{
  "best_metric": 0.239013671875,
  "best_model_checkpoint": "./results_morgangen_auto/checkpoint-240000",
  "epoch": 0.0024,
  "eval_steps": 20000,
  "global_step": 240000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1e-06,
      "grad_norm": 7.192080497741699,
      "learning_rate": 4.944309013222119e-06,
      "loss": 3.578,
      "step": 100
    },
    {
      "epoch": 2e-06,
      "grad_norm": 5.5353922843933105,
      "learning_rate": 5.725086528406295e-06,
      "loss": 1.9606,
      "step": 200
    },
    {
      "epoch": 3e-06,
      "grad_norm": 6.314189910888672,
      "learning_rate": 6.1745550399454076e-06,
      "loss": 1.5525,
      "step": 300
    },
    {
      "epoch": 4e-06,
      "grad_norm": 4.842369556427002,
      "learning_rate": 6.4914927390661495e-06,
      "loss": 1.3692,
      "step": 400
    },
    {
      "epoch": 5e-06,
      "grad_norm": 6.878078460693359,
      "learning_rate": 6.736512997333922e-06,
      "loss": 1.2631,
      "step": 500
    },
    {
      "epoch": 6e-06,
      "grad_norm": 5.2255377769470215,
      "learning_rate": 6.936292414321374e-06,
      "loss": 1.2033,
      "step": 600
    },
    {
      "epoch": 7e-06,
      "grad_norm": 6.255804538726807,
      "learning_rate": 7.104962011475284e-06,
      "loss": 1.176,
      "step": 700
    },
    {
      "epoch": 8e-06,
      "grad_norm": 5.792397499084473,
      "learning_rate": 7.250917821641176e-06,
      "loss": 1.1391,
      "step": 800
    },
    {
      "epoch": 9e-06,
      "grad_norm": 6.366394996643066,
      "learning_rate": 7.37955758828978e-06,
      "loss": 1.0937,
      "step": 900
    },
    {
      "epoch": 1e-05,
      "grad_norm": 5.171053886413574,
      "learning_rate": 7.494557701864313e-06,
      "loss": 1.0665,
      "step": 1000
    },
    {
      "epoch": 1.1e-05,
      "grad_norm": 5.637969017028809,
      "learning_rate": 7.598535297940343e-06,
      "loss": 1.0483,
      "step": 1100
    },
    {
      "epoch": 1.2e-05,
      "grad_norm": 5.52170467376709,
      "learning_rate": 7.69341976321039e-06,
      "loss": 1.0372,
      "step": 1200
    },
    {
      "epoch": 1.3e-05,
      "grad_norm": 4.622586727142334,
      "learning_rate": 7.780674421043177e-06,
      "loss": 1.0008,
      "step": 1300
    },
    {
      "epoch": 1.4e-05,
      "grad_norm": 4.899344444274902,
      "learning_rate": 7.86143551902404e-06,
      "loss": 0.9829,
      "step": 1400
    },
    {
      "epoch": 1.5e-05,
      "grad_norm": 4.919180870056152,
      "learning_rate": 7.936602981651121e-06,
      "loss": 0.9821,
      "step": 1500
    },
    {
      "epoch": 1.6e-05,
      "grad_norm": 5.980068206787109,
      "learning_rate": 8.006901718483e-06,
      "loss": 0.9506,
      "step": 1600
    },
    {
      "epoch": 1.7e-05,
      "grad_norm": 5.955260276794434,
      "learning_rate": 8.072924256347751e-06,
      "loss": 0.9382,
      "step": 1700
    },
    {
      "epoch": 1.8e-05,
      "grad_norm": 4.828701019287109,
      "learning_rate": 8.135161132285844e-06,
      "loss": 0.923,
      "step": 1800
    },
    {
      "epoch": 1.9e-05,
      "grad_norm": 5.517646789550781,
      "learning_rate": 8.194023035760226e-06,
      "loss": 0.9019,
      "step": 1900
    },
    {
      "epoch": 2e-05,
      "grad_norm": 4.49784517288208,
      "learning_rate": 8.249857250056917e-06,
      "loss": 0.9013,
      "step": 2000
    },
    {
      "epoch": 2.1e-05,
      "grad_norm": 4.274550914764404,
      "learning_rate": 8.302960068255784e-06,
      "loss": 0.9038,
      "step": 2100
    },
    {
      "epoch": 2.2e-05,
      "grad_norm": 5.395852565765381,
      "learning_rate": 8.35358631144535e-06,
      "loss": 0.9049,
      "step": 2200
    },
    {
      "epoch": 2.3e-05,
      "grad_norm": 4.8221611976623535,
      "learning_rate": 8.4019567246832e-06,
      "loss": 0.8675,
      "step": 2300
    },
    {
      "epoch": 2.4e-05,
      "grad_norm": 4.574971675872803,
      "learning_rate": 8.448263794376455e-06,
      "loss": 0.8743,
      "step": 2400
    },
    {
      "epoch": 2.5e-05,
      "grad_norm": 4.795435905456543,
      "learning_rate": 8.492676374898522e-06,
      "loss": 0.8473,
      "step": 2500
    },
    {
      "epoch": 2.6e-05,
      "grad_norm": 4.126605987548828,
      "learning_rate": 8.535343405461191e-06,
      "loss": 0.8553,
      "step": 2600
    },
    {
      "epoch": 2.7e-05,
      "grad_norm": 4.7121992111206055,
      "learning_rate": 8.576396923806893e-06,
      "loss": 0.8496,
      "step": 2700
    },
    {
      "epoch": 2.8e-05,
      "grad_norm": 4.441779136657715,
      "learning_rate": 8.615954530556104e-06,
      "loss": 0.8336,
      "step": 2800
    },
    {
      "epoch": 2.9e-05,
      "grad_norm": 4.51931095123291,
      "learning_rate": 8.654121420158637e-06,
      "loss": 0.8287,
      "step": 2900
    },
    {
      "epoch": 3e-05,
      "grad_norm": 4.081902980804443,
      "learning_rate": 8.690992066813325e-06,
      "loss": 0.8168,
      "step": 3000
    },
    {
      "epoch": 3.1e-05,
      "grad_norm": 4.671773910522461,
      "learning_rate": 8.726651633390342e-06,
      "loss": 0.8166,
      "step": 3100
    },
    {
      "epoch": 3.2e-05,
      "grad_norm": 4.348089694976807,
      "learning_rate": 8.76083727950616e-06,
      "loss": 0.81,
      "step": 3200
    },
    {
      "epoch": 3.3e-05,
      "grad_norm": 4.5769829750061035,
      "learning_rate": 8.794308987084338e-06,
      "loss": 0.7987,
      "step": 3300
    },
    {
      "epoch": 3.4e-05,
      "grad_norm": 4.201402187347412,
      "learning_rate": 8.826459649112795e-06,
      "loss": 0.8051,
      "step": 3400
    },
    {
      "epoch": 3.5e-05,
      "grad_norm": 4.153467655181885,
      "learning_rate": 8.857996464094115e-06,
      "loss": 0.8102,
      "step": 3500
    },
    {
      "epoch": 3.6e-05,
      "grad_norm": 4.17486047744751,
      "learning_rate": 8.888643043011622e-06,
      "loss": 0.7847,
      "step": 3600
    },
    {
      "epoch": 3.7e-05,
      "grad_norm": 4.006905555725098,
      "learning_rate": 8.918448269127447e-06,
      "loss": 0.7765,
      "step": 3700
    },
    {
      "epoch": 3.8e-05,
      "grad_norm": 3.959653377532959,
      "learning_rate": 8.947170821665072e-06,
      "loss": 0.7704,
      "step": 3800
    },
    {
      "epoch": 3.9e-05,
      "grad_norm": 5.123909950256348,
      "learning_rate": 8.975432078990786e-06,
      "loss": 0.7671,
      "step": 3900
    },
    {
      "epoch": 4e-05,
      "grad_norm": 3.873758316040039,
      "learning_rate": 9.002976331538332e-06,
      "loss": 0.7567,
      "step": 4000
    },
    {
      "epoch": 4.1e-05,
      "grad_norm": 4.340254783630371,
      "learning_rate": 9.029839062600307e-06,
      "loss": 0.7416,
      "step": 4100
    },
    {
      "epoch": 4.2e-05,
      "grad_norm": 4.056208610534668,
      "learning_rate": 9.056053184939176e-06,
      "loss": 0.7445,
      "step": 4200
    },
    {
      "epoch": 4.3e-05,
      "grad_norm": 3.6428372859954834,
      "learning_rate": 9.081649283234784e-06,
      "loss": 0.7456,
      "step": 4300
    },
    {
      "epoch": 4.4e-05,
      "grad_norm": 3.506011486053467,
      "learning_rate": 9.106655828605087e-06,
      "loss": 0.7342,
      "step": 4400
    },
    {
      "epoch": 4.5e-05,
      "grad_norm": 3.7343637943267822,
      "learning_rate": 9.13109936897355e-06,
      "loss": 0.7435,
      "step": 4500
    },
    {
      "epoch": 4.6e-05,
      "grad_norm": 3.65120267868042,
      "learning_rate": 9.155004698474792e-06,
      "loss": 0.7346,
      "step": 4600
    },
    {
      "epoch": 4.7e-05,
      "grad_norm": 3.66025710105896,
      "learning_rate": 9.17839500860873e-06,
      "loss": 0.7394,
      "step": 4700
    },
    {
      "epoch": 4.8e-05,
      "grad_norm": 3.7857818603515625,
      "learning_rate": 9.201292023453135e-06,
      "loss": 0.7137,
      "step": 4800
    },
    {
      "epoch": 4.9e-05,
      "grad_norm": 4.2525858879089355,
      "learning_rate": 9.22371612091062e-06,
      "loss": 0.7179,
      "step": 4900
    },
    {
      "epoch": 5e-05,
      "grad_norm": 3.6449532508850098,
      "learning_rate": 9.245686441685918e-06,
      "loss": 0.7149,
      "step": 5000
    },
    {
      "epoch": 5.1e-05,
      "grad_norm": 4.4344401359558105,
      "learning_rate": 9.267220987454044e-06,
      "loss": 0.7209,
      "step": 5100
    },
    {
      "epoch": 5.2e-05,
      "grad_norm": 3.7667882442474365,
      "learning_rate": 9.28833670948078e-06,
      "loss": 0.7104,
      "step": 5200
    },
    {
      "epoch": 5.3e-05,
      "grad_norm": 3.8509140014648438,
      "learning_rate": 9.309049588788657e-06,
      "loss": 0.6939,
      "step": 5300
    },
    {
      "epoch": 5.4e-05,
      "grad_norm": 4.200638294219971,
      "learning_rate": 9.329374708818158e-06,
      "loss": 0.6934,
      "step": 5400
    },
    {
      "epoch": 5.5e-05,
      "grad_norm": 3.6815011501312256,
      "learning_rate": 9.349326321411793e-06,
      "loss": 0.6841,
      "step": 5500
    },
    {
      "epoch": 5.6e-05,
      "grad_norm": 4.6741719245910645,
      "learning_rate": 9.368917906844062e-06,
      "loss": 0.7124,
      "step": 5600
    },
    {
      "epoch": 5.7e-05,
      "grad_norm": 3.726712942123413,
      "learning_rate": 9.388162228530614e-06,
      "loss": 0.6749,
      "step": 5700
    },
    {
      "epoch": 5.8e-05,
      "grad_norm": 3.2452657222747803,
      "learning_rate": 9.407071382972726e-06,
      "loss": 0.7073,
      "step": 5800
    },
    {
      "epoch": 5.9e-05,
      "grad_norm": 3.7711005210876465,
      "learning_rate": 9.425656845426483e-06,
      "loss": 0.6816,
      "step": 5900
    },
    {
      "epoch": 6e-05,
      "grad_norm": 3.617072105407715,
      "learning_rate": 9.443929511728523e-06,
      "loss": 0.6788,
      "step": 6000
    },
    {
      "epoch": 6.1e-05,
      "grad_norm": 3.677022933959961,
      "learning_rate": 9.461721498753552e-06,
      "loss": 0.6707,
      "step": 6100
    },
    {
      "epoch": 6.2e-05,
      "grad_norm": 3.563831329345703,
      "learning_rate": 9.479402010032261e-06,
      "loss": 0.6688,
      "step": 6200
    },
    {
      "epoch": 6.3e-05,
      "grad_norm": 5.436884880065918,
      "learning_rate": 9.496799212962515e-06,
      "loss": 0.6667,
      "step": 6300
    },
    {
      "epoch": 6.4e-05,
      "grad_norm": 3.69850754737854,
      "learning_rate": 9.51392204387139e-06,
      "loss": 0.665,
      "step": 6400
    },
    {
      "epoch": 6.5e-05,
      "grad_norm": 4.021072864532471,
      "learning_rate": 9.530779022827808e-06,
      "loss": 0.6652,
      "step": 6500
    },
    {
      "epoch": 6.6e-05,
      "grad_norm": 3.9425106048583984,
      "learning_rate": 9.547378279100432e-06,
      "loss": 0.6593,
      "step": 6600
    },
    {
      "epoch": 6.7e-05,
      "grad_norm": 4.435314655303955,
      "learning_rate": 9.563727574698575e-06,
      "loss": 0.6521,
      "step": 6700
    },
    {
      "epoch": 6.8e-05,
      "grad_norm": 3.426171064376831,
      "learning_rate": 9.579674435701252e-06,
      "loss": 0.6592,
      "step": 6800
    },
    {
      "epoch": 6.9e-05,
      "grad_norm": 3.2786061763763428,
      "learning_rate": 9.595548054769063e-06,
      "loss": 0.659,
      "step": 6900
    },
    {
      "epoch": 7e-05,
      "grad_norm": 3.6172714233398438,
      "learning_rate": 9.611192939364202e-06,
      "loss": 0.6425,
      "step": 7000
    },
    {
      "epoch": 7.1e-05,
      "grad_norm": 3.1984446048736572,
      "learning_rate": 9.626615587957666e-06,
      "loss": 0.6476,
      "step": 7100
    },
    {
      "epoch": 7.2e-05,
      "grad_norm": 4.468976974487305,
      "learning_rate": 9.641822225957206e-06,
      "loss": 0.6394,
      "step": 7200
    },
    {
      "epoch": 7.3e-05,
      "grad_norm": 3.582566261291504,
      "learning_rate": 9.656818820794936e-06,
      "loss": 0.6312,
      "step": 7300
    },
    {
      "epoch": 7.4e-05,
      "grad_norm": 3.2956085205078125,
      "learning_rate": 9.671611095987065e-06,
      "loss": 0.6349,
      "step": 7400
    },
    {
      "epoch": 7.5e-05,
      "grad_norm": 3.747368574142456,
      "learning_rate": 9.686204544248665e-06,
      "loss": 0.641,
      "step": 7500
    },
    {
      "epoch": 7.6e-05,
      "grad_norm": 3.5763564109802246,
      "learning_rate": 9.7006044397387e-06,
      "loss": 0.635,
      "step": 7600
    },
    {
      "epoch": 7.7e-05,
      "grad_norm": 3.662113666534424,
      "learning_rate": 9.714815849503578e-06,
      "loss": 0.6216,
      "step": 7700
    },
    {
      "epoch": 7.8e-05,
      "grad_norm": 4.00654935836792,
      "learning_rate": 9.728843644181411e-06,
      "loss": 0.6385,
      "step": 7800
    },
    {
      "epoch": 7.9e-05,
      "grad_norm": 3.1744225025177,
      "learning_rate": 9.74269250802355e-06,
      "loss": 0.6225,
      "step": 7900
    },
    {
      "epoch": 8e-05,
      "grad_norm": 2.9433505535125732,
      "learning_rate": 9.756366948284976e-06,
      "loss": 0.6218,
      "step": 8000
    },
    {
      "epoch": 8.1e-05,
      "grad_norm": 3.351646661758423,
      "learning_rate": 9.76987130403068e-06,
      "loss": 0.6156,
      "step": 8100
    },
    {
      "epoch": 8.2e-05,
      "grad_norm": 3.297624111175537,
      "learning_rate": 9.783209754401046e-06,
      "loss": 0.6105,
      "step": 8200
    },
    {
      "epoch": 8.3e-05,
      "grad_norm": 3.242459535598755,
      "learning_rate": 9.796386326375682e-06,
      "loss": 0.6121,
      "step": 8300
    },
    {
      "epoch": 8.4e-05,
      "grad_norm": 3.301387071609497,
      "learning_rate": 9.80940490207175e-06,
      "loss": 0.6103,
      "step": 8400
    },
    {
      "epoch": 8.5e-05,
      "grad_norm": 3.537336587905884,
      "learning_rate": 9.822269225609881e-06,
      "loss": 0.5999,
      "step": 8500
    },
    {
      "epoch": 8.6e-05,
      "grad_norm": 3.297769784927368,
      "learning_rate": 9.834856506853153e-06,
      "loss": 0.6137,
      "step": 8600
    },
    {
      "epoch": 8.7e-05,
      "grad_norm": 3.5436480045318604,
      "learning_rate": 9.847424493057225e-06,
      "loss": 0.5982,
      "step": 8700
    },
    {
      "epoch": 8.8e-05,
      "grad_norm": 3.3553879261016846,
      "learning_rate": 9.85984866118054e-06,
      "loss": 0.5987,
      "step": 8800
    },
    {
      "epoch": 8.9e-05,
      "grad_norm": 2.8393940925598145,
      "learning_rate": 9.872010114832027e-06,
      "loss": 0.5954,
      "step": 8900
    },
    {
      "epoch": 9e-05,
      "grad_norm": 3.286961317062378,
      "learning_rate": 9.884157659367727e-06,
      "loss": 0.6128,
      "step": 9000
    },
    {
      "epoch": 9.1e-05,
      "grad_norm": 3.427604913711548,
      "learning_rate": 9.896170795917358e-06,
      "loss": 0.5997,
      "step": 9100
    },
    {
      "epoch": 9.2e-05,
      "grad_norm": 4.3479323387146,
      "learning_rate": 9.908052466307471e-06,
      "loss": 0.592,
      "step": 9200
    },
    {
      "epoch": 9.3e-05,
      "grad_norm": 3.0680465698242188,
      "learning_rate": 9.919805516826294e-06,
      "loss": 0.5973,
      "step": 9300
    },
    {
      "epoch": 9.4e-05,
      "grad_norm": 3.6079111099243164,
      "learning_rate": 9.931432702316388e-06,
      "loss": 0.5995,
      "step": 9400
    },
    {
      "epoch": 9.5e-05,
      "grad_norm": 3.1534416675567627,
      "learning_rate": 9.942936690050469e-06,
      "loss": 0.5899,
      "step": 9500
    },
    {
      "epoch": 9.6e-05,
      "grad_norm": 2.7021210193634033,
      "learning_rate": 9.95432006340404e-06,
      "loss": 0.5963,
      "step": 9600
    },
    {
      "epoch": 9.7e-05,
      "grad_norm": 2.8992557525634766,
      "learning_rate": 9.965585325337488e-06,
      "loss": 0.5919,
      "step": 9700
    },
    {
      "epoch": 9.8e-05,
      "grad_norm": 3.391969680786133,
      "learning_rate": 9.976734901699378e-06,
      "loss": 0.5814,
      "step": 9800
    },
    {
      "epoch": 9.9e-05,
      "grad_norm": 2.826235771179199,
      "learning_rate": 9.987771144361851e-06,
      "loss": 0.5675,
      "step": 9900
    },
    {
      "epoch": 0.0001,
      "grad_norm": 3.625103235244751,
      "learning_rate": 9.998696334198274e-06,
      "loss": 0.589,
      "step": 10000
    },
    {
      "epoch": 0.000101,
      "grad_norm": 3.538278341293335,
      "learning_rate": 1e-05,
      "loss": 0.5863,
      "step": 10100
    },
    {
      "epoch": 0.000102,
      "grad_norm": 3.2610013484954834,
      "learning_rate": 1e-05,
      "loss": 0.5853,
      "step": 10200
    },
    {
      "epoch": 0.000103,
      "grad_norm": 3.018453359603882,
      "learning_rate": 1e-05,
      "loss": 0.581,
      "step": 10300
    },
    {
      "epoch": 0.000104,
      "grad_norm": 2.997459650039673,
      "learning_rate": 1e-05,
      "loss": 0.5749,
      "step": 10400
    },
    {
      "epoch": 0.000105,
      "grad_norm": 3.1793456077575684,
      "learning_rate": 1e-05,
      "loss": 0.5624,
      "step": 10500
    },
    {
      "epoch": 0.000106,
      "grad_norm": 3.571202039718628,
      "learning_rate": 1e-05,
      "loss": 0.574,
      "step": 10600
    },
    {
      "epoch": 0.000107,
      "grad_norm": 3.742325782775879,
      "learning_rate": 1e-05,
      "loss": 0.5736,
      "step": 10700
    },
    {
      "epoch": 0.000108,
      "grad_norm": 3.2514116764068604,
      "learning_rate": 1e-05,
      "loss": 0.5611,
      "step": 10800
    },
    {
      "epoch": 0.000109,
      "grad_norm": 3.0863165855407715,
      "learning_rate": 1e-05,
      "loss": 0.5564,
      "step": 10900
    },
    {
      "epoch": 0.00011,
      "grad_norm": 3.638606071472168,
      "learning_rate": 1e-05,
      "loss": 0.5711,
      "step": 11000
    },
    {
      "epoch": 0.000111,
      "grad_norm": 2.756107807159424,
      "learning_rate": 1e-05,
      "loss": 0.5711,
      "step": 11100
    },
    {
      "epoch": 0.000112,
      "grad_norm": 3.2166287899017334,
      "learning_rate": 1e-05,
      "loss": 0.5781,
      "step": 11200
    },
    {
      "epoch": 0.000113,
      "grad_norm": 2.7138330936431885,
      "learning_rate": 1e-05,
      "loss": 0.5522,
      "step": 11300
    },
    {
      "epoch": 0.000114,
      "grad_norm": 3.1922643184661865,
      "learning_rate": 1e-05,
      "loss": 0.5686,
      "step": 11400
    },
    {
      "epoch": 0.000115,
      "grad_norm": 3.2307920455932617,
      "learning_rate": 1e-05,
      "loss": 0.5621,
      "step": 11500
    },
    {
      "epoch": 0.000116,
      "grad_norm": 2.9271585941314697,
      "learning_rate": 1e-05,
      "loss": 0.554,
      "step": 11600
    },
    {
      "epoch": 0.000117,
      "grad_norm": 2.994710922241211,
      "learning_rate": 1e-05,
      "loss": 0.5564,
      "step": 11700
    },
    {
      "epoch": 0.000118,
      "grad_norm": 3.0383167266845703,
      "learning_rate": 1e-05,
      "loss": 0.5584,
      "step": 11800
    },
    {
      "epoch": 0.000119,
      "grad_norm": 2.635859489440918,
      "learning_rate": 1e-05,
      "loss": 0.5657,
      "step": 11900
    },
    {
      "epoch": 0.00012,
      "grad_norm": 2.850497245788574,
      "learning_rate": 1e-05,
      "loss": 0.5538,
      "step": 12000
    },
    {
      "epoch": 0.000121,
      "grad_norm": 3.060102701187134,
      "learning_rate": 1e-05,
      "loss": 0.5671,
      "step": 12100
    },
    {
      "epoch": 0.000122,
      "grad_norm": 2.882080316543579,
      "learning_rate": 1e-05,
      "loss": 0.5563,
      "step": 12200
    },
    {
      "epoch": 0.000123,
      "grad_norm": 2.516627311706543,
      "learning_rate": 1e-05,
      "loss": 0.5499,
      "step": 12300
    },
    {
      "epoch": 0.000124,
      "grad_norm": 3.2647488117218018,
      "learning_rate": 1e-05,
      "loss": 0.5523,
      "step": 12400
    },
    {
      "epoch": 0.000125,
      "grad_norm": 3.0820746421813965,
      "learning_rate": 1e-05,
      "loss": 0.5551,
      "step": 12500
    },
    {
      "epoch": 0.000126,
      "grad_norm": 3.288663864135742,
      "learning_rate": 1e-05,
      "loss": 0.5421,
      "step": 12600
    },
    {
      "epoch": 0.000127,
      "grad_norm": 2.8991341590881348,
      "learning_rate": 1e-05,
      "loss": 0.5379,
      "step": 12700
    },
    {
      "epoch": 0.000128,
      "grad_norm": 3.0275886058807373,
      "learning_rate": 1e-05,
      "loss": 0.5555,
      "step": 12800
    },
    {
      "epoch": 0.000129,
      "grad_norm": 2.8435568809509277,
      "learning_rate": 1e-05,
      "loss": 0.5475,
      "step": 12900
    },
    {
      "epoch": 0.00013,
      "grad_norm": 3.5080063343048096,
      "learning_rate": 1e-05,
      "loss": 0.5464,
      "step": 13000
    },
    {
      "epoch": 0.000131,
      "grad_norm": 3.4270200729370117,
      "learning_rate": 1e-05,
      "loss": 0.5489,
      "step": 13100
    },
    {
      "epoch": 0.000132,
      "grad_norm": 3.046891689300537,
      "learning_rate": 1e-05,
      "loss": 0.5427,
      "step": 13200
    },
    {
      "epoch": 0.000133,
      "grad_norm": 2.9758501052856445,
      "learning_rate": 1e-05,
      "loss": 0.547,
      "step": 13300
    },
    {
      "epoch": 0.000134,
      "grad_norm": 2.984278917312622,
      "learning_rate": 1e-05,
      "loss": 0.5405,
      "step": 13400
    },
    {
      "epoch": 0.000135,
      "grad_norm": 2.7465741634368896,
      "learning_rate": 1e-05,
      "loss": 0.5449,
      "step": 13500
    },
    {
      "epoch": 0.000136,
      "grad_norm": 2.6770498752593994,
      "learning_rate": 1e-05,
