_rate": 1.9000400000000002e-05,
      "loss": 0.2499,
      "step": 452500
    },
    {
      "epoch": 0.2052,
      "grad_norm": 0.1928393542766571,
      "learning_rate": 1.89604e-05,
      "loss": 0.2671,
      "step": 452600
    },
    {
      "epoch": 0.2054,
      "grad_norm": 0.626834511756897,
      "learning_rate": 1.8920400000000003e-05,
      "loss": 0.2681,
      "step": 452700
    },
    {
      "epoch": 0.2056,
      "grad_norm": 0.2420172393321991,
      "learning_rate": 1.88804e-05,
      "loss": 0.2626,
      "step": 452800
    },
    {
      "epoch": 0.2058,
      "grad_norm": 0.5636458396911621,
      "learning_rate": 1.88404e-05,
      "loss": 0.2625,
      "step": 452900
    },
    {
      "epoch": 0.206,
      "grad_norm": 1.392844319343567,
      "learning_rate": 1.88004e-05,
      "loss": 0.2567,
      "step": 453000
    },
    {
      "epoch": 0.2062,
      "grad_norm": 0.22444498538970947,
      "learning_rate": 1.87604e-05,
      "loss": 0.2405,
      "step": 453100
    },
    {
      "epoch": 0.2064,
      "grad_norm": 0.3851085603237152,
      "learning_rate": 1.87204e-05,
      "loss": 0.2609,
      "step": 453200
    },
    {
      "epoch": 0.2066,
      "grad_norm": 0.2355654239654541,
      "learning_rate": 1.86804e-05,
      "loss": 0.2557,
      "step": 453300
    },
    {
      "epoch": 0.2068,
      "grad_norm": 0.20136545598506927,
      "learning_rate": 1.86404e-05,
      "loss": 0.2569,
      "step": 453400
    },
    {
      "epoch": 0.207,
      "grad_norm": 0.2145121991634369,
      "learning_rate": 1.8600400000000002e-05,
      "loss": 0.2602,
      "step": 453500
    },
    {
      "epoch": 0.2072,
      "grad_norm": 0.20145122706890106,
      "learning_rate": 1.85604e-05,
      "loss": 0.2648,
      "step": 453600
    },
    {
      "epoch": 0.2074,
      "grad_norm": 0.2964533269405365,
      "learning_rate": 1.8520400000000003e-05,
      "loss": 0.2509,
      "step": 453700
    },
    {
      "epoch": 0.2076,
      "grad_norm": 0.22049467265605927,
      "learning_rate": 1.84804e-05,
      "loss": 0.3241,
      "step": 453800
    },
    {
      "epoch": 0.2078,
      "grad_norm": 0.2652212679386139,
      "learning_rate": 1.84404e-05,
      "loss": 0.2527,
      "step": 453900
    },
    {
      "epoch": 0.208,
      "grad_norm": 0.17262853682041168,
      "learning_rate": 1.84004e-05,
      "loss": 0.2512,
      "step": 454000
    },
    {
      "epoch": 0.2082,
      "grad_norm": 0.4729574918746948,
      "learning_rate": 1.83604e-05,
      "loss": 0.257,
      "step": 454100
    },
    {
      "epoch": 0.2084,
      "grad_norm": 0.20896553993225098,
      "learning_rate": 1.83204e-05,
      "loss": 0.257,
      "step": 454200
    },
    {
      "epoch": 0.2086,
      "grad_norm": 0.32045525312423706,
      "learning_rate": 1.82804e-05,
      "loss": 0.2587,
      "step": 454300
    },
    {
      "epoch": 0.2088,
      "grad_norm": 0.27513304352760315,
      "learning_rate": 1.8240400000000002e-05,
      "loss": 0.2748,
      "step": 454400
    },
    {
      "epoch": 0.209,
      "grad_norm": 0.3870587944984436,
      "learning_rate": 1.82004e-05,
      "loss": 0.2626,
      "step": 454500
    },
    {
      "epoch": 0.2092,
      "grad_norm": 0.18776573240756989,
      "learning_rate": 1.81604e-05,
      "loss": 0.2574,
      "step": 454600
    },
    {
      "epoch": 0.2094,
      "grad_norm": 0.2663273513317108,
      "learning_rate": 1.81204e-05,
      "loss": 0.2613,
      "step": 454700
    },
    {
      "epoch": 0.2096,
      "grad_norm": 0.20890645682811737,
      "learning_rate": 1.80804e-05,
      "loss": 0.2612,
      "step": 454800
    },
    {
      "epoch": 0.2098,
      "grad_norm": 0.19033282995224,
      "learning_rate": 1.80404e-05,
      "loss": 0.2611,
      "step": 454900
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.1746765375137329,
      "learning_rate": 1.80004e-05,
      "loss": 0.2456,
      "step": 455000
    }
  ],
  "logging_steps": 100,
  "max_steps": 500000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 5000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 4.8875100438528e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}