      "loss": 0.5313,
      "step": 13600
    },
    {
      "epoch": 0.000137,
      "grad_norm": 3.1820075511932373,
      "learning_rate": 1e-05,
      "loss": 0.5444,
      "step": 13700
    },
    {
      "epoch": 0.000138,
      "grad_norm": 2.7672953605651855,
      "learning_rate": 1e-05,
      "loss": 0.5366,
      "step": 13800
    },
    {
      "epoch": 0.000139,
      "grad_norm": 3.25970458984375,
      "learning_rate": 1e-05,
      "loss": 0.5331,
      "step": 13900
    },
    {
      "epoch": 0.00014,
      "grad_norm": 2.8809070587158203,
      "learning_rate": 1e-05,
      "loss": 0.5332,
      "step": 14000
    },
    {
      "epoch": 0.000141,
      "grad_norm": 2.7005808353424072,
      "learning_rate": 1e-05,
      "loss": 0.5335,
      "step": 14100
    },
    {
      "epoch": 0.000142,
      "grad_norm": 3.1988399028778076,
      "learning_rate": 1e-05,
      "loss": 0.533,
      "step": 14200
    },
    {
      "epoch": 0.000143,
      "grad_norm": 2.7792532444000244,
      "learning_rate": 1e-05,
      "loss": 0.5277,
      "step": 14300
    },
    {
      "epoch": 0.000144,
      "grad_norm": 3.010068655014038,
      "learning_rate": 1e-05,
      "loss": 0.5314,
      "step": 14400
    },
    {
      "epoch": 0.000145,
      "grad_norm": 3.3190596103668213,
      "learning_rate": 1e-05,
      "loss": 0.5308,
      "step": 14500
    },
    {
      "epoch": 0.000146,
      "grad_norm": 3.3294529914855957,
      "learning_rate": 1e-05,
      "loss": 0.5369,
      "step": 14600
    },
    {
      "epoch": 0.000147,
      "grad_norm": 3.0750784873962402,
      "learning_rate": 1e-05,
      "loss": 0.5399,
      "step": 14700
    },
    {
      "epoch": 0.000148,
      "grad_norm": 2.720137357711792,
      "learning_rate": 1e-05,
      "loss": 0.5255,
      "step": 14800
    },
    {
      "epoch": 0.000149,
      "grad_norm": 3.3225038051605225,
      "learning_rate": 1e-05,
      "loss": 0.5324,
      "step": 14900
    },
    {
      "epoch": 0.00015,
      "grad_norm": 2.890933036804199,
      "learning_rate": 1e-05,
      "loss": 0.5356,
      "step": 15000
    },
    {
      "epoch": 0.000151,
      "grad_norm": 3.118818998336792,
      "learning_rate": 1e-05,
      "loss": 0.5209,
      "step": 15100
    },
    {
      "epoch": 0.000152,
      "grad_norm": 2.549999237060547,
      "learning_rate": 1e-05,
      "loss": 0.5327,
      "step": 15200
    },
    {
      "epoch": 0.000153,
      "grad_norm": 2.8534693717956543,
      "learning_rate": 1e-05,
      "loss": 0.5161,
      "step": 15300
    },
    {
      "epoch": 0.000154,
      "grad_norm": 3.0132813453674316,
      "learning_rate": 1e-05,
      "loss": 0.5225,
      "step": 15400
    },
    {
      "epoch": 0.000155,
      "grad_norm": 2.934532880783081,
      "learning_rate": 1e-05,
      "loss": 0.5185,
      "step": 15500
    },
    {
      "epoch": 0.000156,
      "grad_norm": 3.079315423965454,
      "learning_rate": 1e-05,
      "loss": 0.5228,
      "step": 15600
    },
    {
      "epoch": 0.000157,
      "grad_norm": 2.9515817165374756,
      "learning_rate": 1e-05,
      "loss": 0.5183,
      "step": 15700
    },
    {
      "epoch": 0.000158,
      "grad_norm": 2.9063994884490967,
      "learning_rate": 1e-05,
      "loss": 0.5221,
      "step": 15800
    },
    {
      "epoch": 0.000159,
      "grad_norm": 2.873257875442505,
      "learning_rate": 1e-05,
      "loss": 0.5184,
      "step": 15900
    },
    {
      "epoch": 0.00016,
      "grad_norm": 3.2980196475982666,
      "learning_rate": 1e-05,
      "loss": 0.5208,
      "step": 16000
    },
    {
      "epoch": 0.000161,
      "grad_norm": 3.3925833702087402,
      "learning_rate": 1e-05,
      "loss": 0.5165,
      "step": 16100
    },
    {
      "epoch": 0.000162,
      "grad_norm": 3.8521616458892822,
      "learning_rate": 1e-05,
      "loss": 0.525,
      "step": 16200
    },
    {
      "epoch": 0.000163,
      "grad_norm": 2.9561917781829834,
      "learning_rate": 1e-05,
      "loss": 0.5266,
      "step": 16300
    },
    {
      "epoch": 0.000164,
      "grad_norm": 3.1145403385162354,
      "learning_rate": 1e-05,
      "loss": 0.51,
      "step": 16400
    },
    {
      "epoch": 0.000165,
      "grad_norm": 2.615156888961792,
      "learning_rate": 1e-05,
      "loss": 0.5253,
      "step": 16500
    },
    {
      "epoch": 0.000166,
      "grad_norm": 2.6065399646759033,
      "learning_rate": 1e-05,
      "loss": 0.5121,
      "step": 16600
    },
    {
      "epoch": 0.000167,
      "grad_norm": 3.8787238597869873,
      "learning_rate": 1e-05,
      "loss": 0.518,
      "step": 16700
    },
    {
      "epoch": 0.000168,
      "grad_norm": 2.8339273929595947,
      "learning_rate": 1e-05,
      "loss": 0.5098,
      "step": 16800
    },
    {
      "epoch": 0.000169,
      "grad_norm": 3.400110960006714,
      "learning_rate": 1e-05,
      "loss": 0.5139,
      "step": 16900
    },
    {
      "epoch": 0.00017,
      "grad_norm": 2.8647141456604004,
      "learning_rate": 1e-05,
      "loss": 0.51,
      "step": 17000
    },
    {
      "epoch": 0.000171,
      "grad_norm": 3.0346758365631104,
      "learning_rate": 1e-05,
      "loss": 0.52,
      "step": 17100
    },
    {
      "epoch": 0.000172,
      "grad_norm": 2.840468168258667,
      "learning_rate": 1e-05,
      "loss": 0.5126,
      "step": 17200
    },
    {
      "epoch": 0.000173,
      "grad_norm": 2.32464861869812,
      "learning_rate": 1e-05,
      "loss": 0.5141,
      "step": 17300
    },
    {
      "epoch": 0.000174,
      "grad_norm": 2.6493277549743652,
      "learning_rate": 1e-05,
      "loss": 0.5061,
      "step": 17400
    },
    {
      "epoch": 0.000175,
      "grad_norm": 2.9182634353637695,
      "learning_rate": 1e-05,
      "loss": 0.5079,
      "step": 17500
    },
    {
      "epoch": 0.000176,
      "grad_norm": 2.817209243774414,
      "learning_rate": 1e-05,
      "loss": 0.5209,
      "step": 17600
    },
    {
      "epoch": 0.000177,
      "grad_norm": 3.2138490676879883,
      "learning_rate": 1e-05,
      "loss": 0.5008,
      "step": 17700
    },
    {
      "epoch": 0.000178,
      "grad_norm": 2.925663471221924,
      "learning_rate": 1e-05,
      "loss": 0.5045,
      "step": 17800
    },
    {
      "epoch": 0.000179,
      "grad_norm": 3.3058528900146484,
      "learning_rate": 1e-05,
      "loss": 0.5036,
      "step": 17900
    },
    {
      "epoch": 0.00018,
      "grad_norm": 2.73296856880188,
      "learning_rate": 1e-05,
      "loss": 0.5023,
      "step": 18000
    },
    {
      "epoch": 0.000181,
      "grad_norm": 3.112847089767456,
      "learning_rate": 1e-05,
      "loss": 0.5178,
      "step": 18100
    },
    {
      "epoch": 0.000182,
      "grad_norm": 2.471736431121826,
      "learning_rate": 1e-05,
      "loss": 0.5082,
      "step": 18200
    },
    {
      "epoch": 0.000183,
      "grad_norm": 2.6374621391296387,
      "learning_rate": 1e-05,
      "loss": 0.5069,
      "step": 18300
    },
    {
      "epoch": 0.000184,
      "grad_norm": 2.838254690170288,
      "learning_rate": 1e-05,
      "loss": 0.5101,
      "step": 18400
    },
    {
      "epoch": 0.000185,
      "grad_norm": 2.9953255653381348,
      "learning_rate": 1e-05,
      "loss": 0.5076,
      "step": 18500
    },
    {
      "epoch": 0.000186,
      "grad_norm": 2.7611818313598633,
      "learning_rate": 1e-05,
      "loss": 0.4942,
      "step": 18600
    },
    {
      "epoch": 0.000187,
      "grad_norm": 2.739598274230957,
      "learning_rate": 1e-05,
      "loss": 0.508,
      "step": 18700
    },
    {
      "epoch": 0.000188,
      "grad_norm": 2.6710898876190186,
      "learning_rate": 1e-05,
      "loss": 0.5125,
      "step": 18800
    },
    {
      "epoch": 0.000189,
      "grad_norm": 2.7952322959899902,
      "learning_rate": 1e-05,
      "loss": 0.4844,
      "step": 18900
    },
    {
      "epoch": 0.00019,
      "grad_norm": 2.4689576625823975,
      "learning_rate": 1e-05,
      "loss": 0.5072,
      "step": 19000
    },
    {
      "epoch": 0.000191,
      "grad_norm": 2.6042397022247314,
      "learning_rate": 1e-05,
      "loss": 0.491,
      "step": 19100
    },
    {
      "epoch": 0.000192,
      "grad_norm": 2.915821075439453,
      "learning_rate": 1e-05,
      "loss": 0.5029,
      "step": 19200
    },
    {
      "epoch": 0.000193,
      "grad_norm": 3.0724246501922607,
      "learning_rate": 1e-05,
      "loss": 0.5085,
      "step": 19300
    },
    {
      "epoch": 0.000194,
      "grad_norm": 2.755842924118042,
      "learning_rate": 1e-05,
      "loss": 0.5073,
      "step": 19400
    },
    {
      "epoch": 0.000195,
      "grad_norm": 2.6537370681762695,
      "learning_rate": 1e-05,
      "loss": 0.4953,
      "step": 19500
    },
    {
      "epoch": 0.000196,
      "grad_norm": 2.8526852130889893,
      "learning_rate": 1e-05,
      "loss": 0.4943,
      "step": 19600
    },
    {
      "epoch": 0.000197,
      "grad_norm": 2.4322919845581055,
      "learning_rate": 1e-05,
      "loss": 0.4985,
      "step": 19700
    },
    {
      "epoch": 0.000198,
      "grad_norm": 3.18966007232666,
      "learning_rate": 1e-05,
      "loss": 0.4946,
      "step": 19800
    },
    {
      "epoch": 0.000199,
      "grad_norm": 2.5495989322662354,
      "learning_rate": 1e-05,
      "loss": 0.4962,
      "step": 19900
    },
    {
      "epoch": 0.0002,
      "grad_norm": 3.0627942085266113,
      "learning_rate": 1e-05,
      "loss": 0.4876,
      "step": 20000
    },
    {
      "epoch": 0.0002,
      "eval_loss": 0.46728515625,
      "eval_runtime": 111.4705,
      "eval_samples_per_second": 448.549,
      "eval_steps_per_second": 28.034,
      "step": 20000
    },
    {
      "epoch": 0.000201,
      "grad_norm": 2.799205780029297,
      "learning_rate": 1e-05,
      "loss": 0.5117,
      "step": 20100
    },
    {
      "epoch": 0.000202,
      "grad_norm": 2.509645700454712,
      "learning_rate": 1e-05,
      "loss": 0.4903,
      "step": 20200
    },
    {
      "epoch": 0.000203,
      "grad_norm": 3.120729684829712,
      "learning_rate": 1e-05,
      "loss": 0.5027,
      "step": 20300
    },
    {
      "epoch": 0.000204,
      "grad_norm": 2.390143394470215,
      "learning_rate": 1e-05,
      "loss": 0.4989,
      "step": 20400
    },
    {
      "epoch": 0.000205,
      "grad_norm": 2.5487399101257324,
      "learning_rate": 1e-05,
      "loss": 0.4945,
      "step": 20500
    },
    {
      "epoch": 0.000206,
      "grad_norm": 2.9931600093841553,
      "learning_rate": 1e-05,
      "loss": 0.5048,
      "step": 20600
    },
    {
      "epoch": 0.000207,
      "grad_norm": 2.5760886669158936,
      "learning_rate": 1e-05,
      "loss": 0.4923,
      "step": 20700
    },
    {
      "epoch": 0.000208,
      "grad_norm": 3.149047613143921,
      "learning_rate": 1e-05,
      "loss": 0.4937,
      "step": 20800
    },
    {
      "epoch": 0.000209,
      "grad_norm": 2.982687473297119,
      "learning_rate": 1e-05,
      "loss": 0.4901,
      "step": 20900
    },
    {
      "epoch": 0.00021,
      "grad_norm": 2.6681571006774902,
      "learning_rate": 1e-05,
      "loss": 0.4938,
      "step": 21000
    },
    {
      "epoch": 0.000211,
      "grad_norm": 2.9358367919921875,
      "learning_rate": 1e-05,
      "loss": 0.4962,
      "step": 21100
    },
    {
      "epoch": 0.000212,
      "grad_norm": 2.8525304794311523,
      "learning_rate": 1e-05,
      "loss": 0.4998,
      "step": 21200
    },
    {
      "epoch": 0.000213,
      "grad_norm": 2.7039895057678223,
      "learning_rate": 1e-05,
      "loss": 0.4876,
      "step": 21300
    },
    {
      "epoch": 0.000214,
      "grad_norm": 2.591728448867798,
      "learning_rate": 1e-05,
      "loss": 0.4895,
      "step": 21400
    },
    {
      "epoch": 0.000215,
      "grad_norm": 4.136421203613281,
      "learning_rate": 1e-05,
      "loss": 0.4892,
      "step": 21500
    },
    {
      "epoch": 0.000216,
      "grad_norm": 2.715740919113159,
      "learning_rate": 1e-05,
      "loss": 0.4966,
      "step": 21600
    },
    {
      "epoch": 0.000217,
      "grad_norm": 2.804382562637329,
      "learning_rate": 1e-05,
      "loss": 0.5002,
      "step": 21700
    },
    {
      "epoch": 0.000218,
      "grad_norm": 2.568103551864624,
      "learning_rate": 1e-05,
      "loss": 0.4925,
      "step": 21800
    },
    {
      "epoch": 0.000219,
      "grad_norm": 2.671935796737671,
      "learning_rate": 1e-05,
      "loss": 0.4855,
      "step": 21900
    },
    {
      "epoch": 0.00022,
      "grad_norm": 3.6780941486358643,
      "learning_rate": 1e-05,
      "loss": 0.4778,
      "step": 22000
    },
    {
      "epoch": 0.000221,
      "grad_norm": 2.6798088550567627,
      "learning_rate": 1e-05,
      "loss": 0.4893,
      "step": 22100
    },
    {
      "epoch": 0.000222,
      "grad_norm": 2.930389642715454,
      "learning_rate": 1e-05,
      "loss": 0.4904,
      "step": 22200
    },
    {
      "epoch": 0.000223,
      "grad_norm": 2.2273404598236084,
      "learning_rate": 1e-05,
      "loss": 0.4864,
      "step": 22300
    },
    {
      "epoch": 0.000224,
      "grad_norm": 2.7305030822753906,
      "learning_rate": 1e-05,
      "loss": 0.4962,
      "step": 22400
    },
    {
      "epoch": 0.000225,
      "grad_norm": 2.6182594299316406,
      "learning_rate": 1e-05,
      "loss": 0.4843,
      "step": 22500
    },
    {
      "epoch": 0.000226,
      "grad_norm": 2.7195889949798584,
      "learning_rate": 1e-05,
      "loss": 0.4848,
      "step": 22600
    },
    {
      "epoch": 0.000227,
      "grad_norm": 2.842867136001587,
      "learning_rate": 1e-05,
      "loss": 0.4868,
      "step": 22700
    },
    {
      "epoch": 0.000228,
      "grad_norm": 2.6133205890655518,
      "learning_rate": 1e-05,
      "loss": 0.4854,
      "step": 22800
    },
    {
      "epoch": 0.000229,
      "grad_norm": 2.9739954471588135,
      "learning_rate": 1e-05,
      "loss": 0.4799,
      "step": 22900
    },
    {
      "epoch": 0.00023,
      "grad_norm": 2.704303503036499,
      "learning_rate": 1e-05,
      "loss": 0.4787,
      "step": 23000
    },
    {
      "epoch": 0.000231,
      "grad_norm": 2.6430766582489014,
      "learning_rate": 1e-05,
      "loss": 0.4838,
      "step": 23100
    },
    {
      "epoch": 0.000232,
      "grad_norm": 2.57578444480896,
      "learning_rate": 1e-05,
      "loss": 0.4818,
      "step": 23200
    },
    {
      "epoch": 0.000233,
      "grad_norm": 2.553027868270874,
      "learning_rate": 1e-05,
      "loss": 0.4878,
      "step": 23300
    },
    {
      "epoch": 0.000234,
      "grad_norm": 2.853264331817627,
      "learning_rate": 1e-05,
      "loss": 0.4841,
      "step": 23400
    },
    {
      "epoch": 0.000235,
      "grad_norm": 2.591419219970703,
      "learning_rate": 1e-05,
      "loss": 0.4818,
      "step": 23500
    },
    {
      "epoch": 0.000236,
      "grad_norm": 2.8946421146392822,
      "learning_rate": 1e-05,
      "loss": 0.4808,
      "step": 23600
    },
    {
      "epoch": 0.000237,
      "grad_norm": 2.9158196449279785,
      "learning_rate": 1e-05,
      "loss": 0.4755,
      "step": 23700
    },
    {
      "epoch": 0.000238,
      "grad_norm": 2.578831195831299,
      "learning_rate": 1e-05,
      "loss": 0.4759,
      "step": 23800
    },
    {
      "epoch": 0.000239,
      "grad_norm": 2.6290273666381836,
      "learning_rate": 1e-05,
      "loss": 0.4858,
      "step": 23900
    },
    {
      "epoch": 0.00024,
      "grad_norm": 2.525026321411133,
      "learning_rate": 1e-05,
      "loss": 0.4913,
      "step": 24000
    },
    {
      "epoch": 0.000241,
      "grad_norm": 4.174901008605957,
      "learning_rate": 1e-05,
      "loss": 0.474,
      "step": 24100
    },
    {
      "epoch": 0.000242,
      "grad_norm": 2.6417720317840576,
      "learning_rate": 1e-05,
      "loss": 0.4831,
      "step": 24200
    },
    {
      "epoch": 0.000243,
      "grad_norm": 2.4943110942840576,
      "learning_rate": 1e-05,
      "loss": 0.4763,
      "step": 24300
    },
    {
      "epoch": 0.000244,
      "grad_norm": 2.609255075454712,
      "learning_rate": 1e-05,
      "loss": 0.4868,
      "step": 24400
    },
    {
      "epoch": 0.000245,
      "grad_norm": 3.1849722862243652,
      "learning_rate": 1e-05,
      "loss": 0.48,
      "step": 24500
    },
    {
      "epoch": 0.000246,
      "grad_norm": 2.6401076316833496,
      "learning_rate": 1e-05,
      "loss": 0.4873,
      "step": 24600
    },
    {
      "epoch": 0.000247,
      "grad_norm": 2.936086416244507,
      "learning_rate": 1e-05,
      "loss": 0.4769,
      "step": 24700
    },
    {
      "epoch": 0.000248,
      "grad_norm": 2.6310338973999023,
      "learning_rate": 1e-05,
      "loss": 0.4794,
      "step": 24800
    },
    {
      "epoch": 0.000249,
      "grad_norm": 2.3982949256896973,
      "learning_rate": 1e-05,
      "loss": 0.4818,
      "step": 24900
    },
    {
      "epoch": 0.00025,
      "grad_norm": 2.8052144050598145,
      "learning_rate": 1e-05,
      "loss": 0.4674,
      "step": 25000
    },
    {
      "epoch": 0.000251,
      "grad_norm": 3.136012315750122,
      "learning_rate": 1e-05,
      "loss": 0.4739,
      "step": 25100
    },
    {
      "epoch": 0.000252,
      "grad_norm": 2.721803665161133,
      "learning_rate": 1e-05,
      "loss": 0.4653,
      "step": 25200
    },
    {
      "epoch": 0.000253,
      "grad_norm": 2.9835872650146484,
      "learning_rate": 1e-05,
      "loss": 0.4752,
      "step": 25300
    },
    {
      "epoch": 0.000254,
      "grad_norm": 2.4551830291748047,
      "learning_rate": 1e-05,
      "loss": 0.4732,
      "step": 25400
    },
    {
      "epoch": 0.000255,
      "grad_norm": 3.3453078269958496,
      "learning_rate": 1e-05,
      "loss": 0.4766,
      "step": 25500
    },
    {
      "epoch": 0.000256,
      "grad_norm": 2.7177910804748535,
      "learning_rate": 1e-05,
      "loss": 0.4761,
      "step": 25600
    },
    {
      "epoch": 0.000257,
      "grad_norm": 2.9529471397399902,
      "learning_rate": 1e-05,
      "loss": 0.4816,
      "step": 25700
    },
    {
      "epoch": 0.000258,
      "grad_norm": 2.5339162349700928,
      "learning_rate": 1e-05,
      "loss": 0.4719,
      "step": 25800
    },
    {
      "epoch": 0.000259,
      "grad_norm": 2.5781121253967285,
      "learning_rate": 1e-05,
      "loss": 0.4742,
      "step": 25900
    },
    {
      "epoch": 0.00026,
      "grad_norm": 2.7583415508270264,
      "learning_rate": 1e-05,
      "loss": 0.478,
      "step": 26000
    },
    {
      "epoch": 0.000261,
      "grad_norm": 2.307037591934204,
      "learning_rate": 1e-05,
      "loss": 0.4767,
      "step": 26100
    },
    {
      "epoch": 0.000262,
      "grad_norm": 2.9219844341278076,
      "learning_rate": 1e-05,
      "loss": 0.4783,
      "step": 26200
    },
    {
      "epoch": 0.000263,
      "grad_norm": 2.4626011848449707,
      "learning_rate": 1e-05,
      "loss": 0.4806,
      "step": 26300
    },
    {
      "epoch": 0.000264,
      "grad_norm": 2.7708845138549805,
      "learning_rate": 1e-05,
      "loss": 0.4717,
      "step": 26400
    },
    {
      "epoch": 0.000265,
      "grad_norm": 2.806086301803589,
      "learning_rate": 1e-05,
      "loss": 0.4641,
      "step": 26500
    },
    {
      "epoch": 0.000266,
      "grad_norm": 2.395228385925293,
      "learning_rate": 1e-05,
      "loss": 0.4769,
      "step": 26600
    },
    {
      "epoch": 0.000267,
      "grad_norm": 2.482196807861328,
      "learning_rate": 1e-05,
      "loss": 0.4637,
      "step": 26700
    },
    {
      "epoch": 0.000268,
      "grad_norm": 2.3239686489105225,
      "learning_rate": 1e-05,
      "loss": 0.4728,
      "step": 26800
    },
    {
      "epoch": 0.000269,
      "grad_norm": 2.8760108947753906,
      "learning_rate": 1e-05,
      "loss": 0.4683,
      "step": 26900
    },
    {
      "epoch": 0.00027,
      "grad_norm": 3.0095269680023193,
      "learning_rate": 1e-05,
      "loss": 0.4641,
      "step": 27000
    },
    {
      "epoch": 0.000271,
      "grad_norm": 2.7603840827941895,
      "learning_rate": 1e-05,
      "loss": 0.4824,
      "step": 27100
    },
    {
      "epoch": 0.000272,
      "grad_norm": 2.6380269527435303,
      "learning_rate": 1e-05,
      "loss": 0.4626,
      "step": 27200
    },
    {
      "epoch": 0.000273,
      "grad_norm": 3.095323324203491,
      "learning_rate": 1e-05,
      "loss": 0.4671,
      "step": 27300
    },
    {
      "epoch": 0.000274,
      "grad_norm": 2.6990623474121094,
      "learning_rate": 1e-05,
      "loss": 0.4766,
      "step": 27400
    },
    {
      "epoch": 0.000275,
      "grad_norm": 2.502337694168091,
      "learning_rate": 1e-05,
      "loss": 0.4608,
      "step": 27500
    },
    {
      "epoch": 0.000276,
      "grad_norm": 2.3775081634521484,
      "learning_rate": 1e-05,
      "loss": 0.4749,
      "step": 27600
    },
    {
      "epoch": 0.000277,
      "grad_norm": 2.7666544914245605,
      "learning_rate": 1e-05,
      "loss": 0.465,
      "step": 27700
    },
    {
      "epoch": 0.000278,
      "grad_norm": 2.4668657779693604,
      "learning_rate": 1e-05,
      "loss": 0.4605,
      "step": 27800
    },
    {
      "epoch": 0.000279,
      "grad_norm": 3.91645884513855,
      "learning_rate": 1e-05,
      "loss": 0.4684,
      "step": 27900
    },
    {
      "epoch": 0.00028,
      "grad_norm": 2.781068801879883,
      "learning_rate": 1e-05,
      "loss": 0.4612,
      "step": 28000
    },
    {
      "epoch": 0.000281,
      "grad_norm": 2.4313833713531494,
      "learning_rate": 1e-05,
      "loss": 0.4661,
      "step": 28100
    },
    {
      "epoch": 0.000282,
      "grad_norm": 2.236158847808838,
      "learning_rate": 1e-05,
      "loss": 0.4728,
      "step": 28200
    },
    {
      "epoch": 0.000283,
      "grad_norm": 2.7676749229431152,
      "learning_rate": 1e-05,
      "loss": 0.4667,
      "step": 28300
    },
    {
      "epoch": 0.000284,
      "grad_norm": 2.1664578914642334,
      "learning_rate": 1e-05,
      "loss": 0.4632,
      "step": 28400
    },
    {
      "epoch": 0.000285,
      "grad_norm": 2.4924814701080322,
      "learning_rate": 1e-05,
      "loss": 0.4699,
      "step": 28500
    },
    {
      "epoch": 0.000286,
      "grad_norm": 3.364351749420166,
      "learning_rate": 1e-05,
      "loss": 0.4563,
      "step": 28600
    },
    {
      "epoch": 0.000287,
      "grad_norm": 2.29882550239563,
      "learning_rate": 1e-05,
      "loss": 0.4689,
      "step": 28700
    },
    {
      "epoch": 0.000288,
      "grad_norm": 2.626985549926758,
      "learning_rate": 1e-05,
      "loss": 0.463,
      "step": 28800
    },
    {
      "epoch": 0.000289,
      "grad_norm": 2.7008321285247803,
      "learning_rate": 1e-05,
      "loss": 0.4601,
      "step": 28900
    },
    {
      "epoch": 0.00029,
      "grad_norm": 2.5816690921783447,
      "learning_rate": 1e-05,
      "loss": 0.4597,
      "step": 29000
    },
    {
      "epoch": 0.000291,
      "grad_norm": 2.4688684940338135,
      "learning_rate": 1e-05,
      "loss": 0.4599,
      "step": 29100
    },
    {
      "epoch": 0.000292,
      "grad_norm": 2.839632749557495,
      "learning_rate": 1e-05,
      "loss": 0.4677,
      "step": 29200
    },
    {
      "epoch": 0.000293,
      "grad_norm": 3.800483465194702,
      "learning_rate": 1e-05,
      "loss": 0.4565,
      "step": 29300
    },
    {
      "epoch": 0.000294,
      "grad_norm": 4.7663726806640625,
      "learning_rate": 1e-05,
      "loss": 0.4522,
      "step": 29400
    },
    {
      "epoch": 0.000295,
      "grad_norm": 2.4576992988586426,
      "learning_rate": 1e-05,
      "loss": 0.4638,
      "step": 29500
    },
    {
      "epoch": 0.000296,
      "grad_norm": 2.6215529441833496,
      "learning_rate": 1e-05,
      "loss": 0.4522,
      "step": 29600
    },
    {
      "epoch": 0.000297,
      "grad_norm": 2.35202693939209,
      "learning_rate": 1e-05,
      "loss": 0.452,
      "step": 29700
    },
    {
      "epoch": 0.000298,
      "grad_norm": 2.1658172607421875,
      "learning_rate": 1e-05,
      "loss": 0.4638,
      "step": 29800
    },
    {
      "epoch": 0.000299,
      "grad_norm": 2.6954879760742188,
      "learning_rate": 1e-05,
      "loss": 0.4533,
      "step": 29900
    },
    {
      "epoch": 0.0003,
      "grad_norm": 2.465700626373291,
      "learning_rate": 1e-05,
      "loss": 0.4595,
      "step": 30000
    },
    {
      "epoch": 0.000301,
      "grad_norm": 2.5099382400512695,
      "learning_rate": 1e-05,
      "loss": 0.464,
      "step": 30100
    },
    {
      "epoch": 0.000302,
      "grad_norm": 2.632709264755249,
      "learning_rate": 1e-05,
      "loss": 0.4583,
      "step": 30200
    },
    {
      "epoch": 0.000303,
      "grad_norm": 3.451253890991211,
      "learning_rate": 1e-05,
      "loss": 0.4641,
      "step": 30300
    },
    {
      "epoch": 0.000304,
      "grad_norm": 2.5576229095458984,
      "learning_rate": 1e-05,
      "loss": 0.4643,
      "step": 30400
    },
    {
      "epoch": 0.000305,
      "grad_norm": 2.63854718208313,
      "learning_rate": 1e-05,
      "loss": 0.4509,
      "step": 30500
    },
    {
      "epoch": 0.000306,
      "grad_norm": 2.7660105228424072,
      "learning_rate": 1e-05,
      "loss": 0.4533,
      "step": 30600
    },
    {
      "epoch": 0.000307,
      "grad_norm": 2.862382411956787,
      "learning_rate": 1e-05,
      "loss": 0.4536,
      "step": 30700
    },
    {
      "epoch": 0.000308,
      "grad_norm": 2.6443052291870117,
      "learning_rate": 1e-05,
      "loss": 0.4485,
      "step": 30800
    },
    {
      "epoch": 0.000309,
      "grad_norm": 2.525301456451416,
      "learning_rate": 1e-05,
      "loss": 0.4593,
      "step": 30900
    },
    {
      "epoch": 0.00031,
      "grad_norm": 2.4334826469421387,
      "learning_rate": 1e-05,
      "loss": 0.4495,
      "step": 31000
    },
    {
      "epoch": 0.000311,
      "grad_norm": 2.443531036376953,
      "learning_rate": 1e-05,
      "loss": 0.4563,
      "step": 31100
    },
    {
      "epoch": 0.000312,
      "grad_norm": 2.3432857990264893,
      "learning_rate": 1e-05,
      "loss": 0.4587,
      "step": 31200
    },
    {
      "epoch": 0.000313,
      "grad_norm": 2.470900058746338,
      "learning_rate": 1e-05,
      "loss": 0.4466,
      "step": 31300
    },
    {
      "epoch": 0.000314,
      "grad_norm": 2.719302177429199,
      "learning_rate": 1e-05,
      "loss": 0.4633,
      "step": 31400
    },
    {
      "epoch": 0.000315,
      "grad_norm": 2.7963156700134277,
      "learning_rate": 1e-05,
      "loss": 0.4536,
      "step": 31500
    },
    {
      "epoch": 0.000316,
      "grad_norm": 2.750457763671875,
      "learning_rate": 1e-05,
      "loss": 0.4515,
      "step": 31600
    },
    {
      "epoch": 0.000317,
      "grad_norm": 2.4325835704803467,
      "learning_rate": 1e-05,
      "loss": 0.4555,
      "step": 31700
    },
    {
      "epoch": 0.000318,
      "grad_norm": 2.4605915546417236,
      "learning_rate": 1e-05,
      "loss": 0.4571,
      "step": 31800
    },
    {
      "epoch": 0.000319,
      "grad_norm": 2.3610308170318604,
      "learning_rate": 1e-05,
      "loss": 0.455,
      "step": 31900
    },
    {
      "epoch": 0.00032,
      "grad_norm": 2.4374446868896484,
      "learning_rate": 1e-05,
      "loss": 0.4577,
      "step": 32000
    },
    {
      "epoch": 0.000321,
      "grad_norm": 2.618852138519287,
      "learning_rate": 1e-05,
      "loss": 0.4434,
      "step": 32100
    },
    {
      "epoch": 0.000322,
      "grad_norm": 3.4312095642089844,
      "learning_rate": 1e-05,
      "loss": 0.448,
      "step": 32200
    },
    {
      "epoch": 0.000323,
      "grad_norm": 3.933258056640625,
      "learning_rate": 1e-05,
      "loss": 0.4542,
      "step": 32300
    },
    {
      "epoch": 0.000324,
      "grad_norm": 2.4201653003692627,
      "learning_rate": 1e-05,
      "loss": 0.4501,
      "step": 32400
    },
    {
      "epoch": 0.000325,
      "grad_norm": 2.638230085372925,
      "learning_rate": 1e-05,
      "loss": 0.4587,
      "step": 32500
    },
    {
      "epoch": 0.000326,
      "grad_norm": 2.310612201690674,
      "learning_rate": 1e-05,
      "loss": 0.4647,
      "step": 32600
    },
    {
      "epoch": 0.000327,
      "grad_norm": 2.466186285018921,
      "learning_rate": 1e-05,
      "loss": 0.4462,
      "step": 32700
    },
    {
      "epoch": 0.000328,
      "grad_norm": 2.8925211429595947,
      "learning_rate": 1e-05,
      "loss": 0.4556,
      "step": 32800
    },
    {
      "epoch": 0.000329,
      "grad_norm": 2.5643179416656494,
      "learning_rate": 1e-05,
      "loss": 0.4479,
      "step": 32900
    },
    {
      "epoch": 0.00033,
      "grad_norm": 2.372391939163208,
      "learning_rate": 1e-05,
      "loss": 0.4583,
      "step": 33000
    },
    {
      "epoch": 0.000331,
      "grad_norm": 2.6954376697540283,
      "learning_rate": 1e-05,
      "loss": 0.4514,
      "step": 33100
    },
    {
      "epoch": 0.000332,
      "grad_norm": 2.380615234375,
      "learning_rate": 1e-05,
      "loss": 0.4468,
      "step": 33200
    },
    {
      "epoch": 0.000333,
      "grad_norm": 2.5895583629608154,
      "learning_rate": 1e-05,
      "loss": 0.4503,
      "step": 33300
    },
    {
      "epoch": 0.000334,
      "grad_norm": 3.745288133621216,
      "learning_rate": 1e-05,
      "loss": 0.4453,
      "step": 33400
    },
    {
      "epoch": 0.000335,
      "grad_norm": 2.346338987350464,
      "learning_rate": 1e-05,
      "loss": 0.4605,
      "step": 33500
    },
    {
      "epoch": 0.000336,
      "grad_norm": 3.660322666168213,
      "learning_rate": 1e-05,
      "loss": 0.4518,
      "step": 33600
    },
    {
      "epoch": 0.000337,
      "grad_norm": 2.2425618171691895,
      "learning_rate": 1e-05,
      "loss": 0.4433,
      "step": 33700
    },
    {
      "epoch": 0.000338,
      "grad_norm": 2.385923385620117,
      "learning_rate": 1e-05,
      "loss": 0.4452,
      "step": 33800
    },
    {
      "epoch": 0.000339,
      "grad_norm": 2.3236701488494873,
      "learning_rate": 1e-05,
      "loss": 0.4428,
      "step": 33900
    },
    {
      "epoch": 0.00034,
      "grad_norm": 2.2188169956207275,
      "learning_rate": 1e-05,
      "loss": 0.4536,
      "step": 34000
    },
    {
      "epoch": 0.000341,
      "grad_norm": 2.760098934173584,
      "learning_rate": 1e-05,
      "loss": 0.4487,
      "step": 34100
    },
    {
      "epoch": 0.000342,
      "grad_norm": 2.2766711711883545,
      "learning_rate": 1e-05,
      "loss": 0.4545,
      "step": 34200
    },
    {
      "epoch": 0.000343,
      "grad_norm": 2.6107327938079834,
      "learning_rate": 1e-05,
      "loss": 0.4503,
      "step": 34300
    },
    {
      "epoch": 0.000344,
      "grad_norm": 2.5845329761505127,
      "learning_rate": 1e-05,
      "loss": 0.4503,
      "step": 34400
    },
    {
      "epoch": 0.000345,
      "grad_norm": 2.1708152294158936,
      "learning_rate": 1e-05,
      "loss": 0.4474,
      "step": 34500
    },
    {
      "epoch": 0.000346,
      "grad_norm": 2.408508062362671,
      "learning_rate": 1e-05,
      "loss": 0.4504,
      "step": 34600
    },
    {
      "epoch": 0.000347,
      "grad_norm": 2.5341970920562744,
      "learning_rate": 1e-05,
      "loss": 0.4531,
      "step": 34700
    },
    {
      "epoch": 0.000348,
      "grad_norm": 2.5573482513427734,
      "learning_rate": 1e-05,
      "loss": 0.4472,
      "step": 34800
    },
    {
      "epoch": 0.000349,
      "grad_norm": 2.318730115890503,
      "learning_rate": 1e-05,
      "loss": 0.4462,
      "step": 34900
    },
    {
      "epoch": 0.00035,
      "grad_norm": 2.341620922088623,
      "learning_rate": 1e-05,
      "loss": 0.4423,
      "step": 35000
    },
    {
      "epoch": 0.000351,
      "grad_norm": 2.1792995929718018,
      "learning_rate": 1e-05,
      "loss": 0.4519,
      "step": 35100
    },
    {
      "epoch": 0.000352,
      "grad_norm": 2.4928019046783447,
      "learning_rate": 1e-05,
      "loss": 0.4519,
      "step": 35200
    },
    {
      "epoch": 0.000353,
      "grad_norm": 2.512012481689453,
      "learning_rate": 1e-05,
      "loss": 0.4578,
      "step": 35300
    },
    {
      "epoch": 0.000354,
      "grad_norm": 2.510221242904663,
      "learning_rate": 1e-05,
      "loss": 0.4506,
      "step": 35400
    },
    {
      "epoch": 0.000355,
      "grad_norm": 2.637925624847412,
      "learning_rate": 1e-05,
      "loss": 0.4454,
      "step": 35500
    },
    {
      "epoch": 0.000356,
      "grad_norm": 2.6336724758148193,
      "learning_rate": 1e-05,
      "loss": 0.4477,
      "step": 35600
    },
    {
      "epoch": 0.000357,
      "grad_norm": 2.6396801471710205,
      "learning_rate": 1e-05,
      "loss": 0.4424,
      "step": 35700
    },
    {
      "epoch": 0.000358,
      "grad_norm": 2.2982890605926514,
      "learning_rate": 1e-05,
      "loss": 0.4434,
      "step": 35800
    },
    {
      "epoch": 0.000359,
      "grad_norm": 2.617039680480957,
      "learning_rate": 1e-05,
      "loss": 0.443,
      "step": 35900
    },
    {
      "epoch": 0.00036,
      "grad_norm": 2.3763229846954346,
      "learning_rate": 1e-05,
      "loss": 0.4462,
      "step": 36000
    },
    {
      "epoch": 0.000361,
      "grad_norm": 2.4981770515441895,
      "learning_rate": 1e-05,
      "loss": 0.445,
      "step": 36100
    },
    {
      "epoch": 0.000362,
      "grad_norm": 2.2644827365875244,
      "learning_rate": 1e-05,
      "loss": 0.4423,
      "step": 36200
    },
    {
      "epoch": 0.000363,
      "grad_norm": 2.1267762184143066,
      "learning_rate": 1e-05,
      "loss": 0.4391,
      "step": 36300
    },
    {
      "epoch": 0.000364,
      "grad_norm": 2.2303924560546875,
      "learning_rate": 1e-05,
      "loss": 0.445,
      "step": 36400
    },
    {
      "epoch": 0.000365,
      "grad_norm": 2.383427619934082,
      "learning_rate": 1e-05,
      "loss": 0.4391,
      "step": 36500
    },
    {
      "epoch": 0.000366,
      "grad_norm": 2.600222587585449,
      "learning_rate": 1e-05,
      "loss": 0.446,
      "step": 36600
    },
    {
      "epoch": 0.000367,
      "grad_norm": 2.557803153991699,
      "learning_rate": 1e-05,
      "loss": 0.4519,
      "step": 36700
    },
    {
      "epoch": 0.000368,
      "grad_norm": 2.5691545009613037,
      "learning_rate": 1e-05,
      "loss": 0.4385,
      "step": 36800
    },
    {
      "epoch": 0.000369,
      "grad_norm": 2.3497157096862793,
      "learning_rate": 1e-05,
      "loss": 0.4383,
      "step": 36900
    },
    {
      "epoch": 0.00037,
      "grad_norm": 2.87781023979187,
      "learning_rate": 1e-05,
      "loss": 0.4409,
      "step": 37000
    },
    {
      "epoch": 0.000371,
      "grad_norm": 2.447251558303833,
      "learning_rate": 1e-05,
      "loss": 0.4485,
      "step": 37100
    },
    {
      "epoch": 0.000372,
      "grad_norm": 2.536022424697876,
      "learning_rate": 1e-05,
      "loss": 0.4406,
      "step": 37200
    },
    {
      "epoch": 0.000373,
      "grad_norm": 2.4963696002960205,
      "learning_rate": 1e-05,
      "loss": 0.4311,
      "step": 37300
    },
    {
      "epoch": 0.000374,
      "grad_norm": 2.701169967651367,
      "learning_rate": 1e-05,
      "loss": 0.4385,
      "step": 37400
    },
    {
      "epoch": 0.000375,
      "grad_norm": 2.514227867126465,
      "learning_rate": 1e-05,
      "loss": 0.4357,
      "step": 37500
    },
    {
      "epoch": 0.000376,
      "grad_norm": 2.5205540657043457,
      "learning_rate": 1e-05,
      "loss": 0.4447,
      "step": 37600
    },
    {
      "epoch": 0.000377,
      "grad_norm": 2.7338292598724365,
      "learning_rate": 1e-05,
      "loss": 0.4452,
      "step": 37700
    },
    {
      "epoch": 0.000378,
      "grad_norm": 2.3861348628997803,
      "learning_rate": 1e-05,
      "loss": 0.4377,
      "step": 37800
    },
    {
      "epoch": 0.000379,
      "grad_norm": 2.41011381149292,
      "learning_rate": 1e-05,
      "loss": 0.4405,
      "step": 37900
    },
    {
      "epoch": 0.00038,
      "grad_norm": 2.449092388153076,
      "learning_rate": 1e-05,
      "loss": 0.4451,
      "step": 38000
    },
    {
      "epoch": 0.000381,
      "grad_norm": 2.571415662765503,
      "learning_rate": 1e-05,
      "loss": 0.439,
      "step": 38100
    },
    {
      "epoch": 0.000382,
      "grad_norm": 2.169980764389038,
      "learning_rate": 1e-05,
      "loss": 0.4384,
      "step": 38200
    },
    {
      "epoch": 0.000383,
      "grad_norm": 3.838111162185669,
      "learning_rate": 1e-05,
      "loss": 0.4322,
      "step": 38300
    },
    {
      "epoch": 0.000384,
      "grad_norm": 3.1614363193511963,
      "learning_rate": 1e-05,
      "loss": 0.4426,
      "step": 38400
    },
    {
      "epoch": 0.000385,
      "grad_norm": 1.975185513496399,
      "learning_rate": 1e-05,
      "loss": 0.4375,
      "step": 38500
    },
    {
      "epoch": 0.000386,
      "grad_norm": 2.3980181217193604,
      "learning_rate": 1e-05,
      "loss": 0.4411,
      "step": 38600
    },
    {
      "epoch": 0.000387,
      "grad_norm": 2.372525453567505,
      "learning_rate": 1e-05,
      "loss": 0.4362,
      "step": 38700
    },
    {
      "epoch": 0.000388,
      "grad_norm": 2.3161978721618652,
      "learning_rate": 1e-05,
      "loss": 0.4405,
      "step": 38800
    },
    {
      "epoch": 0.000389,
      "grad_norm": 2.4494197368621826,
      "learning_rate": 1e-05,
      "loss": 0.4374,
      "step": 38900
    },
    {
      "epoch": 0.00039,
      "grad_norm": 2.3522799015045166,
      "learning_rate": 1e-05,
      "loss": 0.4432,
      "step": 39000
    },
    {
      "epoch": 0.000391,
      "grad_norm": 2.3715996742248535,
      "learning_rate": 1e-05,
      "loss": 0.4237,
      "step": 39100
    },
    {
      "epoch": 0.000392,
      "grad_norm": 2.429914951324463,
      "learning_rate": 1e-05,
      "loss": 0.4332,
      "step": 39200
    },
    {
      "epoch": 0.000393,
      "grad_norm": 2.4736123085021973,
      "learning_rate": 1e-05,
      "loss": 0.4345,
      "step": 39300
    },
    {
      "epoch": 0.000394,
      "grad_norm": 2.033489942550659,
      "learning_rate": 1e-05,
      "loss": 0.4371,
      "step": 39400
    },
    {
      "epoch": 0.000395,
      "grad_norm": 2.22459077835083,
      "learning_rate": 1e-05,
      "loss": 0.4336,
      "step": 39500
    },
    {
      "epoch": 0.000396,
      "grad_norm": 2.475951910018921,
      "learning_rate": 1e-05,
      "loss": 0.4349,
      "step": 39600
    },
    {
      "epoch": 0.000397,
      "grad_norm": 2.2297749519348145,
      "learning_rate": 1e-05,
      "loss": 0.4274,
      "step": 39700
    },
    {
      "epoch": 0.000398,
      "grad_norm": 2.445439338684082,
      "learning_rate": 1e-05,
      "loss": 0.4366,
      "step": 39800
    },
    {
      "epoch": 0.000399,
      "grad_norm": 2.4138917922973633,
      "learning_rate": 1e-05,
      "loss": 0.4351,
      "step": 39900
    },
    {
      "epoch": 0.0004,
      "grad_norm": 2.551255226135254,
      "learning_rate": 1e-05,
      "loss": 0.4333,
      "step": 40000
    },
    {
      "epoch": 0.0004,
      "eval_loss": 0.4091796875,
      "eval_runtime": 110.8814,
      "eval_samples_per_second": 450.932,
      "eval_steps_per_second": 28.183,
      "step": 40000
    },
    {
      "epoch": 0.000401,
      "grad_norm": 2.6333506107330322,
      "learning_rate": 1e-05,
      "loss": 0.4332,
      "step": 40100
    },
    {
      "epoch": 0.000402,
      "grad_norm": 2.0419421195983887,
      "learning_rate": 1e-05,
      "loss": 0.4324,
      "step": 40200
    },
    {
      "epoch": 0.000403,
      "grad_norm": 2.1599907875061035,
      "learning_rate": 1e-05,
      "loss": 0.4261,
      "step": 40300
    },
    {
      "epoch": 0.000404,
      "grad_norm": 2.1608216762542725,
      "learning_rate": 1e-05,
      "loss": 0.4265,
      "step": 40400
    },
    {
      "epoch": 0.000405,
      "grad_norm": 2.0742979049682617,
      "learning_rate": 1e-05,
      "loss": 0.4298,
      "step": 40500
    },
    {
      "epoch": 0.000406,
      "grad_norm": 2.487959146499634,
      "learning_rate": 1e-05,
      "loss": 0.43,
      "step": 40600
    },
    {
      "epoch": 0.000407,
      "grad_norm": 2.436591148376465,
      "learning_rate": 1e-05,
      "loss": 0.4356,
      "step": 40700
    },
    {
      "epoch": 0.000408,
      "grad_norm": 2.2447760105133057,
      "learning_rate": 1e-05,
      "loss": 0.4269,
      "step": 40800
    },
    {
      "epoch": 0.000409,
      "grad_norm": 2.1390585899353027,
      "learning_rate": 1e-05,
      "loss": 0.4268,
      "step": 40900
    },
    {
      "epoch": 0.00041,
      "grad_norm": 2.390690326690674,
      "learning_rate": 1e-05,
      "loss": 0.4233,
      "step": 41000
    },
    {
      "epoch": 0.000411,
      "grad_norm": 1.8873522281646729,
      "learning_rate": 1e-05,
      "loss": 0.4259,
      "step": 41100
    },
    {
      "epoch": 0.000412,
      "grad_norm": 2.3923439979553223,
      "learning_rate": 1e-05,
      "loss": 0.4296,
      "step": 41200
    },
    {
      "epoch": 0.000413,
      "grad_norm": 2.7505736351013184,
      "learning_rate": 1e-05,
      "loss": 0.4265,
      "step": 41300
    },
    {
      "epoch": 0.000414,
      "grad_norm": 2.2666115760803223,
      "learning_rate": 1e-05,
      "loss": 0.4339,
      "step": 41400
    },
    {
      "epoch": 0.000415,
      "grad_norm": 2.128662586212158,
      "learning_rate": 1e-05,
      "loss": 0.4279,
      "step": 41500
    },
    {
      "epoch": 0.000416,
      "grad_norm": 2.4806056022644043,
      "learning_rate": 1e-05,
      "loss": 0.4335,
      "step": 41600
    },
    {
      "epoch": 0.000417,
      "grad_norm": 2.5022566318511963,
      "learning_rate": 1e-05,
      "loss": 0.4321,
      "step": 41700
    },
    {
      "epoch": 0.000418,
      "grad_norm": 2.682896137237549,
      "learning_rate": 1e-05,
      "loss": 0.4373,
      "step": 41800
    },
    {
      "epoch": 0.000419,
      "grad_norm": 2.7449374198913574,
      "learning_rate": 1e-05,
      "loss": 0.4362,
      "step": 41900
    },
    {
      "epoch": 0.00042,
      "grad_norm": 2.0306496620178223,
      "learning_rate": 1e-05,
      "loss": 0.4264,
      "step": 42000
    },
    {
      "epoch": 0.000421,
      "grad_norm": 2.0226821899414062,
      "learning_rate": 1e-05,
      "loss": 0.4301,
      "step": 42100
    },
    {
      "epoch": 0.000422,
      "grad_norm": 2.372490406036377,
      "learning_rate": 1e-05,
      "loss": 0.429,
      "step": 42200
    },
    {
      "epoch": 0.000423,
      "grad_norm": 2.4113259315490723,
      "learning_rate": 1e-05,
      "loss": 0.4347,
      "step": 42300
    },
    {
      "epoch": 0.000424,
      "grad_norm": 2.3437299728393555,
      "learning_rate": 1e-05,
      "loss": 0.4204,
      "step": 42400
    },
    {
      "epoch": 0.000425,
      "grad_norm": 2.460440158843994,
      "learning_rate": 1e-05,
      "loss": 0.4195,
      "step": 42500
    },
    {
      "epoch": 0.000426,
      "grad_norm": 2.3652024269104004,
      "learning_rate": 1e-05,
      "loss": 0.418,
      "step": 42600
    },
    {
      "epoch": 0.000427,
      "grad_norm": 2.48496150970459,
      "learning_rate": 1e-05,
      "loss": 0.4263,
      "step": 42700
    },
    {
      "epoch": 0.000428,
      "grad_norm": 2.4841647148132324,
      "learning_rate": 1e-05,
      "loss": 0.4259,
      "step": 42800
    },
    {
      "epoch": 0.000429,
      "grad_norm": 2.573284149169922,
      "learning_rate": 1e-05,
      "loss": 0.4302,
      "step": 42900
    },
    {
      "epoch": 0.00043,
      "grad_norm": 1.9319133758544922,
      "learning_rate": 1e-05,
      "loss": 0.4227,
      "step": 43000
    },
    {
      "epoch": 0.000431,
      "grad_norm": 2.4806172847747803,
      "learning_rate": 1e-05,
      "loss": 0.4325,
      "step": 43100
    },
    {
      "epoch": 0.000432,
      "grad_norm": 2.4880504608154297,
      "learning_rate": 1e-05,
      "loss": 0.4229,
      "step": 43200
    },
    {
      "epoch": 0.000433,
      "grad_norm": 2.395817995071411,
      "learning_rate": 1e-05,
      "loss": 0.43,
      "step": 43300
    },
    {
      "epoch": 0.000434,
      "grad_norm": 2.6356828212738037,
      "learning_rate": 1e-05,
      "loss": 0.4207,
      "step": 43400
    },
    {
      "epoch": 0.000435,
      "grad_norm": 2.1054494380950928,
      "learning_rate": 1e-05,
      "loss": 0.4284,
      "step": 43500
    },
    {
      "epoch": 0.000436,
      "grad_norm": 2.3987913131713867,
      "learning_rate": 1e-05,
      "loss": 0.4267,
      "step": 43600
    },
    {
      "epoch": 0.000437,
      "grad_norm": 2.606956720352173,
      "learning_rate": 1e-05,
      "loss": 0.4203,
      "step": 43700
    },
    {
      "epoch": 0.000438,
      "grad_norm": 2.5072622299194336,
      "learning_rate": 1e-05,
      "loss": 0.4276,
      "step": 43800
    },
    {
      "epoch": 0.000439,
      "grad_norm": 2.300851345062256,
      "learning_rate": 1e-05,
      "loss": 0.427,
      "step": 43900
    },
    {
      "epoch": 0.00044,
      "grad_norm": 2.514756441116333,
      "learning_rate": 1e-05,
      "loss": 0.4233,
      "step": 44000
    },
    {
      "epoch": 0.000441,
      "grad_norm": 2.5738296508789062,
      "learning_rate": 1e-05,
      "loss": 0.4255,
      "step": 44100
    },
    {
      "epoch": 0.000442,
      "grad_norm": 2.324410915374756,
      "learning_rate": 1e-05,
      "loss": 0.4225,
      "step": 44200
    },
    {
      "epoch": 0.000443,
      "grad_norm": 2.1797661781311035,
      "learning_rate": 1e-05,
      "loss": 0.4204,
      "step": 44300
    },
    {
      "epoch": 0.000444,
      "grad_norm": 2.667961359024048,
      "learning_rate": 1e-05,
      "loss": 0.4298,
      "step": 44400
    },
    {
      "epoch": 0.000445,
      "grad_norm": 2.6222057342529297,
      "learning_rate": 1e-05,
      "loss": 0.4159,
      "step": 44500
    },
    {
      "epoch": 0.000446,
      "grad_norm": 2.4339888095855713,
      "learning_rate": 1e-05,
      "loss": 0.4298,
      "step": 44600
    },
    {
      "epoch": 0.000447,
      "grad_norm": 2.394127368927002,
      "learning_rate": 1e-05,
      "loss": 0.4299,
      "step": 44700
    },
    {
      "epoch": 0.000448,
      "grad_norm": 2.3612658977508545,
      "learning_rate": 1e-05,
      "loss": 0.4285,
      "step": 44800
    },
    {
      "epoch": 0.000449,
      "grad_norm": 2.4719297885894775,
      "learning_rate": 1e-05,
      "loss": 0.4218,
      "step": 44900
    },
    {
      "epoch": 0.00045,
      "grad_norm": 2.1661250591278076,
      "learning_rate": 1e-05,
      "loss": 0.4213,
      "step": 45000
    },
    {
      "epoch": 0.000451,
      "grad_norm": 2.559985637664795,
      "learning_rate": 1e-05,
      "loss": 0.4289,
      "step": 45100
    },
    {
      "epoch": 0.000452,
      "grad_norm": 2.452289342880249,
      "learning_rate": 1e-05,
      "loss": 0.4257,
      "step": 45200
    },
    {
      "epoch": 0.000453,
      "grad_norm": 2.0307326316833496,
      "learning_rate": 1e-05,
      "loss": 0.4229,
      "step": 45300
    },
    {
      "epoch": 0.000454,
      "grad_norm": 2.489323377609253,
      "learning_rate": 1e-05,
      "loss": 0.4269,
      "step": 45400
    },
    {
      "epoch": 0.000455,
      "grad_norm": 2.5684876441955566,
      "learning_rate": 1e-05,
      "loss": 0.4173,
      "step": 45500
    },
    {
      "epoch": 0.000456,
      "grad_norm": 2.4012162685394287,
      "learning_rate": 1e-05,
      "loss": 0.4139,
      "step": 45600
    },
    {
      "epoch": 0.000457,
      "grad_norm": 2.4833133220672607,
      "learning_rate": 1e-05,
      "loss": 0.4233,
      "step": 45700
    },
    {
      "epoch": 0.000458,
      "grad_norm": 2.4413490295410156,
      "learning_rate": 1e-05,
      "loss": 0.4089,
      "step": 45800
    },
    {
      "epoch": 0.000459,
      "grad_norm": 2.192959785461426,
      "learning_rate": 1e-05,
      "loss": 0.4156,
      "step": 45900
    },
    {
      "epoch": 0.00046,
      "grad_norm": 2.3064184188842773,
      "learning_rate": 1e-05,
      "loss": 0.4187,
      "step": 46000
    },
    {
      "epoch": 0.000461,
      "grad_norm": 2.4842922687530518,
      "learning_rate": 1e-05,
      "loss": 0.4131,
      "step": 46100
    },
    {
      "epoch": 0.000462,
      "grad_norm": 2.074312925338745,
      "learning_rate": 1e-05,
      "loss": 0.4265,
      "step": 46200
    },
    {
      "epoch": 0.000463,
      "grad_norm": 2.4513862133026123,
      "learning_rate": 1e-05,
      "loss": 0.4153,
      "step": 46300
    },
    {
      "epoch": 0.000464,
      "grad_norm": 2.21403431892395,
      "learning_rate": 1e-05,
      "loss": 0.417,
      "step": 46400
    },
    {
      "epoch": 0.000465,
      "grad_norm": 2.469252109527588,
      "learning_rate": 1e-05,
      "loss": 0.4181,
      "step": 46500
    },
    {
      "epoch": 0.000466,
      "grad_norm": 2.3535096645355225,
      "learning_rate": 1e-05,
      "loss": 0.4159,
      "step": 46600
    },
    {
      "epoch": 0.000467,
      "grad_norm": 2.8495819568634033,
      "learning_rate": 1e-05,
      "loss": 0.4118,
      "step": 46700
    },
    {
      "epoch": 0.000468,
      "grad_norm": 2.2924575805664062,
      "learning_rate": 1e-05,
      "loss": 0.4221,
      "step": 46800
    },
    {
      "epoch": 0.000469,
      "grad_norm": 2.1148035526275635,
      "learning_rate": 1e-05,
      "loss": 0.4221,
      "step": 46900
    },
    {
      "epoch": 0.00047,
      "grad_norm": 2.5257456302642822,
      "learning_rate": 1e-05,
      "loss": 0.4129,
      "step": 47000
    },
    {
      "epoch": 0.000471,
      "grad_norm": 2.734550714492798,
      "learning_rate": 1e-05,
      "loss": 0.4214,
      "step": 47100
    },
    {
      "epoch": 0.000472,
      "grad_norm": 2.844151496887207,
      "learning_rate": 1e-05,
      "loss": 0.4133,
      "step": 47200
    },
    {
      "epoch": 0.000473,
      "grad_norm": 2.2594943046569824,
      "learning_rate": 1e-05,
      "loss": 0.4154,
      "step": 47300
    },
    {
      "epoch": 0.000474,
      "grad_norm": 2.037102699279785,
      "learning_rate": 1e-05,
      "loss": 0.4095,
      "step": 47400
    },
    {
      "epoch": 0.000475,
      "grad_norm": 2.472301483154297,
      "learning_rate": 1e-05,
      "loss": 0.4156,
      "step": 47500
    },
    {
      "epoch": 0.000476,
      "grad_norm": 2.0751333236694336,
      "learning_rate": 1e-05,
      "loss": 0.4196,
      "step": 47600
    },
    {
      "epoch": 0.000477,
      "grad_norm": 2.197103977203369,
      "learning_rate": 1e-05,
      "loss": 0.4129,
      "step": 47700
    },
    {
      "epoch": 0.000478,
      "grad_norm": 2.5141637325286865,
      "learning_rate": 1e-05,
      "loss": 0.4085,
      "step": 47800
    },
    {
      "epoch": 0.000479,
      "grad_norm": 2.441049337387085,
      "learning_rate": 1e-05,
      "loss": 0.4137,
      "step": 47900
    },
    {
      "epoch": 0.00048,
      "grad_norm": 2.2101807594299316,
      "learning_rate": 1e-05,
      "loss": 0.415,
      "step": 48000
    },
    {
      "epoch": 0.000481,
      "grad_norm": 2.235775947570801,
      "learning_rate": 1e-05,
      "loss": 0.4169,
      "step": 48100
    },
    {
      "epoch": 0.000482,
      "grad_norm": 2.0968542098999023,
      "learning_rate": 1e-05,
      "loss": 0.4131,
      "step": 48200
    },
    {
      "epoch": 0.000483,
      "grad_norm": 2.1529128551483154,
      "learning_rate": 1e-05,
      "loss": 0.42,
      "step": 48300
    },
    {
      "epoch": 0.000484,
      "grad_norm": 2.2251384258270264,
      "learning_rate": 1e-05,
      "loss": 0.4153,
      "step": 48400
    },
    {
      "epoch": 0.000485,
      "grad_norm": 2.5039467811584473,
      "learning_rate": 1e-05,
      "loss": 0.4133,
      "step": 48500
    },
    {
      "epoch": 0.000486,
      "grad_norm": 2.3165180683135986,
      "learning_rate": 1e-05,
      "loss": 0.4104,
      "step": 48600
    },
    {
      "epoch": 0.000487,
      "grad_norm": 2.2625648975372314,
      "learning_rate": 1e-05,
      "loss": 0.4047,
      "step": 48700
    },
    {
      "epoch": 0.000488,
      "grad_norm": 2.4777987003326416,
      "learning_rate": 1e-05,
      "loss": 0.4112,
      "step": 48800
    },
    {
      "epoch": 0.000489,
      "grad_norm": 2.794090986251831,
      "learning_rate": 1e-05,
      "loss": 0.4143,
      "step": 48900
    },
    {
      "epoch": 0.00049,
      "grad_norm": 2.3887550830841064,
      "learning_rate": 1e-05,
      "loss": 0.4172,
      "step": 49000
    },
    {
      "epoch": 0.000491,
      "grad_norm": 2.0181326866149902,
      "learning_rate": 1e-05,
      "loss": 0.4147,
      "step": 49100
    },
    {
      "epoch": 0.000492,
      "grad_norm": 2.037066698074341,
      "learning_rate": 1e-05,
      "loss": 0.4079,
      "step": 49200
    },
    {
      "epoch": 0.000493,
      "grad_norm": 2.349827289581299,
      "learning_rate": 1e-05,
      "loss": 0.4203,
      "step": 49300
    },
    {
      "epoch": 0.000494,
      "grad_norm": 2.35591459274292,
      "learning_rate": 1e-05,
      "loss": 0.4096,
      "step": 49400
    },
    {
      "epoch": 0.000495,
      "grad_norm": 2.994199752807617,
      "learning_rate": 1e-05,
      "loss": 0.4171,
      "step": 49500
    },
    {
      "epoch": 0.000496,
      "grad_norm": 2.415408134460449,
      "learning_rate": 1e-05,
      "loss": 0.4027,
      "step": 49600
    },
    {
      "epoch": 0.000497,
      "grad_norm": 2.205004930496216,
      "learning_rate": 1e-05,
      "loss": 0.4208,
      "step": 49700
    },
    {
      "epoch": 0.000498,
      "grad_norm": 2.1636242866516113,
      "learning_rate": 1e-05,
      "loss": 0.4131,
      "step": 49800
    },
    {
      "epoch": 0.000499,
      "grad_norm": 2.4507057666778564,
      "learning_rate": 1e-05,
      "loss": 0.4135,
      "step": 49900
    },
    {
      "epoch": 0.0005,
      "grad_norm": 2.1614506244659424,
      "learning_rate": 1e-05,
      "loss": 0.41,
      "step": 50000
    },
    {
      "epoch": 0.000501,
      "grad_norm": 2.070063591003418,
      "learning_rate": 1e-05,
      "loss": 0.4089,
      "step": 50100
    },
    {
      "epoch": 0.000502,
      "grad_norm": 2.339935779571533,
      "learning_rate": 1e-05,
      "loss": 0.4115,
      "step": 50200
    },
    {
      "epoch": 0.000503,
      "grad_norm": 2.25191330909729,
      "learning_rate": 1e-05,
      "loss": 0.4064,
      "step": 50300
    },
    {
      "epoch": 0.000504,
      "grad_norm": 2.198077440261841,
      "learning_rate": 1e-05,
      "loss": 0.4049,
      "step": 50400
    },
    {
      "epoch": 0.000505,
      "grad_norm": 2.4001047611236572,
      "learning_rate": 1e-05,
      "loss": 0.4146,
      "step": 50500
    },
    {
      "epoch": 0.000506,
      "grad_norm": 2.311879873275757,
      "learning_rate": 1e-05,
      "loss": 0.4149,
      "step": 50600
    },
    {
      "epoch": 0.000507,
      "grad_norm": 2.249931573867798,
      "learning_rate": 1e-05,
      "loss": 0.4091,
      "step": 50700
    },
    {
      "epoch": 0.000508,
      "grad_norm": 1.957440733909607,
      "learning_rate": 1e-05,
      "loss": 0.407,
      "step": 50800
    },
    {
      "epoch": 0.000509,
      "grad_norm": 2.3248322010040283,
      "learning_rate": 1e-05,
      "loss": 0.4138,
      "step": 50900
    },
    {
      "epoch": 0.00051,
      "grad_norm": 2.2417356967926025,
      "learning_rate": 1e-05,
      "loss": 0.4075,
      "step": 51000
    },
    {
      "epoch": 0.000511,
      "grad_norm": 2.2850210666656494,
      "learning_rate": 1e-05,
      "loss": 0.4148,
      "step": 51100
    },
    {
      "epoch": 0.000512,
      "grad_norm": 2.3242995738983154,
      "learning_rate": 1e-05,
      "loss": 0.4134,
      "step": 51200
    },
    {
      "epoch": 0.000513,
      "grad_norm": 2.0709969997406006,
      "learning_rate": 1e-05,
      "loss": 0.4196,
      "step": 51300
    },
    {
      "epoch": 0.000514,
      "grad_norm": 2.149703025817871,
      "learning_rate": 1e-05,
      "loss": 0.4105,
      "step": 51400
    },
    {
      "epoch": 0.000515,
      "grad_norm": 2.2497308254241943,
      "learning_rate": 1e-05,
      "loss": 0.4098,
      "step": 51500
    },
    {
      "epoch": 0.000516,
      "grad_norm": 2.0328240394592285,
      "learning_rate": 1e-05,
      "loss": 0.406,
      "step": 51600
    },
    {
      "epoch": 0.000517,
      "grad_norm": 2.052591562271118,
      "learning_rate": 1e-05,
      "loss": 0.4086,
      "step": 51700
    },
    {
      "epoch": 0.000518,
      "grad_norm": 2.353180170059204,
      "learning_rate": 1e-05,
      "loss": 0.4058,
      "step": 51800
    },
    {
      "epoch": 0.000519,
      "grad_norm": 2.352935791015625,
      "learning_rate": 1e-05,
      "loss": 0.4057,
      "step": 51900
    },
    {
      "epoch": 0.00052,
      "grad_norm": 2.1475372314453125,
      "learning_rate": 1e-05,
      "loss": 0.4043,
      "step": 52000
    },
    {
      "epoch": 0.000521,
      "grad_norm": 2.2819299697875977,
      "learning_rate": 1e-05,
      "loss": 0.4079,
      "step": 52100
    },
    {
      "epoch": 0.000522,
      "grad_norm": 2.23323392868042,
      "learning_rate": 1e-05,
      "loss": 0.4142,
      "step": 52200
    },
    {
      "epoch": 0.000523,
      "grad_norm": 2.1115095615386963,
      "learning_rate": 1e-05,
      "loss": 0.4091,
      "step": 52300
    },
    {
      "epoch": 0.000524,
      "grad_norm": 2.34243106842041,
      "learning_rate": 1e-05,
      "loss": 0.4028,
      "step": 52400
    },
    {
      "epoch": 0.000525,
      "grad_norm": 2.1626434326171875,
      "learning_rate": 1e-05,
      "loss": 0.4042,
      "step": 52500
    },
    {
      "epoch": 0.000526,
      "grad_norm": 2.283756732940674,
      "learning_rate": 1e-05,
      "loss": 0.4068,
      "step": 52600
    },
    {
      "epoch": 0.000527,
      "grad_norm": 2.1026229858398438,
      "learning_rate": 1e-05,
      "loss": 0.4082,
      "step": 52700
    },
    {
      "epoch": 0.000528,
      "grad_norm": 2.1464221477508545,
      "learning_rate": 1e-05,
      "loss": 0.4093,
      "step": 52800
    },
    {
      "epoch": 0.000529,
      "grad_norm": 2.1100659370422363,
      "learning_rate": 1e-05,
      "loss": 0.4049,
      "step": 52900
    },
    {
      "epoch": 0.00053,
      "grad_norm": 2.06082820892334,
      "learning_rate": 1e-05,
      "loss": 0.4174,
      "step": 53000
    },
    {
      "epoch": 0.000531,
      "grad_norm": 2.226346492767334,
      "learning_rate": 1e-05,
      "loss": 0.4084,
      "step": 53100
    },
    {
      "epoch": 0.000532,
      "grad_norm": 2.432999849319458,
      "learning_rate": 1e-05,
      "loss": 0.406,
      "step": 53200
    },
    {
      "epoch": 0.000533,
      "grad_norm": 2.4239957332611084,
      "learning_rate": 1e-05,
      "loss": 0.4073,
      "step": 53300
    },
    {
      "epoch": 0.000534,
      "grad_norm": 2.08341646194458,
      "learning_rate": 1e-05,
      "loss": 0.4092,
      "step": 53400
    },
    {
      "epoch": 0.000535,
      "grad_norm": 2.4562456607818604,
      "learning_rate": 1e-05,
      "loss": 0.3995,
      "step": 53500
    },
    {
      "epoch": 0.000536,
      "grad_norm": 2.1700892448425293,
      "learning_rate": 1e-05,
      "loss": 0.4107,
      "step": 53600
    },
    {
      "epoch": 0.000537,
      "grad_norm": 2.071171998977661,
      "learning_rate": 1e-05,
      "loss": 0.4085,
      "step": 53700
    },
    {
      "epoch": 0.000538,
      "grad_norm": 1.977064847946167,
      "learning_rate": 1e-05,
      "loss": 0.4046,
      "step": 53800
    },
    {
      "epoch": 0.000539,
      "grad_norm": 4.314730644226074,
      "learning_rate": 1e-05,
      "loss": 0.4023,
      "step": 53900
    },
    {
      "epoch": 0.00054,
      "grad_norm": 2.5248172283172607,
      "learning_rate": 1e-05,
      "loss": 0.3983,
      "step": 54000
    },
    {
      "epoch": 0.000541,
      "grad_norm": 2.1219537258148193,
      "learning_rate": 1e-05,
      "loss": 0.4068,
      "step": 54100
    },
    {
      "epoch": 0.000542,
      "grad_norm": 2.4074840545654297,
      "learning_rate": 1e-05,
      "loss": 0.4029,
      "step": 54200
    },
    {
      "epoch": 0.000543,
      "grad_norm": 2.462904930114746,
      "learning_rate": 1e-05,
      "loss": 0.4082,
      "step": 54300
    },
    {
      "epoch": 0.000544,
      "grad_norm": 2.5849449634552,
      "learning_rate": 1e-05,
      "loss": 0.3998,
      "step": 54400
    },
    {
      "epoch": 0.000545,
      "grad_norm": 2.1051547527313232,
      "learning_rate": 1e-05,
      "loss": 0.4018,
      "step": 54500
    },
    {
      "epoch": 0.000546,
      "grad_norm": 2.4176714420318604,
      "learning_rate": 1e-05,
      "loss": 0.3965,
      "step": 54600
    },
    {
      "epoch": 0.000547,
      "grad_norm": 2.1228177547454834,
      "learning_rate": 1e-05,
      "loss": 0.4007,
      "step": 54700
    },
    {
      "epoch": 0.000548,
      "grad_norm": 2.0286078453063965,
      "learning_rate": 1e-05,
      "loss": 0.402,
      "step": 54800
    },
    {
      "epoch": 0.000549,
      "grad_norm": 2.300497531890869,
      "learning_rate": 1e-05,
      "loss": 0.4084,
      "step": 54900
    },
    {
      "epoch": 0.00055,
      "grad_norm": 2.1815927028656006,
      "learning_rate": 1e-05,
      "loss": 0.4031,
      "step": 55000
    },
    {
      "epoch": 0.000551,
      "grad_norm": 2.347383975982666,
      "learning_rate": 1e-05,
      "loss": 0.402,
      "step": 55100
    },
    {
      "epoch": 0.000552,
      "grad_norm": 2.059412717819214,
      "learning_rate": 1e-05,
      "loss": 0.4169,
      "step": 55200
    },
    {
      "epoch": 0.000553,
      "grad_norm": 2.089460849761963,
      "learning_rate": 1e-05,
      "loss": 0.4059,
      "step": 55300
    },
    {
      "epoch": 0.000554,
      "grad_norm": 2.608187675476074,
      "learning_rate": 1e-05,
      "loss": 0.4059,
      "step": 55400
    },
    {
      "epoch": 0.000555,
      "grad_norm": 2.468566656112671,
      "learning_rate": 1e-05,
      "loss": 0.4007,
      "step": 55500
    },
    {
      "epoch": 0.000556,
      "grad_norm": 2.740276336669922,
      "learning_rate": 1e-05,
      "loss": 0.4099,
      "step": 55600
    },
    {
      "epoch": 0.000557,
      "grad_norm": 2.447087526321411,
      "learning_rate": 1e-05,
      "loss": 0.4157,
      "step": 55700
    },
    {
      "epoch": 0.000558,
      "grad_norm": 2.1900322437286377,
      "learning_rate": 1e-05,
      "loss": 0.4018,
      "step": 55800
    },
    {
      "epoch": 0.000559,
      "grad_norm": 2.332939386367798,
      "learning_rate": 1e-05,
      "loss": 0.3949,
      "step": 55900
    },
    {
      "epoch": 0.00056,
      "grad_norm": 2.050628900527954,
      "learning_rate": 1e-05,
      "loss": 0.4062,
      "step": 56000
    },
    {
      "epoch": 0.000561,
      "grad_norm": 2.101712226867676,
      "learning_rate": 1e-05,
      "loss": 0.4014,
      "step": 56100
    },
    {
      "epoch": 0.000562,
      "grad_norm": 2.093705177307129,
      "learning_rate": 1e-05,
      "loss": 0.3951,
      "step": 56200
    },
    {
      "epoch": 0.000563,
      "grad_norm": 2.02903413772583,
      "learning_rate": 1e-05,
      "loss": 0.4059,
      "step": 56300
    },
    {
      "epoch": 0.000564,
      "grad_norm": 2.0588796138763428,
      "learning_rate": 1e-05,
      "loss": 0.4104,
      "step": 56400
    },
    {
      "epoch": 0.000565,
      "grad_norm": 1.968138575553894,
      "learning_rate": 1e-05,
      "loss": 0.3955,
      "step": 56500
    },
    {
      "epoch": 0.000566,
      "grad_norm": 2.0863802433013916,
      "learning_rate": 1e-05,
      "loss": 0.4056,
      "step": 56600
    },
    {
      "epoch": 0.000567,
      "grad_norm": 2.0999319553375244,
      "learning_rate": 1e-05,
      "loss": 0.3999,
      "step": 56700
    },
    {
      "epoch": 0.000568,
      "grad_norm": 2.2497940063476562,
      "learning_rate": 1e-05,
      "loss": 0.3944,
      "step": 56800
    },
    {
      "epoch": 0.000569,
      "grad_norm": 2.327509880065918,
      "learning_rate": 1e-05,
      "loss": 0.4045,
      "step": 56900
    },
    {
      "epoch": 0.00057,
      "grad_norm": 1.9509259462356567,
      "learning_rate": 1e-05,
      "loss": 0.3937,
      "step": 57000
    },
    {
      "epoch": 0.000571,
      "grad_norm": 1.9733527898788452,
      "learning_rate": 1e-05,
      "loss": 0.3994,
      "step": 57100
    },
    {
      "epoch": 0.000572,
      "grad_norm": 2.3149795532226562,
      "learning_rate": 1e-05,
      "loss": 0.3964,
      "step": 57200
    },
    {
      "epoch": 0.000573,
      "grad_norm": 2.2869510650634766,
      "learning_rate": 1e-05,
      "loss": 0.3972,
      "step": 57300
    },
    {
      "epoch": 0.000574,
      "grad_norm": 2.807288885116577,
      "learning_rate": 1e-05,
      "loss": 0.3979,
      "step": 57400
    },
    {
      "epoch": 0.000575,
      "grad_norm": 1.9130806922912598,
      "learning_rate": 1e-05,
      "loss": 0.3909,
      "step": 57500
    },
    {
      "epoch": 0.000576,
      "grad_norm": 2.392228841781616,
      "learning_rate": 1e-05,
      "loss": 0.4081,
      "step": 57600
    },
    {
      "epoch": 0.000577,
      "grad_norm": 2.2016382217407227,
      "learning_rate": 1e-05,
      "loss": 0.3942,
      "step": 57700
    },
    {
      "epoch": 0.000578,
      "grad_norm": 1.9153637886047363,
      "learning_rate": 1e-05,
      "loss": 0.3856,
      "step": 57800
    },
    {
      "epoch": 0.000579,
      "grad_norm": 2.334127902984619,
      "learning_rate": 1e-05,
      "loss": 0.4011,
      "step": 57900
    },
    {
      "epoch": 0.00058,
      "grad_norm": 2.0389389991760254,
      "learning_rate": 1e-05,
      "loss": 0.3964,
      "step": 58000
    },
    {
      "epoch": 0.000581,
      "grad_norm": 1.817014217376709,
      "learning_rate": 1e-05,
      "loss": 0.401,
      "step": 58100
    },
    {
      "epoch": 0.000582,
      "grad_norm": 2.2769718170166016,
      "learning_rate": 1e-05,
      "loss": 0.4025,
      "step": 58200
    },
    {
      "epoch": 0.000583,
      "grad_norm": 2.2681713104248047,
      "learning_rate": 1e-05,
      "loss": 0.3998,
      "step": 58300
    },
    {
      "epoch": 0.000584,
      "grad_norm": 2.0518765449523926,
      "learning_rate": 1e-05,
      "loss": 0.3958,
      "step": 58400
    },
    {
      "epoch": 0.000585,
      "grad_norm": 2.0787107944488525,
      "learning_rate": 1e-05,
      "loss": 0.3979,
      "step": 58500
    },
    {
      "epoch": 0.000586,
      "grad_norm": 2.2114005088806152,
      "learning_rate": 1e-05,
      "loss": 0.3953,
      "step": 58600
    },
    {
      "epoch": 0.000587,
      "grad_norm": 2.2382404804229736,
      "learning_rate": 1e-05,
      "loss": 0.4011,
      "step": 58700
    },
    {
      "epoch": 0.000588,
      "grad_norm": 2.2104434967041016,
      "learning_rate": 1e-05,
      "loss": 0.3897,
      "step": 58800
    },
    {
      "epoch": 0.000589,
      "grad_norm": 2.5010359287261963,
      "learning_rate": 1e-05,
      "loss": 0.3929,
      "step": 58900
    },
    {
      "epoch": 0.00059,
      "grad_norm": 2.6456377506256104,
      "learning_rate": 1e-05,
      "loss": 0.4048,
      "step": 59000
    },
    {
      "epoch": 0.000591,
      "grad_norm": 2.2201075553894043,
      "learning_rate": 1e-05,
      "loss": 0.4011,
      "step": 59100
    },
    {
      "epoch": 0.000592,
      "grad_norm": 2.050746440887451,
      "learning_rate": 1e-05,
      "loss": 0.4003,
      "step": 59200
    },
    {
      "epoch": 0.000593,
      "grad_norm": 4.129772663116455,
      "learning_rate": 1e-05,
      "loss": 0.3889,
      "step": 59300
    },
    {
      "epoch": 0.000594,
      "grad_norm": 2.160189628601074,
      "learning_rate": 1e-05,
      "loss": 0.4046,
      "step": 59400
    },
    {
      "epoch": 0.000595,
      "grad_norm": 2.5370826721191406,
      "learning_rate": 1e-05,
      "loss": 0.3829,
      "step": 59500
    },
    {
      "epoch": 0.000596,
      "grad_norm": 2.0123531818389893,
      "learning_rate": 1e-05,
      "loss": 0.3976,
      "step": 59600
    },
    {
      "epoch": 0.000597,
      "grad_norm": 2.175504207611084,
      "learning_rate": 1e-05,
      "loss": 0.3908,
      "step": 59700
    },
    {
      "epoch": 0.000598,
      "grad_norm": 1.993752360343933,
      "learning_rate": 1e-05,
      "loss": 0.401,
      "step": 59800
    },
    {
      "epoch": 0.000599,
      "grad_norm": 2.1103925704956055,
      "learning_rate": 1e-05,
      "loss": 0.3886,
      "step": 59900
    },
    {
      "epoch": 0.0006,
      "grad_norm": 1.866847038269043,
      "learning_rate": 1e-05,
      "loss": 0.3929,
      "step": 60000
    },
    {
      "epoch": 0.0006,
      "eval_loss": 0.3701171875,
      "eval_runtime": 109.3176,
      "eval_samples_per_second": 457.383,
      "eval_steps_per_second": 28.586,
      "step": 60000
    },
    {
      "epoch": 0.000601,
      "grad_norm": 2.07293701171875,
      "learning_rate": 1e-05,
      "loss": 0.3941,
      "step": 60100
    },
    {
      "epoch": 0.000602,
      "grad_norm": 2.153665542602539,
      "learning_rate": 1e-05,
      "loss": 0.3933,
      "step": 60200
    },
    {
      "epoch": 0.000603,
      "grad_norm": 1.8806813955307007,
      "learning_rate": 1e-05,
      "loss": 0.3908,
      "step": 60300
    },
    {
      "epoch": 0.000604,
      "grad_norm": 1.9428766965866089,
      "learning_rate": 1e-05,
      "loss": 0.3941,
      "step": 60400
    },
    {
      "epoch": 0.000605,
      "grad_norm": 2.4207301139831543,
      "learning_rate": 1e-05,
      "loss": 0.395,
      "step": 60500
    },
    {
      "epoch": 0.000606,
      "grad_norm": 2.292665958404541,
      "learning_rate": 1e-05,
      "loss": 0.3893,
      "step": 60600
    },
    {
      "epoch": 0.000607,
      "grad_norm": 2.2332205772399902,
      "learning_rate": 1e-05,
      "loss": 0.3939,
      "step": 60700
    },
    {
      "epoch": 0.000608,
      "grad_norm": 2.379991054534912,
      "learning_rate": 1e-05,
      "loss": 0.3911,
      "step": 60800
    },
    {
      "epoch": 0.000609,
      "grad_norm": 2.1357316970825195,
      "learning_rate": 1e-05,
      "loss": 0.394,
      "step": 60900
    },
    {
      "epoch": 0.00061,
      "grad_norm": 2.218677520751953,
      "learning_rate": 1e-05,
      "loss": 0.3908,
      "step": 61000
    },
    {
      "epoch": 0.000611,
      "grad_norm": 2.144749641418457,
      "learning_rate": 1e-05,
      "loss": 0.3948,
      "step": 61100
    },
    {
      "epoch": 0.000612,
      "grad_norm": 1.9096667766571045,
      "learning_rate": 1e-05,
      "loss": 0.4055,
      "step": 61200
    },
    {
      "epoch": 0.000613,
      "grad_norm": 1.813551664352417,
      "learning_rate": 1e-05,
      "loss": 0.3909,
      "step": 61300
    },
    {
      "epoch": 0.000614,
      "grad_norm": 2.0957746505737305,
      "learning_rate": 1e-05,
      "loss": 0.3934,
      "step": 61400
    },
    {
      "epoch": 0.000615,
      "grad_norm": 2.288628578186035,
      "learning_rate": 1e-05,
      "loss": 0.3948,
      "step": 61500
    },
    {
      "epoch": 0.000616,
      "grad_norm": 1.8869370222091675,
      "learning_rate": 1e-05,
      "loss": 0.3896,
      "step": 61600
    },
    {
      "epoch": 0.000617,
      "grad_norm": 1.8751919269561768,
      "learning_rate": 1e-05,
      "loss": 0.3946,
      "step": 61700
    },
    {
      "epoch": 0.000618,
      "grad_norm": 2.0423409938812256,
      "learning_rate": 1e-05,
      "loss": 0.3914,
      "step": 61800
    },
    {
      "epoch": 0.000619,
      "grad_norm": 2.154679536819458,
      "learning_rate": 1e-05,
      "loss": 0.3942,
      "step": 61900
    },
    {
      "epoch": 0.00062,
      "grad_norm": 2.293510913848877,
      "learning_rate": 1e-05,
      "loss": 0.3867,
      "step": 62000
    },
    {
      "epoch": 0.000621,
      "grad_norm": 2.034313678741455,
      "learning_rate": 1e-05,
      "loss": 0.3844,
      "step": 62100
    },
    {
      "epoch": 0.000622,
      "grad_norm": 2.105489730834961,
      "learning_rate": 1e-05,
      "loss": 0.3886,
      "step": 62200
    },
    {
      "epoch": 0.000623,
      "grad_norm": 1.9530473947525024,
      "learning_rate": 1e-05,
      "loss": 0.3962,
      "step": 62300
    },
    {
      "epoch": 0.000624,
      "grad_norm": 2.291125774383545,
      "learning_rate": 1e-05,
      "loss": 0.392,
      "step": 62400
    },
    {
      "epoch": 0.000625,
      "grad_norm": 2.169159412384033,
      "learning_rate": 1e-05,
      "loss": 0.3916,
      "step": 62500
    },
    {
      "epoch": 0.000626,
      "grad_norm": 2.2920339107513428,
      "learning_rate": 1e-05,
      "loss": 0.3886,
      "step": 62600
    },
    {
      "epoch": 0.000627,
      "grad_norm": 2.248567819595337,
      "learning_rate": 1e-05,
      "loss": 0.3997,
      "step": 62700
    },
    {
      "epoch": 0.000628,
      "grad_norm": 1.9369299411773682,
      "learning_rate": 1e-05,
      "loss": 0.3945,
      "step": 62800
    },
    {
      "epoch": 0.000629,
      "grad_norm": 1.912782073020935,
      "learning_rate": 1e-05,
      "loss": 0.3911,
      "step": 62900
    },
    {
      "epoch": 0.00063,
      "grad_norm": 1.7592915296554565,
      "learning_rate": 1e-05,
      "loss": 0.3989,
      "step": 63000
    },
    {
      "epoch": 0.000631,
      "grad_norm": 2.0076982975006104,
      "learning_rate": 1e-05,
      "loss": 0.3918,
      "step": 63100
    },
    {
      "epoch": 0.000632,
      "grad_norm": 2.0114753246307373,
      "learning_rate": 1e-05,
      "loss": 0.3932,
      "step": 63200
    },
    {
      "epoch": 0.000633,
      "grad_norm": 2.50410795211792,
      "learning_rate": 1e-05,
      "loss": 0.3958,
      "step": 63300
    },
    {
      "epoch": 0.000634,
      "grad_norm": 2.156872510910034,
      "learning_rate": 1e-05,
      "loss": 0.3875,
      "step": 63400
    },
    {
      "epoch": 0.000635,
      "grad_norm": 2.2408478260040283,
      "learning_rate": 1e-05,
      "loss": 0.3939,
      "step": 63500
    },
    {
      "epoch": 0.000636,
      "grad_norm": 2.0988857746124268,
      "learning_rate": 1e-05,
      "loss": 0.3867,
      "step": 63600
    },
    {
      "epoch": 0.000637,
      "grad_norm": 2.140925407409668,
      "learning_rate": 1e-05,
      "loss": 0.3878,
      "step": 63700
    },
    {
      "epoch": 0.000638,
      "grad_norm": 2.2293543815612793,
      "learning_rate": 1e-05,
      "loss": 0.3866,
      "step": 63800
    },
    {
      "epoch": 0.000639,
      "grad_norm": 2.0480923652648926,
      "learning_rate": 1e-05,
      "loss": 0.3819,
      "step": 63900
    },
    {
      "epoch": 0.00064,
      "grad_norm": 2.129159927368164,
      "learning_rate": 1e-05,
      "loss": 0.3874,
      "step": 64000
    },
    {
      "epoch": 0.000641,
      "grad_norm": 1.907259225845337,
      "learning_rate": 1e-05,
      "loss": 0.3918,
      "step": 64100
    },
    {
      "epoch": 0.000642,
      "grad_norm": 1.9210904836654663,
      "learning_rate": 1e-05,
      "loss": 0.3863,
      "step": 64200
    },
    {
      "epoch": 0.000643,
      "grad_norm": 2.2560818195343018,
      "learning_rate": 1e-05,
      "loss": 0.3986,
      "step": 64300
    },
    {
      "epoch": 0.000644,
      "grad_norm": 1.8635262250900269,
      "learning_rate": 1e-05,
      "loss": 0.3781,
      "step": 64400
    },
    {
      "epoch": 0.000645,
      "grad_norm": 2.076395273208618,
      "learning_rate": 1e-05,
      "loss": 0.3822,
      "step": 64500
    },
    {
      "epoch": 0.000646,
      "grad_norm": 1.7710347175598145,
      "learning_rate": 1e-05,
      "loss": 0.3794,
      "step": 64600
    },
    {
      "epoch": 0.000647,
      "grad_norm": 2.0143582820892334,
      "learning_rate": 1e-05,
      "loss": 0.3825,
      "step": 64700
    },
    {
      "epoch": 0.000648,
      "grad_norm": 2.2155025005340576,
      "learning_rate": 1e-05,
      "loss": 0.3938,
      "step": 64800
    },
    {
      "epoch": 0.000649,
      "grad_norm": 1.8567825555801392,
      "learning_rate": 1e-05,
      "loss": 0.3799,
      "step": 64900
    },
    {
      "epoch": 0.00065,
      "grad_norm": 2.183415412902832,
      "learning_rate": 1e-05,
      "loss": 0.3824,
      "step": 65000
    },
    {
      "epoch": 0.000651,
      "grad_norm": 2.005911350250244,
      "learning_rate": 1e-05,
      "loss": 0.3931,
      "step": 65100
    },
    {
      "epoch": 0.000652,
      "grad_norm": 1.8332974910736084,
      "learning_rate": 1e-05,
      "loss": 0.3811,
      "step": 65200
    },
    {
      "epoch": 0.000653,
      "grad_norm": 2.1285884380340576,
      "learning_rate": 1e-05,
      "loss": 0.3918,
      "step": 65300
    },
    {
      "epoch": 0.000654,
      "grad_norm": 2.158264636993408,
      "learning_rate": 1e-05,
      "loss": 0.3863,
      "step": 65400
    },
    {
      "epoch": 0.000655,
      "grad_norm": 1.8876869678497314,
      "learning_rate": 1e-05,
      "loss": 0.3934,
      "step": 65500
    },
    {
      "epoch": 0.000656,
      "grad_norm": 1.8769333362579346,
      "learning_rate": 1e-05,
      "loss": 0.3923,
      "step": 65600
    },
    {
      "epoch": 0.000657,
      "grad_norm": 2.019409656524658,
      "learning_rate": 1e-05,
      "loss": 0.3826,
      "step": 65700
    },
    {
      "epoch": 0.000658,
      "grad_norm": 2.0213446617126465,
      "learning_rate": 1e-05,
      "loss": 0.3922,
      "step": 65800
    },
    {
      "epoch": 0.000659,
      "grad_norm": 2.2089147567749023,
      "learning_rate": 1e-05,
      "loss": 0.3933,
      "step": 65900
    },
    {
      "epoch": 0.00066,
      "grad_norm": 1.837319254875183,
      "learning_rate": 1e-05,
      "loss": 0.3868,
      "step": 66000
    },
    {
      "epoch": 0.000661,
      "grad_norm": 1.8547362089157104,
      "learning_rate": 1e-05,
      "loss": 0.3887,
      "step": 66100
    },
    {
      "epoch": 0.000662,
      "grad_norm": 1.8269295692443848,
      "learning_rate": 1e-05,
      "loss": 0.384,
      "step": 66200
    },
    {
      "epoch": 0.000663,
      "grad_norm": 2.052025318145752,
      "learning_rate": 1e-05,
      "loss": 0.3838,
      "step": 66300
    },
    {
      "epoch": 0.000664,
      "grad_norm": 2.030297040939331,
      "learning_rate": 1e-05,
      "loss": 0.3873,
      "step": 66400
    },
    {
      "epoch": 0.000665,
      "grad_norm": 2.019329309463501,
      "learning_rate": 1e-05,
      "loss": 0.3828,
      "step": 66500
    },
    {
      "epoch": 0.000666,
      "grad_norm": 1.8459995985031128,
      "learning_rate": 1e-05,
      "loss": 0.3879,
      "step": 66600
    },
    {
      "epoch": 0.000667,
      "grad_norm": 1.9611304998397827,
      "learning_rate": 1e-05,
      "loss": 0.3893,
      "step": 66700
    },
    {
      "epoch": 0.000668,
      "grad_norm": 1.8976935148239136,
      "learning_rate": 1e-05,
      "loss": 0.3858,
      "step": 66800
    },
    {
      "epoch": 0.000669,
      "grad_norm": 1.9818809032440186,
      "learning_rate": 1e-05,
      "loss": 0.3838,
      "step": 66900
    },
    {
      "epoch": 0.00067,
      "grad_norm": 1.7839868068695068,
      "learning_rate": 1e-05,
      "loss": 0.3809,
      "step": 67000
    },
    {
      "epoch": 0.000671,
      "grad_norm": 2.1452698707580566,
      "learning_rate": 1e-05,
      "loss": 0.3843,
      "step": 67100
    },
    {
      "epoch": 0.000672,
      "grad_norm": 2.077277660369873,
      "learning_rate": 1e-05,
      "loss": 0.3837,
      "step": 67200
    },
    {
      "epoch": 0.000673,
      "grad_norm": 2.0192837715148926,
      "learning_rate": 1e-05,
      "loss": 0.3974,
      "step": 67300
    },
    {
      "epoch": 0.000674,
      "grad_norm": 2.134225606918335,
      "learning_rate": 1e-05,
      "loss": 0.3777,
      "step": 67400
    },
    {
      "epoch": 0.000675,
      "grad_norm": 2.0650904178619385,
      "learning_rate": 1e-05,
      "loss": 0.3837,
      "step": 67500
    },
    {
      "epoch": 0.000676,
      "grad_norm": 1.857259750366211,
      "learning_rate": 1e-05,
      "loss": 0.3777,
      "step": 67600
    },
    {
      "epoch": 0.000677,
      "grad_norm": 1.8561601638793945,
      "learning_rate": 1e-05,
      "loss": 0.3854,
      "step": 67700
    },
    {
      "epoch": 0.000678,
      "grad_norm": 2.27827525138855,
      "learning_rate": 1e-05,
      "loss": 0.3772,
      "step": 67800
    },
    {
      "epoch": 0.000679,
      "grad_norm": 1.9206945896148682,
      "learning_rate": 1e-05,
      "loss": 0.3882,
      "step": 67900
    },
    {
      "epoch": 0.00068,
      "grad_norm": 1.849585771560669,
      "learning_rate": 1e-05,
      "loss": 0.3903,
      "step": 68000
    },
    {
      "epoch": 0.000681,
      "grad_norm": 2.1679646968841553,
      "learning_rate": 1e-05,
      "loss": 0.3844,
      "step": 68100
    },
    {
      "epoch": 0.000682,
      "grad_norm": 2.105186700820923,
      "learning_rate": 1e-05,
      "loss": 0.3843,
      "step": 68200
    },
    {
      "epoch": 0.000683,
      "grad_norm": 1.8043280839920044,
      "learning_rate": 1e-05,
      "loss": 0.3782,
      "step": 68300
    },
    {
      "epoch": 0.000684,
      "grad_norm": 1.9731149673461914,
      "learning_rate": 1e-05,
      "loss": 0.3876,
      "step": 68400
    },
    {
      "epoch": 0.000685,
      "grad_norm": 1.9924131631851196,
      "learning_rate": 1e-05,
      "loss": 0.386,
      "step": 68500
    },
    {
      "epoch": 0.000686,
      "grad_norm": 1.8679152727127075,
      "learning_rate": 1e-05,
      "loss": 0.3797,
      "step": 68600
    },
    {
      "epoch": 0.000687,
      "grad_norm": 2.01244854927063,
      "learning_rate": 1e-05,
      "loss": 0.3803,
      "step": 68700
    },
    {
      "epoch": 0.000688,
      "grad_norm": 1.9184852838516235,
      "learning_rate": 1e-05,
      "loss": 0.3872,
      "step": 68800
    },
    {
      "epoch": 0.000689,
      "grad_norm": 2.039447546005249,
      "learning_rate": 1e-05,
      "loss": 0.3808,
      "step": 68900
    },
    {
      "epoch": 0.00069,
      "grad_norm": 2.367798089981079,
      "learning_rate": 1e-05,
      "loss": 0.3858,
      "step": 69000
    },
    {
      "epoch": 0.000691,
      "grad_norm": 2.0003209114074707,
      "learning_rate": 1e-05,
      "loss": 0.3808,
      "step": 69100
    },
    {
      "epoch": 0.000692,
      "grad_norm": 1.9453091621398926,
      "learning_rate": 1e-05,
      "loss": 0.3707,
      "step": 69200
    },
    {
      "epoch": 0.000693,
      "grad_norm": 1.6954456567764282,
      "learning_rate": 1e-05,
      "loss": 0.3829,
      "step": 69300
    },
    {
      "epoch": 0.000694,
      "grad_norm": 2.1012470722198486,
      "learning_rate": 1e-05,
      "loss": 0.3833,
      "step": 69400
    },
    {
      "epoch": 0.000695,
      "grad_norm": 1.8490900993347168,
      "learning_rate": 1e-05,
      "loss": 0.3873,
      "step": 69500
    },
    {
      "epoch": 0.000696,
      "grad_norm": 1.8682618141174316,
      "learning_rate": 1e-05,
      "loss": 0.3862,
      "step": 69600
    },
    {
      "epoch": 0.000697,
      "grad_norm": 2.068352460861206,
      "learning_rate": 1e-05,
      "loss": 0.3802,
      "step": 69700
    },
    {
      "epoch": 0.000698,
      "grad_norm": 2.118117094039917,
      "learning_rate": 1e-05,
      "loss": 0.3764,
      "step": 69800
    },
    {
      "epoch": 0.000699,
      "grad_norm": 1.8571758270263672,
      "learning_rate": 1e-05,
      "loss": 0.378,
      "step": 69900
    },
    {
      "epoch": 0.0007,
      "grad_norm": 2.103874921798706,
      "learning_rate": 1e-05,
      "loss": 0.3798,
      "step": 70000
    },
    {
      "epoch": 0.000701,
      "grad_norm": 2.4420368671417236,
      "learning_rate": 1e-05,
      "loss": 0.3796,
      "step": 70100
    },
    {
      "epoch": 0.000702,
      "grad_norm": 2.143949270248413,
      "learning_rate": 1e-05,
      "loss": 0.3735,
      "step": 70200
    },
    {
      "epoch": 0.000703,
      "grad_norm": 2.070586681365967,
      "learning_rate": 1e-05,
      "loss": 0.3813,
      "step": 70300
    },
    {
      "epoch": 0.000704,
      "grad_norm": 2.0714941024780273,
      "learning_rate": 1e-05,
      "loss": 0.383,
      "step": 70400
    },
    {
      "epoch": 0.000705,
      "grad_norm": 2.0592539310455322,
      "learning_rate": 1e-05,
      "loss": 0.3769,
      "step": 70500
    },
    {
      "epoch": 0.000706,
      "grad_norm": 2.0504090785980225,
      "learning_rate": 1e-05,
      "loss": 0.3791,
      "step": 70600
    },
    {
      "epoch": 0.000707,
      "grad_norm": 1.6406168937683105,
      "learning_rate": 1e-05,
      "loss": 0.3758,
      "step": 70700
    },
    {
      "epoch": 0.000708,
      "grad_norm": 2.1220123767852783,
      "learning_rate": 1e-05,
      "loss": 0.3868,
      "step": 70800
    },
    {
      "epoch": 0.000709,
      "grad_norm": 2.0536298751831055,
      "learning_rate": 1e-05,
      "loss": 0.3805,
      "step": 70900
    },
    {
      "epoch": 0.00071,
      "grad_norm": 2.076979875564575,
      "learning_rate": 1e-05,
      "loss": 0.3807,
      "step": 71000
    },
    {
      "epoch": 0.000711,
      "grad_norm": 2.6225621700286865,
      "learning_rate": 1e-05,
      "loss": 0.373,
      "step": 71100
    },
    {
      "epoch": 0.000712,
      "grad_norm": 2.2727653980255127,
      "learning_rate": 1e-05,
      "loss": 0.3762,
      "step": 71200
    },
    {
      "epoch": 0.000713,
      "grad_norm": 2.0625195503234863,
      "learning_rate": 1e-05,
      "loss": 0.3841,
      "step": 71300
    },
    {
      "epoch": 0.000714,
      "grad_norm": 1.9859055280685425,
      "learning_rate": 1e-05,
      "loss": 0.3801,
      "step": 71400
    },
    {
      "epoch": 0.000715,
      "grad_norm": 1.9635552167892456,
      "learning_rate": 1e-05,
      "loss": 0.3848,
      "step": 71500
    },
    {
      "epoch": 0.000716,
      "grad_norm": 2.121825933456421,
      "learning_rate": 1e-05,
      "loss": 0.3745,
      "step": 71600
    },
    {
      "epoch": 0.000717,
      "grad_norm": 1.9133636951446533,
      "learning_rate": 1e-05,
      "loss": 0.3814,
      "step": 71700
    },
    {
      "epoch": 0.000718,
      "grad_norm": 2.1131491661071777,
      "learning_rate": 1e-05,
      "loss": 0.37,
      "step": 71800
    },
    {
      "epoch": 0.000719,
      "grad_norm": 2.0350754261016846,
      "learning_rate": 1e-05,
      "loss": 0.3814,
      "step": 71900
    },
    {
      "epoch": 0.00072,
      "grad_norm": 2.757786750793457,
      "learning_rate": 1e-05,
      "loss": 0.3754,
      "step": 72000
    },
    {
      "epoch": 0.000721,
      "grad_norm": 1.797782063484192,
      "learning_rate": 1e-05,
      "loss": 0.3712,
      "step": 72100
    },
    {
      "epoch": 0.000722,
      "grad_norm": 2.0632424354553223,
      "learning_rate": 1e-05,
      "loss": 0.3881,
      "step": 72200
    },
    {
      "epoch": 0.000723,
      "grad_norm": 1.7604708671569824,
      "learning_rate": 1e-05,
      "loss": 0.3787,
      "step": 72300
    },
    {
      "epoch": 0.000724,
      "grad_norm": 3.3791792392730713,
      "learning_rate": 1e-05,
      "loss": 0.3776,
      "step": 72400
    },
    {
      "epoch": 0.000725,
      "grad_norm": 2.1998651027679443,
      "learning_rate": 1e-05,
      "loss": 0.3701,
      "step": 72500
    },
    {
      "epoch": 0.000726,
      "grad_norm": 2.309633731842041,
      "learning_rate": 1e-05,
      "loss": 0.3742,
      "step": 72600
    },
    {
      "epoch": 0.000727,
      "grad_norm": 2.0794286727905273,
      "learning_rate": 1e-05,
      "loss": 0.3752,
      "step": 72700
    },
    {
      "epoch": 0.000728,
      "grad_norm": 1.98604154586792,
      "learning_rate": 1e-05,
      "loss": 0.3841,
      "step": 72800
    },
    {
      "epoch": 0.000729,
      "grad_norm": 2.0682222843170166,
      "learning_rate": 1e-05,
      "loss": 0.376,
      "step": 72900
    },
    {
      "epoch": 0.00073,
      "grad_norm": 1.9491254091262817,
      "learning_rate": 1e-05,
      "loss": 0.3704,
      "step": 73000
    },
    {
      "epoch": 0.000731,
      "grad_norm": 1.809173822402954,
      "learning_rate": 1e-05,
      "loss": 0.376,
      "step": 73100
    },
    {
      "epoch": 0.000732,
      "grad_norm": 1.7224321365356445,
      "learning_rate": 1e-05,
      "loss": 0.3737,
      "step": 73200
    },
    {
      "epoch": 0.000733,
      "grad_norm": 1.7145380973815918,
      "learning_rate": 1e-05,
      "loss": 0.3806,
      "step": 73300
    },
    {
      "epoch": 0.000734,
      "grad_norm": 2.0233635902404785,
      "learning_rate": 1e-05,
      "loss": 0.3741,
      "step": 73400
    },
    {
      "epoch": 0.000735,
      "grad_norm": 1.9742248058319092,
      "learning_rate": 1e-05,
      "loss": 0.3645,
      "step": 73500
    },
    {
      "epoch": 0.000736,
      "grad_norm": 1.889393925666809,
      "learning_rate": 1e-05,
      "loss": 0.3696,
      "step": 73600
    },
    {
      "epoch": 0.000737,
      "grad_norm": 2.075669050216675,
      "learning_rate": 1e-05,
      "loss": 0.3735,
      "step": 73700
    },
    {
      "epoch": 0.000738,
      "grad_norm": 2.1420507431030273,
      "learning_rate": 1e-05,
      "loss": 0.3701,
      "step": 73800
    },
    {
      "epoch": 0.000739,
      "grad_norm": 2.1469383239746094,
      "learning_rate": 1e-05,
      "loss": 0.379,
      "step": 73900
    },
    {
      "epoch": 0.00074,
      "grad_norm": 2.0224719047546387,
      "learning_rate": 1e-05,
      "loss": 0.372,
      "step": 74000
    },
    {
      "epoch": 0.000741,
      "grad_norm": 1.8598190546035767,
      "learning_rate": 1e-05,
      "loss": 0.3792,
      "step": 74100
    },
    {
      "epoch": 0.000742,
      "grad_norm": 2.1243066787719727,
      "learning_rate": 1e-05,
      "loss": 0.3689,
      "step": 74200
    },
    {
      "epoch": 0.000743,
      "grad_norm": 1.8850631713867188,
      "learning_rate": 1e-05,
      "loss": 0.3762,
      "step": 74300
    },
    {
      "epoch": 0.000744,
      "grad_norm": 2.0598785877227783,
      "learning_rate": 1e-05,
      "loss": 0.3715,
      "step": 74400
    },
    {
      "epoch": 0.000745,
      "grad_norm": 2.120824098587036,
      "learning_rate": 1e-05,
      "loss": 0.3696,
      "step": 74500
    },
    {
      "epoch": 0.000746,
      "grad_norm": 1.7642192840576172,
      "learning_rate": 1e-05,
      "loss": 0.3771,
      "step": 74600
    },
    {
      "epoch": 0.000747,
      "grad_norm": 2.1491034030914307,
      "learning_rate": 1e-05,
      "loss": 0.3883,
      "step": 74700
    },
    {
      "epoch": 0.000748,
      "grad_norm": 1.8905261754989624,
      "learning_rate": 1e-05,
      "loss": 0.3756,
      "step": 74800
    },
    {
      "epoch": 0.000749,
      "grad_norm": 2.4035165309906006,
      "learning_rate": 1e-05,
      "loss": 0.3776,
      "step": 74900
    },
    {
      "epoch": 0.00075,
      "grad_norm": 2.104729652404785,
      "learning_rate": 1e-05,
      "loss": 0.3762,
      "step": 75000
    },
    {
      "epoch": 0.000751,
      "grad_norm": 2.0208077430725098,
      "learning_rate": 1e-05,
      "loss": 0.3696,
      "step": 75100
    },
    {
      "epoch": 0.000752,
      "grad_norm": 2.1069564819335938,
      "learning_rate": 1e-05,
      "loss": 0.3766,
      "step": 75200
    },
    {
      "epoch": 0.000753,
      "grad_norm": 1.9399853944778442,
      "learning_rate": 1e-05,
      "loss": 0.3767,
      "step": 75300
    },
    {
      "epoch": 0.000754,
      "grad_norm": 2.072504758834839,
      "learning_rate": 1e-05,
      "loss": 0.3787,
      "step": 75400
    },
    {
      "epoch": 0.000755,
      "grad_norm": 1.6709392070770264,
      "learning_rate": 1e-05,
      "loss": 0.3739,
      "step": 75500
    },
    {
      "epoch": 0.000756,
      "grad_norm": 1.9638621807098389,
      "learning_rate": 1e-05,
      "loss": 0.3736,
      "step": 75600
    },
    {
      "epoch": 0.000757,
      "grad_norm": 1.9661500453948975,
      "learning_rate": 1e-05,
      "loss": 0.3742,
      "step": 75700
    },
    {
      "epoch": 0.000758,
      "grad_norm": 1.7618420124053955,
      "learning_rate": 1e-05,
      "loss": 0.3725,
      "step": 75800
    },
    {
      "epoch": 0.000759,
      "grad_norm": 2.090238332748413,
      "learning_rate": 1e-05,
      "loss": 0.3771,
      "step": 75900
    },
    {
      "epoch": 0.00076,
      "grad_norm": 1.9758362770080566,
      "learning_rate": 1e-05,
      "loss": 0.3666,
      "step": 76000
    },
    {
      "epoch": 0.000761,
      "grad_norm": 2.023850917816162,
      "learning_rate": 1e-05,
      "loss": 0.379,
      "step": 76100
    },
    {
      "epoch": 0.000762,
      "grad_norm": 1.9351980686187744,
      "learning_rate": 1e-05,
      "loss": 0.3705,
      "step": 76200
    },
    {
      "epoch": 0.000763,
      "grad_norm": 1.8853991031646729,
      "learning_rate": 1e-05,
      "loss": 0.3743,
      "step": 76300
    },
    {
      "epoch": 0.000764,
      "grad_norm": 2.0006158351898193,
      "learning_rate": 1e-05,
      "loss": 0.3704,
      "step": 76400
    },
    {
      "epoch": 0.000765,
      "grad_norm": 2.1643075942993164,
      "learning_rate": 1e-05,
      "loss": 0.3705,
      "step": 76500
    },
    {
      "epoch": 0.000766,
      "grad_norm": 1.9577935934066772,
      "learning_rate": 1e-05,
      "loss": 0.3742,
      "step": 76600
    },
    {
      "epoch": 0.000767,
      "grad_norm": 2.128188371658325,
      "learning_rate": 1e-05,
      "loss": 0.3698,
      "step": 76700
    },
    {
      "epoch": 0.000768,
      "grad_norm": 1.9895089864730835,
      "learning_rate": 1e-05,
      "loss": 0.3753,
      "step": 76800
    },
    {
      "epoch": 0.000769,
      "grad_norm": 2.1536855697631836,
      "learning_rate": 1e-05,
      "loss": 0.3644,
      "step": 76900
    },
    {
      "epoch": 0.00077,
      "grad_norm": 1.9444348812103271,
      "learning_rate": 1e-05,
      "loss": 0.3671,
      "step": 77000
    },
    {
      "epoch": 0.000771,
      "grad_norm": 1.8287049531936646,
      "learning_rate": 1e-05,
      "loss": 0.3735,
      "step": 77100
    },
    {
      "epoch": 0.000772,
      "grad_norm": 1.8443069458007812,
      "learning_rate": 1e-05,
      "loss": 0.3686,
      "step": 77200
    },
    {
      "epoch": 0.000773,
      "grad_norm": 1.8012452125549316,
      "learning_rate": 1e-05,
      "loss": 0.3751,
      "step": 77300
    },
    {
      "epoch": 0.000774,
      "grad_norm": 1.9977177381515503,
      "learning_rate": 1e-05,
      "loss": 0.3684,
      "step": 77400
    },
    {
      "epoch": 0.000775,
      "grad_norm": 1.9906736612319946,
      "learning_rate": 1e-05,
      "loss": 0.3712,
      "step": 77500
    },
    {
      "epoch": 0.000776,
      "grad_norm": 1.9918975830078125,
      "learning_rate": 1e-05,
      "loss": 0.3687,
      "step": 77600
    },
    {
      "epoch": 0.000777,
      "grad_norm": 1.9965052604675293,
      "learning_rate": 1e-05,
      "loss": 0.3668,
      "step": 77700
    },
    {
      "epoch": 0.000778,
      "grad_norm": 1.9064897298812866,
      "learning_rate": 1e-05,
      "loss": 0.3719,
      "step": 77800
    },
    {
      "epoch": 0.000779,
      "grad_norm": 1.7971402406692505,
      "learning_rate": 1e-05,
      "loss": 0.3645,
      "step": 77900
    },
    {
      "epoch": 0.00078,
      "grad_norm": 1.8232814073562622,
      "learning_rate": 1e-05,
      "loss": 0.3705,
      "step": 78000
    },
    {
      "epoch": 0.000781,
      "grad_norm": 2.1162238121032715,
      "learning_rate": 1e-05,
      "loss": 0.3675,
      "step": 78100
    },
    {
      "epoch": 0.000782,
      "grad_norm": 1.8850531578063965,
      "learning_rate": 1e-05,
      "loss": 0.3668,
      "step": 78200
    },
    {
      "epoch": 0.000783,
      "grad_norm": 1.864730715751648,
      "learning_rate": 1e-05,
      "loss": 0.3652,
      "step": 78300
    },
    {
      "epoch": 0.000784,
      "grad_norm": 1.9205899238586426,
      "learning_rate": 1e-05,
      "loss": 0.3716,
      "step": 78400
    },
    {
      "epoch": 0.000785,
      "grad_norm": 2.325000524520874,
      "learning_rate": 1e-05,
      "loss": 0.3738,
      "step": 78500
    },
    {
      "epoch": 0.000786,
      "grad_norm": 2.2757534980773926,
      "learning_rate": 1e-05,
      "loss": 0.3722,
      "step": 78600
    },
    {
      "epoch": 0.000787,
      "grad_norm": 1.7619765996932983,
      "learning_rate": 1e-05,
      "loss": 0.3799,
      "step": 78700
    },
    {
      "epoch": 0.000788,
      "grad_norm": 1.802307367324829,
      "learning_rate": 1e-05,
      "loss": 0.3744,
      "step": 78800
    },
    {
      "epoch": 0.000789,
      "grad_norm": 1.8677384853363037,
      "learning_rate": 1e-05,
      "loss": 0.3757,
      "step": 78900
    },
    {
      "epoch": 0.00079,
      "grad_norm": 2.1615066528320312,
      "learning_rate": 1e-05,
      "loss": 0.3769,
      "step": 79000
    },
    {
      "epoch": 0.000791,
      "grad_norm": 1.6998400688171387,
      "learning_rate": 1e-05,
      "loss": 0.3705,
      "step": 79100
    },
    {
      "epoch": 0.000792,
      "grad_norm": 1.7555445432662964,
      "learning_rate": 1e-05,
      "loss": 0.373,
      "step": 79200
    },
    {
      "epoch": 0.000793,
      "grad_norm": 2.0142476558685303,
      "learning_rate": 1e-05,
      "loss": 0.3723,
      "step": 79300
    },
    {
      "epoch": 0.000794,
      "grad_norm": 2.3179373741149902,
      "learning_rate": 1e-05,
      "loss": 0.3683,
      "step": 79400
    },
    {
      "epoch": 0.000795,
      "grad_norm": 1.9455734491348267,
      "learning_rate": 1e-05,
      "loss": 0.3677,
      "step": 79500
    },
    {
      "epoch": 0.000796,
      "grad_norm": 2.0112357139587402,
      "learning_rate": 1e-05,
      "loss": 0.3667,
      "step": 79600
    },
    {
      "epoch": 0.000797,
      "grad_norm": 2.257429361343384,
      "learning_rate": 1e-05,
      "loss": 0.366,
      "step": 79700
    },
    {
      "epoch": 0.000798,
      "grad_norm": 1.7353073358535767,
      "learning_rate": 1e-05,
      "loss": 0.3693,
      "step": 79800
    },
    {
      "epoch": 0.000799,
      "grad_norm": 1.989250898361206,
      "learning_rate": 1e-05,
      "loss": 0.3658,
      "step": 79900
    },
    {
      "epoch": 0.0008,
      "grad_norm": 1.8010023832321167,
      "learning_rate": 1e-05,
      "loss": 0.3687,
      "step": 80000
    },
    {
      "epoch": 0.0008,
      "eval_loss": 0.343017578125,
      "eval_runtime": 111.2417,
      "eval_samples_per_second": 449.472,
      "eval_steps_per_second": 28.092,
      "step": 80000
    },
    {
      "epoch": 0.000801,
      "grad_norm": 1.8883432149887085,
      "learning_rate": 1e-05,
      "loss": 0.3739,
      "step": 80100
    },
    {
      "epoch": 0.000802,
      "grad_norm": 1.9652680158615112,
      "learning_rate": 1e-05,
      "loss": 0.359,
      "step": 80200
    },
    {
      "epoch": 0.000803,
      "grad_norm": 2.1151764392852783,
      "learning_rate": 1e-05,
      "loss": 0.3626,
      "step": 80300
    },
    {
      "epoch": 0.000804,
      "grad_norm": 1.736228108406067,
      "learning_rate": 1e-05,
      "loss": 0.3737,
      "step": 80400
    },
    {
      "epoch": 0.000805,
      "grad_norm": 1.800878643989563,
      "learning_rate": 1e-05,
      "loss": 0.3731,
      "step": 80500
    },
    {
      "epoch": 0.000806,
      "grad_norm": 1.8532053232192993,
      "learning_rate": 1e-05,
      "loss": 0.3723,
      "step": 80600
    },
    {
      "epoch": 0.000807,
      "grad_norm": 1.838671088218689,
      "learning_rate": 1e-05,
      "loss": 0.3666,
      "step": 80700
    },
    {
      "epoch": 0.000808,
      "grad_norm": 1.7083035707473755,
      "learning_rate": 1e-05,
      "loss": 0.3761,
      "step": 80800
    },
    {
      "epoch": 0.000809,
      "grad_norm": 1.8676607608795166,
      "learning_rate": 1e-05,
      "loss": 0.3684,
      "step": 80900
    },
    {
      "epoch": 0.00081,
      "grad_norm": 2.0752341747283936,
      "learning_rate": 1e-05,
      "loss": 0.3657,
      "step": 81000
    },
    {
      "epoch": 0.000811,
      "grad_norm": 1.793967604637146,
      "learning_rate": 1e-05,
      "loss": 0.3748,
      "step": 81100
    },
    {
      "epoch": 0.000812,
      "grad_norm": 1.6681337356567383,
      "learning_rate": 1e-05,
      "loss": 0.3661,
      "step": 81200
    },
    {
      "epoch": 0.000813,
      "grad_norm": 2.46724534034729,
      "learning_rate": 1e-05,
      "loss": 0.3609,
      "step": 81300
    },
    {
      "epoch": 0.000814,
      "grad_norm": 1.8258310556411743,
      "learning_rate": 1e-05,
      "loss": 0.3632,
      "step": 81400
    },
    {
      "epoch": 0.000815,
      "grad_norm": 1.9003719091415405,
      "learning_rate": 1e-05,
      "loss": 0.365,
      "step": 81500
    },
    {
      "epoch": 0.000816,
      "grad_norm": 1.8004292249679565,
      "learning_rate": 1e-05,
      "loss": 0.3675,
      "step": 81600
    },
    {
      "epoch": 0.000817,
      "grad_norm": 1.8981424570083618,
      "learning_rate": 1e-05,
      "loss": 0.369,
      "step": 81700
    },
    {
      "epoch": 0.000818,
      "grad_norm": 2.1236822605133057,
      "learning_rate": 1e-05,
      "loss": 0.368,
      "step": 81800
    },
    {
      "epoch": 0.000819,
      "grad_norm": 1.9143141508102417,
      "learning_rate": 1e-05,
      "loss": 0.3628,
      "step": 81900
    },
    {
      "epoch": 0.00082,
      "grad_norm": 1.9660463333129883,
      "learning_rate": 1e-05,
      "loss": 0.3627,
      "step": 82000
    },
    {
      "epoch": 0.000821,
      "grad_norm": 2.1098835468292236,
      "learning_rate": 1e-05,
      "loss": 0.3657,
      "step": 82100
    },
    {
      "epoch": 0.000822,
      "grad_norm": 1.8893609046936035,
      "learning_rate": 1e-05,
      "loss": 0.3621,
      "step": 82200
    },
    {
      "epoch": 0.000823,
      "grad_norm": 1.848253607749939,
      "learning_rate": 1e-05,
      "loss": 0.3547,
      "step": 82300
    },
    {
      "epoch": 0.000824,
      "grad_norm": 1.8891757726669312,
      "learning_rate": 1e-05,
      "loss": 0.3654,
      "step": 82400
    },
    {
      "epoch": 0.000825,
      "grad_norm": 1.9906656742095947,
      "learning_rate": 1e-05,
      "loss": 0.37,
      "step": 82500
    },
    {
      "epoch": 0.000826,
      "grad_norm": 2.026745080947876,
      "learning_rate": 1e-05,
      "loss": 0.3598,
      "step": 82600
    },
    {
      "epoch": 0.000827,
      "grad_norm": 1.8796215057373047,
      "learning_rate": 1e-05,
      "loss": 0.3608,
      "step": 82700
    },
    {
      "epoch": 0.000828,
      "grad_norm": 1.8934880495071411,
      "learning_rate": 1e-05,
      "loss": 0.3633,
      "step": 82800
    },
    {
      "epoch": 0.000829,
      "grad_norm": 2.241187572479248,
      "learning_rate": 1e-05,
      "loss": 0.3623,
      "step": 82900
    },
    {
      "epoch": 0.00083,
      "grad_norm": 1.8311808109283447,
      "learning_rate": 1e-05,
      "loss": 0.3643,
      "step": 83000
    },
    {
      "epoch": 0.000831,
      "grad_norm": 1.7869751453399658,
      "learning_rate": 1e-05,
      "loss": 0.3718,
      "step": 83100
    },
    {
      "epoch": 0.000832,
      "grad_norm": 1.894146203994751,
      "learning_rate": 1e-05,
      "loss": 0.3629,
      "step": 83200
    },
    {
      "epoch": 0.000833,
      "grad_norm": 1.7418984174728394,
      "learning_rate": 1e-05,
      "loss": 0.358,
      "step": 83300
    },
    {
      "epoch": 0.000834,
      "grad_norm": 2.2200584411621094,
      "learning_rate": 1e-05,
      "loss": 0.3602,
      "step": 83400
    },
    {
      "epoch": 0.000835,
      "grad_norm": 1.7402255535125732,
      "learning_rate": 1e-05,
      "loss": 0.3648,
      "step": 83500
    },
    {
      "epoch": 0.000836,
      "grad_norm": 1.7476297616958618,
      "learning_rate": 1e-05,
      "loss": 0.3603,
      "step": 83600
    },
    {
      "epoch": 0.000837,
      "grad_norm": 2.0509250164031982,
      "learning_rate": 1e-05,
      "loss": 0.3698,
      "step": 83700
    },
    {
      "epoch": 0.000838,
      "grad_norm": 1.819290041923523,
      "learning_rate": 1e-05,
      "loss": 0.349,
      "step": 83800
    },
    {
      "epoch": 0.000839,
      "grad_norm": 1.9946727752685547,
      "learning_rate": 1e-05,
      "loss": 0.3666,
      "step": 83900
    },
    {
      "epoch": 0.00084,
      "grad_norm": 1.6956796646118164,
      "learning_rate": 1e-05,
      "loss": 0.3686,
      "step": 84000
    },
    {
      "epoch": 0.000841,
      "grad_norm": 1.8575202226638794,
      "learning_rate": 1e-05,
      "loss": 0.3666,
      "step": 84100
    },
    {
      "epoch": 0.000842,
      "grad_norm": 1.7518588304519653,
      "learning_rate": 1e-05,
      "loss": 0.3624,
      "step": 84200
    },
    {
      "epoch": 0.000843,
      "grad_norm": 1.6281752586364746,
      "learning_rate": 1e-05,
      "loss": 0.3567,
      "step": 84300
    },
    {
      "epoch": 0.000844,
      "grad_norm": 1.8025518655776978,
      "learning_rate": 1e-05,
      "loss": 0.355,
      "step": 84400
    },
    {
      "epoch": 0.000845,
      "grad_norm": 1.787426471710205,
      "learning_rate": 1e-05,
      "loss": 0.3666,
      "step": 84500
    },
    {
      "epoch": 0.000846,
      "grad_norm": 1.8636668920516968,
      "learning_rate": 1e-05,
      "loss": 0.3539,
      "step": 84600
    },
    {
      "epoch": 0.000847,
      "grad_norm": 1.999342441558838,
      "learning_rate": 1e-05,
      "loss": 0.3573,
      "step": 84700
    },
    {
      "epoch": 0.000848,
      "grad_norm": 1.7526439428329468,
      "learning_rate": 1e-05,
      "loss": 0.3638,
      "step": 84800
    },
    {
      "epoch": 0.000849,
      "grad_norm": 1.9818848371505737,
      "learning_rate": 1e-05,
      "loss": 0.3622,
      "step": 84900
    },
    {
      "epoch": 0.00085,
      "grad_norm": 2.2633955478668213,
      "learning_rate": 1e-05,
      "loss": 0.3685,
      "step": 85000
    },
    {
      "epoch": 0.000851,
      "grad_norm": 2.037205696105957,
      "learning_rate": 1e-05,
      "loss": 0.3607,
      "step": 85100
    },
    {
      "epoch": 0.000852,
      "grad_norm": 2.008530616760254,
      "learning_rate": 1e-05,
      "loss": 0.3666,
      "step": 85200
    },
    {
      "epoch": 0.000853,
      "grad_norm": 1.7828373908996582,
      "learning_rate": 1e-05,
      "loss": 0.3642,
      "step": 85300
    },
    {
      "epoch": 0.000854,
      "grad_norm": 2.0201706886291504,
      "learning_rate": 1e-05,
      "loss": 0.3597,
      "step": 85400
    },
    {
      "epoch": 0.000855,
      "grad_norm": 2.0106265544891357,
      "learning_rate": 1e-05,
      "loss": 0.3604,
      "step": 85500
    },
    {
      "epoch": 0.000856,
      "grad_norm": 1.896898627281189,
      "learning_rate": 1e-05,
      "loss": 0.366,
      "step": 85600
    },
    {
      "epoch": 0.000857,
      "grad_norm": 1.9812458753585815,
      "learning_rate": 1e-05,
      "loss": 0.3596,
      "step": 85700
    },
    {
      "epoch": 0.000858,
      "grad_norm": 2.0447208881378174,
      "learning_rate": 1e-05,
      "loss": 0.3615,
      "step": 85800
    },
    {
      "epoch": 0.000859,
      "grad_norm": 1.9249247312545776,
      "learning_rate": 1e-05,
      "loss": 0.3532,
      "step": 85900
    },
    {
      "epoch": 0.00086,
      "grad_norm": 1.700594186782837,
      "learning_rate": 1e-05,
      "loss": 0.3567,
      "step": 86000
    },
    {
      "epoch": 0.000861,
      "grad_norm": 1.9149887561798096,
      "learning_rate": 1e-05,
      "loss": 0.3585,
      "step": 86100
    },
    {
      "epoch": 0.000862,
      "grad_norm": 2.202561855316162,
      "learning_rate": 1e-05,
      "loss": 0.3667,
      "step": 86200
    },
    {
      "epoch": 0.000863,
      "grad_norm": 1.8370987176895142,
      "learning_rate": 1e-05,
      "loss": 0.3606,
      "step": 86300
    },
    {
      "epoch": 0.000864,
      "grad_norm": 2.0995025634765625,
      "learning_rate": 1e-05,
      "loss": 0.366,
      "step": 86400
    },
    {
      "epoch": 0.000865,
      "grad_norm": 1.7918909788131714,
      "learning_rate": 1e-05,
      "loss": 0.3524,
      "step": 86500
    },
    {
      "epoch": 0.000866,
      "grad_norm": 1.870877742767334,
      "learning_rate": 1e-05,
      "loss": 0.3631,
      "step": 86600
    },
    {
      "epoch": 0.000867,
      "grad_norm": 2.0287795066833496,
      "learning_rate": 1e-05,
      "loss": 0.3555,
      "step": 86700
    },
    {
      "epoch": 0.000868,
      "grad_norm": 1.9686987400054932,
      "learning_rate": 1e-05,
      "loss": 0.3622,
      "step": 86800
    },
    {
      "epoch": 0.000869,
      "grad_norm": 1.714966893196106,
      "learning_rate": 1e-05,
      "loss": 0.3585,
      "step": 86900
    },
    {
      "epoch": 0.00087,
      "grad_norm": 2.0388360023498535,
      "learning_rate": 1e-05,
      "loss": 0.3605,
      "step": 87000
    },
    {
      "epoch": 0.000871,
      "grad_norm": 1.8588838577270508,
      "learning_rate": 1e-05,
      "loss": 0.3588,
      "step": 87100
    },
    {
      "epoch": 0.000872,
      "grad_norm": 2.2491447925567627,
      "learning_rate": 1e-05,
      "loss": 0.3598,
      "step": 87200
    },
    {
      "epoch": 0.000873,
      "grad_norm": 2.023857831954956,
      "learning_rate": 1e-05,
      "loss": 0.3618,
      "step": 87300
    },
    {
      "epoch": 0.000874,
      "grad_norm": 1.5411655902862549,
      "learning_rate": 1e-05,
      "loss": 0.3681,
      "step": 87400
    },
    {
      "epoch": 0.000875,
      "grad_norm": 1.5299054384231567,
      "learning_rate": 1e-05,
      "loss": 0.3594,
      "step": 87500
    },
    {
      "epoch": 0.000876,
      "grad_norm": 1.9162421226501465,
      "learning_rate": 1e-05,
      "loss": 0.3569,
      "step": 87600
    },
    {
      "epoch": 0.000877,
      "grad_norm": 2.160090684890747,
      "learning_rate": 1e-05,
      "loss": 0.358,
      "step": 87700
    },
    {
      "epoch": 0.000878,
      "grad_norm": 2.044666051864624,
      "learning_rate": 1e-05,
      "loss": 0.3609,
      "step": 87800
    },
    {
      "epoch": 0.000879,
      "grad_norm": 1.7112947702407837,
      "learning_rate": 1e-05,
      "loss": 0.3544,
      "step": 87900
    },
    {
      "epoch": 0.00088,
      "grad_norm": 1.8648561239242554,
      "learning_rate": 1e-05,
      "loss": 0.365,
      "step": 88000
    },
    {
      "epoch": 0.000881,
      "grad_norm": 1.8748390674591064,
      "learning_rate": 1e-05,
      "loss": 0.3668,
      "step": 88100
    },
    {
      "epoch": 0.000882,
      "grad_norm": 2.2753427028656006,
      "learning_rate": 1e-05,
      "loss": 0.3631,
      "step": 88200
    },
    {
      "epoch": 0.000883,
      "grad_norm": 1.8260302543640137,
      "learning_rate": 1e-05,
      "loss": 0.3614,
      "step": 88300
    },
    {
      "epoch": 0.000884,
      "grad_norm": 1.8950936794281006,
      "learning_rate": 1e-05,
      "loss": 0.3555,
      "step": 88400
    },
    {
      "epoch": 0.000885,
      "grad_norm": 1.7748656272888184,
      "learning_rate": 1e-05,
      "loss": 0.3607,
      "step": 88500
    },
    {
      "epoch": 0.000886,
      "grad_norm": 1.7580374479293823,
      "learning_rate": 1e-05,
      "loss": 0.352,
      "step": 88600
    },
    {
      "epoch": 0.000887,
      "grad_norm": 1.9820917844772339,
      "learning_rate": 1e-05,
      "loss": 0.3549,
      "step": 88700
    },
    {
      "epoch": 0.000888,
      "grad_norm": 1.842002272605896,
      "learning_rate": 1e-05,
      "loss": 0.3606,
      "step": 88800
    },
    {
      "epoch": 0.000889,
      "grad_norm": 1.8876936435699463,
      "learning_rate": 1e-05,
      "loss": 0.3607,
      "step": 88900
    },
    {
      "epoch": 0.00089,
      "grad_norm": 1.7980290651321411,
      "learning_rate": 1e-05,
      "loss": 0.3555,
      "step": 89000
    },
    {
      "epoch": 0.000891,
      "grad_norm": 2.300755739212036,
      "learning_rate": 1e-05,
      "loss": 0.3582,
      "step": 89100
    },
    {
      "epoch": 0.000892,
      "grad_norm": 2.7439591884613037,
      "learning_rate": 1e-05,
      "loss": 0.3545,
      "step": 89200
    },
    {
      "epoch": 0.000893,
      "grad_norm": 1.7525635957717896,
      "learning_rate": 1e-05,
      "loss": 0.3636,
      "step": 89300
    },
    {
      "epoch": 0.000894,
      "grad_norm": 1.6738208532333374,
      "learning_rate": 1e-05,
      "loss": 0.3569,
      "step": 89400
    },
    {
      "epoch": 0.000895,
      "grad_norm": 1.5723131895065308,
      "learning_rate": 1e-05,
      "loss": 0.3619,
      "step": 89500
    },
    {
      "epoch": 0.000896,
      "grad_norm": 1.865443468093872,
      "learning_rate": 1e-05,
      "loss": 0.3546,
      "step": 89600
    },
    {
      "epoch": 0.000897,
      "grad_norm": 1.8413885831832886,
      "learning_rate": 1e-05,
      "loss": 0.3519,
      "step": 89700
    },
    {
      "epoch": 0.000898,
      "grad_norm": 1.760122537612915,
      "learning_rate": 1e-05,
      "loss": 0.3525,
      "step": 89800
    },
    {
      "epoch": 0.000899,
      "grad_norm": 2.0341832637786865,
      "learning_rate": 1e-05,
      "loss": 0.3478,
      "step": 89900
    },
    {
      "epoch": 0.0009,
      "grad_norm": 1.8548213243484497,
      "learning_rate": 1e-05,
      "loss": 0.3486,
      "step": 90000
    },
    {
      "epoch": 0.000901,
      "grad_norm": 2.114245891571045,
      "learning_rate": 1e-05,
      "loss": 0.3494,
      "step": 90100
    },
    {
      "epoch": 0.000902,
      "grad_norm": 2.117030143737793,
      "learning_rate": 1e-05,
      "loss": 0.3514,
      "step": 90200
    },
    {
      "epoch": 0.000903,
      "grad_norm": 1.8521121740341187,
      "learning_rate": 1e-05,
      "loss": 0.3571,
      "step": 90300
    },
    {
      "epoch": 0.000904,
      "grad_norm": 1.9864593744277954,
      "learning_rate": 1e-05,
      "loss": 0.3657,
      "step": 90400
    },
    {
      "epoch": 0.000905,
      "grad_norm": 1.9219348430633545,
      "learning_rate": 1e-05,
      "loss": 0.3541,
      "step": 90500
    },
    {
      "epoch": 0.000906,
      "grad_norm": 2.13183856010437,
      "learning_rate": 1e-05,
      "loss": 0.3566,
      "step": 90600
    },
    {
      "epoch": 0.000907,
      "grad_norm": 1.7505743503570557,
      "learning_rate": 1e-05,
      "loss": 0.351,
      "step": 90700
    },
    {
      "epoch": 0.000908,
      "grad_norm": 1.7294330596923828,
      "learning_rate": 1e-05,
      "loss": 0.3536,
      "step": 90800
    },
    {
      "epoch": 0.000909,
      "grad_norm": 1.8986823558807373,
      "learning_rate": 1e-05,
      "loss": 0.3598,
      "step": 90900
    },
    {
      "epoch": 0.00091,
      "grad_norm": 1.6649383306503296,
      "learning_rate": 1e-05,
      "loss": 0.3568,
      "step": 91000
    },
    {
      "epoch": 0.000911,
      "grad_norm": 2.0748260021209717,
      "learning_rate": 1e-05,
      "loss": 0.3527,
      "step": 91100
    },
    {
      "epoch": 0.000912,
      "grad_norm": 1.905617117881775,
      "learning_rate": 1e-05,
      "loss": 0.3535,
      "step": 91200
    },
    {
      "epoch": 0.000913,
      "grad_norm": 1.764633059501648,
      "learning_rate": 1e-05,
      "loss": 0.3517,
      "step": 91300
    },
    {
      "epoch": 0.000914,
      "grad_norm": 1.822187900543213,
      "learning_rate": 1e-05,
      "loss": 0.3584,
      "step": 91400
    },
    {
      "epoch": 0.000915,
      "grad_norm": 1.845644235610962,
      "learning_rate": 1e-05,
      "loss": 0.3561,
      "step": 91500
    },
    {
      "epoch": 0.000916,
      "grad_norm": 2.082502603530884,
      "learning_rate": 1e-05,
      "loss": 0.3563,
      "step": 91600
    },
    {
      "epoch": 0.000917,
      "grad_norm": 2.198960065841675,
      "learning_rate": 1e-05,
      "loss": 0.3557,
      "step": 91700
    },
    {
      "epoch": 0.000918,
      "grad_norm": 1.6692492961883545,
      "learning_rate": 1e-05,
      "loss": 0.3558,
      "step": 91800
    },
    {
      "epoch": 0.000919,
      "grad_norm": 2.025036334991455,
      "learning_rate": 1e-05,
      "loss": 0.3427,
      "step": 91900
    },
    {
      "epoch": 0.00092,
      "grad_norm": 1.8072044849395752,
      "learning_rate": 1e-05,
      "loss": 0.3534,
      "step": 92000
    },
    {
      "epoch": 0.000921,
      "grad_norm": 1.989229679107666,
      "learning_rate": 1e-05,
      "loss": 0.3579,
      "step": 92100
    },
    {
      "epoch": 0.000922,
      "grad_norm": 1.943912386894226,
      "learning_rate": 1e-05,
      "loss": 0.3625,
      "step": 92200
    },
    {
      "epoch": 0.000923,
      "grad_norm": 5.172427654266357,
      "learning_rate": 1e-05,
      "loss": 0.3467,
      "step": 92300
    },
    {
      "epoch": 0.000924,
      "grad_norm": 1.854652762413025,
      "learning_rate": 1e-05,
      "loss": 0.3526,
      "step": 92400
    },
    {
      "epoch": 0.000925,
      "grad_norm": 1.7196903228759766,
      "learning_rate": 1e-05,
      "loss": 0.3505,
      "step": 92500
    },
    {
      "epoch": 0.000926,
      "grad_norm": 1.6658947467803955,
      "learning_rate": 1e-05,
      "loss": 0.3564,
      "step": 92600
    },
    {
      "epoch": 0.000927,
      "grad_norm": 2.8138256072998047,
      "learning_rate": 1e-05,
      "loss": 0.3571,
      "step": 92700
    },
    {
      "epoch": 0.000928,
      "grad_norm": 1.700640320777893,
      "learning_rate": 1e-05,
      "loss": 0.3576,
      "step": 92800
    },
    {
      "epoch": 0.000929,
      "grad_norm": 1.738922119140625,
      "learning_rate": 1e-05,
      "loss": 0.3482,
      "step": 92900
    },
    {
      "epoch": 0.00093,
      "grad_norm": 1.7264224290847778,
      "learning_rate": 1e-05,
      "loss": 0.3515,
      "step": 93000
    },
    {
      "epoch": 0.000931,
      "grad_norm": 1.7760035991668701,
      "learning_rate": 1e-05,
      "loss": 0.3546,
      "step": 93100
    },
    {
      "epoch": 0.000932,
      "grad_norm": 1.684767484664917,
      "learning_rate": 1e-05,
      "loss": 0.3509,
      "step": 93200
    },
    {
      "epoch": 0.000933,
      "grad_norm": 1.9357808828353882,
      "learning_rate": 1e-05,
      "loss": 0.3613,
      "step": 93300
    },
    {
      "epoch": 0.000934,
      "grad_norm": 1.853598952293396,
      "learning_rate": 1e-05,
      "loss": 0.3523,
      "step": 93400
    },
    {
      "epoch": 0.000935,
      "grad_norm": 3.272063732147217,
      "learning_rate": 1e-05,
      "loss": 0.3515,
      "step": 93500
    },
    {
      "epoch": 0.000936,
      "grad_norm": 1.8037041425704956,
      "learning_rate": 1e-05,
      "loss": 0.3526,
      "step": 93600
    },
    {
      "epoch": 0.000937,
      "grad_norm": 1.989990472793579,
      "learning_rate": 1e-05,
      "loss": 0.3512,
      "step": 93700
    },
    {
      "epoch": 0.000938,
      "grad_norm": 1.7665644884109497,
      "learning_rate": 1e-05,
      "loss": 0.3503,
      "step": 93800
    },
    {
      "epoch": 0.000939,
      "grad_norm": 2.230848550796509,
      "learning_rate": 1e-05,
      "loss": 0.3532,
      "step": 93900
    },
    {
      "epoch": 0.00094,
      "grad_norm": 1.8637299537658691,
      "learning_rate": 1e-05,
      "loss": 0.3558,
      "step": 94000
    },
    {
      "epoch": 0.000941,
      "grad_norm": 1.9153410196304321,
      "learning_rate": 1e-05,
      "loss": 0.3534,
      "step": 94100
    },
    {
      "epoch": 0.000942,
      "grad_norm": 1.9178539514541626,
      "learning_rate": 1e-05,
      "loss": 0.3578,
      "step": 94200
    },
    {
      "epoch": 0.000943,
      "grad_norm": 1.9506075382232666,
      "learning_rate": 1e-05,
      "loss": 0.3557,
      "step": 94300
    },
    {
      "epoch": 0.000944,
      "grad_norm": 1.97675621509552,
      "learning_rate": 1e-05,
      "loss": 0.3496,
      "step": 94400
    },
    {
      "epoch": 0.000945,
      "grad_norm": 1.571119785308838,
      "learning_rate": 1e-05,
      "loss": 0.3549,
      "step": 94500
    },
    {
      "epoch": 0.000946,
      "grad_norm": 1.84198796749115,
      "learning_rate": 1e-05,
      "loss": 0.3564,
      "step": 94600
    },
    {
      "epoch": 0.000947,
      "grad_norm": 1.6789623498916626,
      "learning_rate": 1e-05,
      "loss": 0.3459,
      "step": 94700
    },
    {
      "epoch": 0.000948,
      "grad_norm": 1.7345160245895386,
      "learning_rate": 1e-05,
      "loss": 0.348,
      "step": 94800
    },
    {
      "epoch": 0.000949,
      "grad_norm": 1.626235008239746,
      "learning_rate": 1e-05,
      "loss": 0.3579,
      "step": 94900
    },
    {
      "epoch": 0.00095,
      "grad_norm": 1.8632274866104126,
      "learning_rate": 1e-05,
      "loss": 0.3555,
      "step": 95000
    },
    {
      "epoch": 0.000951,
      "grad_norm": 1.5302915573120117,
      "learning_rate": 1e-05,
      "loss": 0.3515,
      "step": 95100
    },
    {
      "epoch": 0.000952,
      "grad_norm": 1.759491205215454,
      "learning_rate": 1e-05,
      "loss": 0.3451,
      "step": 95200
    },
    {
      "epoch": 0.000953,
      "grad_norm": 2.1866915225982666,
      "learning_rate": 1e-05,
      "loss": 0.3392,
      "step": 95300
    },
    {
      "epoch": 0.000954,
      "grad_norm": 1.6935898065567017,
      "learning_rate": 1e-05,
      "loss": 0.3497,
      "step": 95400
    },
    {
      "epoch": 0.000955,
      "grad_norm": 1.9268600940704346,
      "learning_rate": 1e-05,
      "loss": 0.3463,
      "step": 95500
    },
    {
      "epoch": 0.000956,
      "grad_norm": 1.9195621013641357,
      "learning_rate": 1e-05,
      "loss": 0.351,
      "step": 95600
    },
    {
      "epoch": 0.000957,
      "grad_norm": 1.845158338546753,
      "learning_rate": 1e-05,
      "loss": 0.3465,
      "step": 95700
    },
    {
      "epoch": 0.000958,
      "grad_norm": 2.0196573734283447,
      "learning_rate": 1e-05,
      "loss": 0.3525,
      "step": 95800
    },
    {
      "epoch": 0.000959,
      "grad_norm": 1.8416608572006226,
      "learning_rate": 1e-05,
      "loss": 0.3502,
      "step": 95900
    },
    {
      "epoch": 0.00096,
      "grad_norm": 1.83146071434021,
      "learning_rate": 1e-05,
      "loss": 0.3535,
      "step": 96000
    },
    {
      "epoch": 0.000961,
      "grad_norm": 1.8110991716384888,
      "learning_rate": 1e-05,
      "loss": 0.3536,
      "step": 96100
    },
    {
      "epoch": 0.000962,
      "grad_norm": 1.798935055732727,
      "learning_rate": 1e-05,
      "loss": 0.3513,
      "step": 96200
    },
    {
      "epoch": 0.000963,
      "grad_norm": 1.7838218212127686,
      "learning_rate": 1e-05,
      "loss": 0.353,
      "step": 96300
    },
    {
      "epoch": 0.000964,
      "grad_norm": 1.8357594013214111,
      "learning_rate": 1e-05,
      "loss": 0.35,
      "step": 96400
    },
    {
      "epoch": 0.000965,
      "grad_norm": 1.9533332586288452,
      "learning_rate": 1e-05,
      "loss": 0.3497,
      "step": 96500
    },
    {
      "epoch": 0.000966,
      "grad_norm": 1.8430505990982056,
      "learning_rate": 1e-05,
      "loss": 0.3535,
      "step": 96600
    },
    {
      "epoch": 0.000967,
      "grad_norm": 1.842871069908142,
      "learning_rate": 1e-05,
      "loss": 0.3455,
      "step": 96700
    },
    {
      "epoch": 0.000968,
      "grad_norm": 1.8501172065734863,
      "learning_rate": 1e-05,
      "loss": 0.3469,
      "step": 96800
    },
    {
      "epoch": 0.000969,
      "grad_norm": 1.8171736001968384,
      "learning_rate": 1e-05,
      "loss": 0.3503,
      "step": 96900
    },
    {
      "epoch": 0.00097,
      "grad_norm": 1.8180707693099976,
      "learning_rate": 1e-05,
      "loss": 0.3509,
      "step": 97000
    },
    {
      "epoch": 0.000971,
      "grad_norm": 1.6564078330993652,
      "learning_rate": 1e-05,
      "loss": 0.3449,
      "step": 97100
    },
    {
      "epoch": 0.000972,
      "grad_norm": 1.9035217761993408,
      "learning_rate": 1e-05,
      "loss": 0.3464,
      "step": 97200
    },
    {
      "epoch": 0.000973,
      "grad_norm": 1.7870876789093018,
      "learning_rate": 1e-05,
      "loss": 0.358,
      "step": 97300
    },
    {
      "epoch": 0.000974,
      "grad_norm": 2.026207447052002,
      "learning_rate": 1e-05,
      "loss": 0.3469,
      "step": 97400
    },
    {
      "epoch": 0.000975,
      "grad_norm": 1.839242935180664,
      "learning_rate": 1e-05,
      "loss": 0.3527,
      "step": 97500
    },
    {
      "epoch": 0.000976,
      "grad_norm": 2.1023123264312744,
      "learning_rate": 1e-05,
      "loss": 0.3575,
      "step": 97600
    },
    {
      "epoch": 0.000977,
      "grad_norm": 1.7062361240386963,
      "learning_rate": 1e-05,
      "loss": 0.3427,
      "step": 97700
    },
    {
      "epoch": 0.000978,
      "grad_norm": 1.8973636627197266,
      "learning_rate": 1e-05,
      "loss": 0.3496,
      "step": 97800
    },
    {
      "epoch": 0.000979,
      "grad_norm": 4.865823745727539,
      "learning_rate": 1e-05,
      "loss": 0.3586,
      "step": 97900
    },
    {
      "epoch": 0.00098,
      "grad_norm": 1.6862282752990723,
      "learning_rate": 1e-05,
      "loss": 0.3483,
      "step": 98000
    },
    {
      "epoch": 0.000981,
      "grad_norm": 1.7278543710708618,
      "learning_rate": 1e-05,
      "loss": 0.345,
      "step": 98100
    },
    {
      "epoch": 0.000982,
      "grad_norm": 1.9642508029937744,
      "learning_rate": 1e-05,
      "loss": 0.3552,
      "step": 98200
    },
    {
      "epoch": 0.000983,
      "grad_norm": 1.6919240951538086,
      "learning_rate": 1e-05,
      "loss": 0.3481,
      "step": 98300
    },
    {
      "epoch": 0.000984,
      "grad_norm": 1.7211792469024658,
      "learning_rate": 1e-05,
      "loss": 0.3456,
      "step": 98400
    },
    {
      "epoch": 0.000985,
      "grad_norm": 1.8794984817504883,
      "learning_rate": 1e-05,
      "loss": 0.3547,
      "step": 98500
    },
    {
      "epoch": 0.000986,
      "grad_norm": 1.7422791719436646,
      "learning_rate": 1e-05,
      "loss": 0.3459,
      "step": 98600
    },
    {
      "epoch": 0.000987,
      "grad_norm": 1.7812235355377197,
      "learning_rate": 1e-05,
      "loss": 0.3534,
      "step": 98700
    },
    {
      "epoch": 0.000988,
      "grad_norm": 1.7994880676269531,
      "learning_rate": 1e-05,
      "loss": 0.3406,
      "step": 98800
    },
    {
      "epoch": 0.000989,
      "grad_norm": 1.766994595527649,
      "learning_rate": 1e-05,
      "loss": 0.3454,
      "step": 98900
    },
    {
      "epoch": 0.00099,
      "grad_norm": 1.9302865266799927,
      "learning_rate": 1e-05,
      "loss": 0.3434,
      "step": 99000
    },
    {
      "epoch": 0.000991,
      "grad_norm": 1.6279524564743042,
      "learning_rate": 1e-05,
      "loss": 0.3443,
      "step": 99100
    },
    {
      "epoch": 0.000992,
      "grad_norm": 1.878088116645813,
      "learning_rate": 1e-05,
      "loss": 0.3433,
      "step": 99200
    },
    {
      "epoch": 0.000993,
      "grad_norm": 1.9811022281646729,
      "learning_rate": 1e-05,
      "loss": 0.3444,
      "step": 99300
    },
    {
      "epoch": 0.000994,
      "grad_norm": 1.9504814147949219,
      "learning_rate": 1e-05,
      "loss": 0.3448,
      "step": 99400
    },
    {
      "epoch": 0.000995,
      "grad_norm": 1.7477716207504272,
      "learning_rate": 1e-05,
      "loss": 0.3372,
      "step": 99500
    },
    {
      "epoch": 0.000996,
      "grad_norm": 1.9687480926513672,
      "learning_rate": 1e-05,
      "loss": 0.346,
      "step": 99600
    },
    {
      "epoch": 0.000997,
      "grad_norm": 2.0356996059417725,
      "learning_rate": 1e-05,
      "loss": 0.3508,
      "step": 99700
    },
    {
      "epoch": 0.000998,
      "grad_norm": 1.816023349761963,
      "learning_rate": 1e-05,
      "loss": 0.3524,
      "step": 99800
    },
    {
      "epoch": 0.000999,
      "grad_norm": 2.0732617378234863,
      "learning_rate": 1e-05,
      "loss": 0.3468,
      "step": 99900
    },
    {
      "epoch": 0.001,
      "grad_norm": 1.8265982866287231,
      "learning_rate": 1e-05,
      "loss": 0.3485,
      "step": 100000
    },
    {
      "epoch": 0.001,
      "eval_loss": 0.322509765625,
      "eval_runtime": 109.8279,
      "eval_samples_per_second": 455.258,
      "eval_steps_per_second": 28.454,
      "step": 100000
    },
    {
      "epoch": 0.001001,
      "grad_norm": 1.6607346534729004,
      "learning_rate": 1e-05,
      "loss": 0.35,
      "step": 100100
    },
    {
      "epoch": 0.001002,
      "grad_norm": 1.8534297943115234,
      "learning_rate": 1e-05,
      "loss": 0.3515,
      "step": 100200
    },
    {
      "epoch": 0.001003,
      "grad_norm": 1.7251813411712646,
      "learning_rate": 1e-05,
      "loss": 0.3449,
      "step": 100300
    },
    {
      "epoch": 0.001004,
      "grad_norm": 2.1961374282836914,
      "learning_rate": 1e-05,
      "loss": 0.3442,
      "step": 100400
    },
    {
      "epoch": 0.001005,
      "grad_norm": 1.8674982786178589,
      "learning_rate": 1e-05,
      "loss": 0.357,
      "step": 100500
    },
    {
      "epoch": 0.001006,
      "grad_norm": 1.9515495300292969,
      "learning_rate": 1e-05,
      "loss": 0.3516,
      "step": 100600
    },
    {
      "epoch": 0.001007,
      "grad_norm": 1.8294038772583008,
      "learning_rate": 1e-05,
      "loss": 0.3475,
      "step": 100700
    },
    {
      "epoch": 0.001008,
      "grad_norm": 1.7040425539016724,
      "learning_rate": 1e-05,
      "loss": 0.3403,
      "step": 100800
    },
    {
      "epoch": 0.001009,
      "grad_norm": 2.464323043823242,
      "learning_rate": 1e-05,
      "loss": 0.3437,
      "step": 100900
    },
    {
      "epoch": 0.00101,
      "grad_norm": 1.5711098909378052,
      "learning_rate": 1e-05,
      "loss": 0.3465,
      "step": 101000
    },
    {
      "epoch": 0.001011,
      "grad_norm": 1.6807917356491089,
      "learning_rate": 1e-05,
      "loss": 0.3462,
      "step": 101100
    },
    {
      "epoch": 0.001012,
      "grad_norm": 2.2576940059661865,
      "learning_rate": 1e-05,
      "loss": 0.3389,
      "step": 101200
    },
    {
      "epoch": 0.001013,
      "grad_norm": 1.7972438335418701,
      "learning_rate": 1e-05,
      "loss": 0.3442,
      "step": 101300
    },
    {
      "epoch": 0.001014,
      "grad_norm": 1.8780492544174194,
      "learning_rate": 1e-05,
      "loss": 0.3424,
      "step": 101400
    },
    {
      "epoch": 0.001015,
      "grad_norm": 1.7459834814071655,
      "learning_rate": 1e-05,
      "loss": 0.3403,
      "step": 101500
    },
    {
      "epoch": 0.001016,
      "grad_norm": 1.549613118171692,
      "learning_rate": 1e-05,
      "loss": 0.3354,
      "step": 101600
    },
    {
      "epoch": 0.001017,
      "grad_norm": 1.737701416015625,
      "learning_rate": 1e-05,
      "loss": 0.3447,
      "step": 101700
    },
    {
      "epoch": 0.001018,
      "grad_norm": 1.936926245689392,
      "learning_rate": 1e-05,
      "loss": 0.3457,
      "step": 101800
    },
    {
      "epoch": 0.001019,
      "grad_norm": 1.579476237297058,
      "learning_rate": 1e-05,
      "loss": 0.3388,
      "step": 101900
    },
    {
      "epoch": 0.00102,
      "grad_norm": 1.620912790298462,
      "learning_rate": 1e-05,
      "loss": 0.342,
      "step": 102000
    },
    {
      "epoch": 0.001021,
      "grad_norm": 1.801440715789795,
      "learning_rate": 1e-05,
      "loss": 0.3479,
      "step": 102100
    },
    {
      "epoch": 0.001022,
      "grad_norm": 1.9294061660766602,
      "learning_rate": 1e-05,
      "loss": 0.3535,
      "step": 102200
    },
    {
      "epoch": 0.001023,
      "grad_norm": 1.7232532501220703,
      "learning_rate": 1e-05,
      "loss": 0.3465,
      "step": 102300
    },
    {
      "epoch": 0.001024,
      "grad_norm": 2.327086925506592,
      "learning_rate": 1e-05,
      "loss": 0.3379,
      "step": 102400
    },
    {
      "epoch": 0.001025,
      "grad_norm": 1.652092695236206,
      "learning_rate": 1e-05,
      "loss": 0.3399,
      "step": 102500
    },
    {
      "epoch": 0.001026,
      "grad_norm": 1.6813795566558838,
      "learning_rate": 1e-05,
      "loss": 0.3489,
      "step": 102600
    },
    {
      "epoch": 0.001027,
      "grad_norm": 1.9206030368804932,
      "learning_rate": 1e-05,
      "loss": 0.3476,
      "step": 102700
    },
    {
      "epoch": 0.001028,
      "grad_norm": 1.5671733617782593,
      "learning_rate": 1e-05,
      "loss": 0.337,
      "step": 102800
    },
    {
      "epoch": 0.001029,
      "grad_norm": 1.8689380884170532,
      "learning_rate": 1e-05,
      "loss": 0.3449,
      "step": 102900
    },
    {
      "epoch": 0.00103,
      "grad_norm": 1.6365469694137573,
      "learning_rate": 1e-05,
      "loss": 0.3421,
      "step": 103000
    },
    {
      "epoch": 0.001031,
      "grad_norm": 2.0219128131866455,
      "learning_rate": 1e-05,
      "loss": 0.3402,
      "step": 103100
    },
    {
      "epoch": 0.001032,
      "grad_norm": 1.6188757419586182,
      "learning_rate": 1e-05,
      "loss": 0.3458,
      "step": 103200
    },
    {
      "epoch": 0.001033,
      "grad_norm": 1.809544324874878,
      "learning_rate": 1e-05,
      "loss": 0.3436,
      "step": 103300
    },
    {
      "epoch": 0.001034,
      "grad_norm": 1.6749653816223145,
      "learning_rate": 1e-05,
      "loss": 0.3436,
      "step": 103400
    },
    {
      "epoch": 0.001035,
      "grad_norm": 1.9452251195907593,
      "learning_rate": 1e-05,
      "loss": 0.3355,
      "step": 103500
    },
    {
      "epoch": 0.001036,
      "grad_norm": 1.790397047996521,
      "learning_rate": 1e-05,
      "loss": 0.3489,
      "step": 103600
    },
    {
      "epoch": 0.001037,
      "grad_norm": 1.5875970125198364,
      "learning_rate": 1e-05,
      "loss": 0.3358,
      "step": 103700
    },
    {
      "epoch": 0.001038,
      "grad_norm": 1.6320905685424805,
      "learning_rate": 1e-05,
      "loss": 0.3436,
      "step": 103800
    },
    {
      "epoch": 0.001039,
      "grad_norm": 1.6067711114883423,
      "learning_rate": 1e-05,
      "loss": 0.3438,
      "step": 103900
    },
    {
      "epoch": 0.00104,
      "grad_norm": 1.8375946283340454,
      "learning_rate": 1e-05,
      "loss": 0.344,
      "step": 104000
    },
    {
      "epoch": 0.001041,
      "grad_norm": 1.708240270614624,
      "learning_rate": 1e-05,
      "loss": 0.3491,
      "step": 104100
    },
    {
      "epoch": 0.001042,
      "grad_norm": 2.3994433879852295,
      "learning_rate": 1e-05,
      "loss": 0.3462,
      "step": 104200
    },
    {
      "epoch": 0.001043,
      "grad_norm": 1.7040139436721802,
      "learning_rate": 1e-05,
      "loss": 0.3459,
      "step": 104300
    },
    {
      "epoch": 0.001044,
      "grad_norm": 2.01163911819458,
      "learning_rate": 1e-05,
      "loss": 0.3463,
      "step": 104400
    },
    {
      "epoch": 0.001045,
      "grad_norm": 1.604658842086792,
      "learning_rate": 1e-05,
      "loss": 0.341,
      "step": 104500
    },
    {
      "epoch": 0.001046,
      "grad_norm": 2.69278883934021,
      "learning_rate": 1e-05,
      "loss": 0.3418,
      "step": 104600
    },
    {
      "epoch": 0.001047,
      "grad_norm": 1.6742432117462158,
      "learning_rate": 1e-05,
      "loss": 0.3376,
      "step": 104700
    },
    {
      "epoch": 0.001048,
      "grad_norm": 1.7139792442321777,
      "learning_rate": 1e-05,
      "loss": 0.3448,
      "step": 104800
    },
    {
      "epoch": 0.001049,
      "grad_norm": 1.9812430143356323,
      "learning_rate": 1e-05,
      "loss": 0.3443,
      "step": 104900
    },
    {
      "epoch": 0.00105,
      "grad_norm": 1.9630818367004395,
      "learning_rate": 1e-05,
      "loss": 0.3367,
      "step": 105000
    },
    {
      "epoch": 0.001051,
      "grad_norm": 1.913673758506775,
      "learning_rate": 1e-05,
      "loss": 0.3433,
      "step": 105100
    },
    {
      "epoch": 0.001052,
      "grad_norm": 1.6645994186401367,
      "learning_rate": 1e-05,
      "loss": 0.3483,
      "step": 105200
    },
    {
      "epoch": 0.001053,
      "grad_norm": 1.6963456869125366,
      "learning_rate": 1e-05,
      "loss": 0.3518,
      "step": 105300
    },
    {
      "epoch": 0.001054,
      "grad_norm": 1.774322748184204,
      "learning_rate": 1e-05,
      "loss": 0.339,
      "step": 105400
    },
    {
      "epoch": 0.001055,
      "grad_norm": 1.6794517040252686,
      "learning_rate": 1e-05,
      "loss": 0.3412,
      "step": 105500
    },
    {
      "epoch": 0.001056,
      "grad_norm": 2.0357189178466797,
      "learning_rate": 1e-05,
      "loss": 0.3406,
      "step": 105600
    },
    {
      "epoch": 0.001057,
      "grad_norm": 1.7818143367767334,
      "learning_rate": 1e-05,
      "loss": 0.3379,
      "step": 105700
    },
    {
      "epoch": 0.001058,
      "grad_norm": 1.819798231124878,
      "learning_rate": 1e-05,
      "loss": 0.3328,
      "step": 105800
    },
    {
      "epoch": 0.001059,
      "grad_norm": 1.751774787902832,
      "learning_rate": 1e-05,
      "loss": 0.3403,
      "step": 105900
    },
    {
      "epoch": 0.00106,
      "grad_norm": 1.720474362373352,
      "learning_rate": 1e-05,
      "loss": 0.3427,
      "step": 106000
    },
    {
      "epoch": 0.001061,
      "grad_norm": 1.7977921962738037,
      "learning_rate": 1e-05,
      "loss": 0.3435,
      "step": 106100
    },
    {
      "epoch": 0.001062,
      "grad_norm": 2.1512701511383057,
      "learning_rate": 1e-05,
      "loss": 0.3443,
      "step": 106200
    },
    {
      "epoch": 0.001063,
      "grad_norm": 1.7027465105056763,
      "learning_rate": 1e-05,
      "loss": 0.3362,
      "step": 106300
    },
    {
      "epoch": 0.001064,
      "grad_norm": 1.663902997970581,
      "learning_rate": 1e-05,
      "loss": 0.3459,
      "step": 106400
    },
    {
      "epoch": 0.001065,
      "grad_norm": 1.7038410902023315,
      "learning_rate": 1e-05,
      "loss": 0.3356,
      "step": 106500
    },
    {
      "epoch": 0.001066,
      "grad_norm": 1.5602421760559082,
      "learning_rate": 1e-05,
      "loss": 0.3354,
      "step": 106600
    },
    {
      "epoch": 0.001067,
      "grad_norm": 1.8753788471221924,
      "learning_rate": 1e-05,
      "loss": 0.3342,
      "step": 106700
    },
    {
      "epoch": 0.001068,
      "grad_norm": 1.731338620185852,
      "learning_rate": 1e-05,
      "loss": 0.34,
      "step": 106800
    },
    {
      "epoch": 0.001069,
      "grad_norm": 1.7717700004577637,
      "learning_rate": 1e-05,
      "loss": 0.3404,
      "step": 106900
    },
    {
      "epoch": 0.00107,
      "grad_norm": 1.5369184017181396,
      "learning_rate": 1e-05,
      "loss": 0.3534,
      "step": 107000
    },
    {
      "epoch": 0.001071,
      "grad_norm": 2.0892210006713867,
      "learning_rate": 1e-05,
      "loss": 0.3396,
      "step": 107100
    },
    {
      "epoch": 0.001072,
      "grad_norm": 1.9275939464569092,
      "learning_rate": 1e-05,
      "loss": 0.3403,
      "step": 107200
    },
    {
      "epoch": 0.001073,
      "grad_norm": 1.9656401872634888,
      "learning_rate": 1e-05,
      "loss": 0.3392,
      "step": 107300
    },
    {
      "epoch": 0.001074,
      "grad_norm": 1.7235068082809448,
      "learning_rate": 1e-05,
      "loss": 0.3416,
      "step": 107400
    },
    {
      "epoch": 0.001075,
      "grad_norm": 1.8416111469268799,
      "learning_rate": 1e-05,
      "loss": 0.3367,
      "step": 107500
    },
    {
      "epoch": 0.001076,
      "grad_norm": 1.7464598417282104,
      "learning_rate": 1e-05,
      "loss": 0.3436,
      "step": 107600
    },
    {
      "epoch": 0.001077,
      "grad_norm": 1.8630284070968628,
      "learning_rate": 1e-05,
      "loss": 0.3432,
      "step": 107700
    },
    {
      "epoch": 0.001078,
      "grad_norm": 1.740233302116394,
      "learning_rate": 1e-05,
      "loss": 0.3384,
      "step": 107800
    },
    {
      "epoch": 0.001079,
      "grad_norm": 1.5509331226348877,
      "learning_rate": 1e-05,
      "loss": 0.3332,
      "step": 107900
    },
    {
      "epoch": 0.00108,
      "grad_norm": 1.7383582592010498,
      "learning_rate": 1e-05,
      "loss": 0.3456,
      "step": 108000
    },
    {
      "epoch": 0.001081,
      "grad_norm": 1.9408977031707764,
      "learning_rate": 1e-05,
      "loss": 0.3396,
      "step": 108100
    },
    {
      "epoch": 0.001082,
      "grad_norm": 1.6888933181762695,
      "learning_rate": 1e-05,
      "loss": 0.3404,
      "step": 108200
    },
    {
      "epoch": 0.001083,
      "grad_norm": 1.9360098838806152,
      "learning_rate": 1e-05,
      "loss": 0.3431,
      "step": 108300
    },
    {
      "epoch": 0.001084,
      "grad_norm": 1.7306195497512817,
      "learning_rate": 1e-05,
      "loss": 0.3348,
      "step": 108400
    },
    {
      "epoch": 0.001085,
      "grad_norm": 1.6970361471176147,
      "learning_rate": 1e-05,
      "loss": 0.3381,
      "step": 108500
    },
    {
      "epoch": 0.001086,
      "grad_norm": 1.664059042930603,
      "learning_rate": 1e-05,
      "loss": 0.3403,
      "step": 108600
    },
    {
      "epoch": 0.001087,
      "grad_norm": 1.7473076581954956,
      "learning_rate": 1e-05,
      "loss": 0.3403,
      "step": 108700
    },
    {
      "epoch": 0.001088,
      "grad_norm": 1.705640196800232,
      "learning_rate": 1e-05,
      "loss": 0.3353,
      "step": 108800
    },
    {
      "epoch": 0.001089,
      "grad_norm": 1.8058274984359741,
      "learning_rate": 1e-05,
      "loss": 0.3309,
      "step": 108900
    },
    {
      "epoch": 0.00109,
      "grad_norm": 1.7639211416244507,
      "learning_rate": 1e-05,
      "loss": 0.3322,
      "step": 109000
    },
    {
      "epoch": 0.001091,
      "grad_norm": 2.020731210708618,
      "learning_rate": 1e-05,
      "loss": 0.3421,
      "step": 109100
    },
    {
      "epoch": 0.001092,
      "grad_norm": 1.7614929676055908,
      "learning_rate": 1e-05,
      "loss": 0.3395,
      "step": 109200
    },
    {
      "epoch": 0.001093,
      "grad_norm": 1.8153364658355713,
      "learning_rate": 1e-05,
      "loss": 0.3413,
      "step": 109300
    },
    {
      "epoch": 0.001094,
      "grad_norm": 1.803002119064331,
      "learning_rate": 1e-05,
      "loss": 0.337,
      "step": 109400
    },
    {
      "epoch": 0.001095,
      "grad_norm": 1.6940698623657227,
      "learning_rate": 1e-05,
      "loss": 0.3372,
      "step": 109500
    },
    {
      "epoch": 0.001096,
      "grad_norm": 1.8647571802139282,
      "learning_rate": 1e-05,
      "loss": 0.338,
      "step": 109600
    },
    {
      "epoch": 0.001097,
      "grad_norm": 1.9329015016555786,
      "learning_rate": 1e-05,
      "loss": 0.3397,
      "step": 109700
    },
    {
      "epoch": 0.001098,
      "grad_norm": 1.8334521055221558,
      "learning_rate": 1e-05,
      "loss": 0.3368,
      "step": 109800
    },
    {
      "epoch": 0.001099,
      "grad_norm": 1.6593636274337769,
      "learning_rate": 1e-05,
      "loss": 0.335,
      "step": 109900
    },
    {
      "epoch": 0.0011,
      "grad_norm": 1.8620237112045288,
      "learning_rate": 1e-05,
      "loss": 0.3383,
      "step": 110000
    },
    {
      "epoch": 0.001101,
      "grad_norm": 1.8764339685440063,
      "learning_rate": 1e-05,
      "loss": 0.3338,
      "step": 110100
    },
    {
      "epoch": 0.001102,
      "grad_norm": 1.9678648710250854,
      "learning_rate": 1e-05,
      "loss": 0.3441,
      "step": 110200
    },
    {
      "epoch": 0.001103,
      "grad_norm": 1.5820297002792358,
      "learning_rate": 1e-05,
      "loss": 0.3388,
      "step": 110300
    },
    {
      "epoch": 0.001104,
      "grad_norm": 1.78269362449646,
      "learning_rate": 1e-05,
      "loss": 0.34,
      "step": 110400
    },
    {
      "epoch": 0.001105,
      "grad_norm": 2.253110647201538,
      "learning_rate": 1e-05,
      "loss": 0.3352,
      "step": 110500
    },
    {
      "epoch": 0.001106,
      "grad_norm": 1.553359866142273,
      "learning_rate": 1e-05,
      "loss": 0.3316,
      "step": 110600
    },
    {
      "epoch": 0.001107,
      "grad_norm": 1.8975584506988525,
      "learning_rate": 1e-05,
      "loss": 0.3363,
      "step": 110700
    },
    {
      "epoch": 0.001108,
      "grad_norm": 1.9635969400405884,
      "learning_rate": 1e-05,
      "loss": 0.3339,
      "step": 110800
    },
    {
      "epoch": 0.001109,
      "grad_norm": 1.6212959289550781,
      "learning_rate": 1e-05,
      "loss": 0.3365,
      "step": 110900
    },
    {
      "epoch": 0.00111,
      "grad_norm": 1.667982578277588,
      "learning_rate": 1e-05,
      "loss": 0.3385,
      "step": 111000
    },
    {
      "epoch": 0.001111,
      "grad_norm": 1.799843192100525,
      "learning_rate": 1e-05,
      "loss": 0.3342,
      "step": 111100
    },
    {
      "epoch": 0.001112,
      "grad_norm": 1.9425872564315796,
      "learning_rate": 1e-05,
      "loss": 0.3445,
      "step": 111200
    },
    {
      "epoch": 0.001113,
      "grad_norm": 1.7052315473556519,
      "learning_rate": 1e-05,
      "loss": 0.3321,
      "step": 111300
    },
    {
      "epoch": 0.001114,
      "grad_norm": 1.8439725637435913,
      "learning_rate": 1e-05,
      "loss": 0.3362,
      "step": 111400
    },
    {
      "epoch": 0.001115,
      "grad_norm": 1.5285687446594238,
      "learning_rate": 1e-05,
      "loss": 0.3362,
      "step": 111500
    },
    {
      "epoch": 0.001116,
      "grad_norm": 1.8127108812332153,
      "learning_rate": 1e-05,
      "loss": 0.3351,
      "step": 111600
    },
    {
      "epoch": 0.001117,
      "grad_norm": 2.099846363067627,
      "learning_rate": 1e-05,
      "loss": 0.3348,
      "step": 111700
    },
    {
      "epoch": 0.001118,
      "grad_norm": 1.8282333612442017,
      "learning_rate": 1e-05,
      "loss": 0.3353,
      "step": 111800
    },
    {
      "epoch": 0.001119,
      "grad_norm": 1.9214797019958496,
      "learning_rate": 1e-05,
      "loss": 0.3384,
      "step": 111900
    },
    {
      "epoch": 0.00112,
      "grad_norm": 1.5597374439239502,
      "learning_rate": 1e-05,
      "loss": 0.339,
      "step": 112000
    },
    {
      "epoch": 0.001121,
      "grad_norm": 1.7652438879013062,
      "learning_rate": 1e-05,
      "loss": 0.336,
      "step": 112100
    },
    {
      "epoch": 0.001122,
      "grad_norm": 1.6596375703811646,
      "learning_rate": 1e-05,
      "loss": 0.3375,
      "step": 112200
    },
    {
      "epoch": 0.001123,
      "grad_norm": 1.7220653295516968,
      "learning_rate": 1e-05,
      "loss": 0.3424,
      "step": 112300
    },
    {
      "epoch": 0.001124,
      "grad_norm": 1.7662781476974487,
      "learning_rate": 1e-05,
      "loss": 0.3389,
      "step": 112400
    },
    {
      "epoch": 0.001125,
      "grad_norm": 2.3776822090148926,
      "learning_rate": 1e-05,
      "loss": 0.3366,
      "step": 112500
    },
    {
      "epoch": 0.001126,
      "grad_norm": 1.8062163591384888,
      "learning_rate": 1e-05,
      "loss": 0.3366,
      "step": 112600
    },
    {
      "epoch": 0.001127,
      "grad_norm": 1.6732230186462402,
      "learning_rate": 1e-05,
      "loss": 0.3338,
      "step": 112700
    },
    {
      "epoch": 0.001128,
      "grad_norm": 1.7580362558364868,
      "learning_rate": 1e-05,
      "loss": 0.33,
      "step": 112800
    },
    {
      "epoch": 0.001129,
      "grad_norm": 1.4763306379318237,
      "learning_rate": 1e-05,
      "loss": 0.3323,
      "step": 112900
    },
    {
      "epoch": 0.00113,
      "grad_norm": 1.7444915771484375,
      "learning_rate": 1e-05,
      "loss": 0.3368,
      "step": 113000
    },
    {
      "epoch": 0.001131,
      "grad_norm": 1.487596869468689,
      "learning_rate": 1e-05,
      "loss": 0.3372,
      "step": 113100
    },
    {
      "epoch": 0.001132,
      "grad_norm": 1.7567662000656128,
      "learning_rate": 1e-05,
      "loss": 0.3363,
      "step": 113200
    },
    {
      "epoch": 0.001133,
      "grad_norm": 1.742226243019104,
      "learning_rate": 1e-05,
      "loss": 0.3353,
      "step": 113300
    },
    {
      "epoch": 0.001134,
      "grad_norm": 1.7816177606582642,
      "learning_rate": 1e-05,
      "loss": 0.3363,
      "step": 113400
    },
    {
      "epoch": 0.001135,
      "grad_norm": 1.9755498170852661,
      "learning_rate": 1e-05,
      "loss": 0.3418,
      "step": 113500
    },
    {
      "epoch": 0.001136,
      "grad_norm": 1.867906093597412,
      "learning_rate": 1e-05,
      "loss": 0.3334,
      "step": 113600
    },
    {
      "epoch": 0.001137,
      "grad_norm": 1.6979436874389648,
      "learning_rate": 1e-05,
      "loss": 0.3229,
      "step": 113700
    },
    {
      "epoch": 0.001138,
      "grad_norm": 1.7003270387649536,
      "learning_rate": 1e-05,
      "loss": 0.339,
      "step": 113800
    },
    {
      "epoch": 0.001139,
      "grad_norm": 1.7175750732421875,
      "learning_rate": 1e-05,
      "loss": 0.328,
      "step": 113900
    },
    {
      "epoch": 0.00114,
      "grad_norm": 1.8313535451889038,
      "learning_rate": 1e-05,
      "loss": 0.334,
      "step": 114000
    },
    {
      "epoch": 0.001141,
      "grad_norm": 1.863562822341919,
      "learning_rate": 1e-05,
      "loss": 0.3331,
      "step": 114100
    },
    {
      "epoch": 0.001142,
      "grad_norm": 1.5223082304000854,
      "learning_rate": 1e-05,
      "loss": 0.3327,
      "step": 114200
    },
    {
      "epoch": 0.001143,
      "grad_norm": 1.6789870262145996,
      "learning_rate": 1e-05,
      "loss": 0.3313,
      "step": 114300
    },
    {
      "epoch": 0.001144,
      "grad_norm": 1.6574594974517822,
      "learning_rate": 1e-05,
      "loss": 0.3376,
      "step": 114400
    },
    {
      "epoch": 0.001145,
      "grad_norm": 1.8169411420822144,
      "learning_rate": 1e-05,
      "loss": 0.3392,
      "step": 114500
    },
    {
      "epoch": 0.001146,
      "grad_norm": 2.384134292602539,
      "learning_rate": 1e-05,
      "loss": 0.3421,
      "step": 114600
    },
    {
      "epoch": 0.001147,
      "grad_norm": 1.8304411172866821,
      "learning_rate": 1e-05,
      "loss": 0.3356,
      "step": 114700
    },
    {
      "epoch": 0.001148,
      "grad_norm": 1.5321639776229858,
      "learning_rate": 1e-05,
      "loss": 0.3317,
      "step": 114800
    },
    {
      "epoch": 0.001149,
      "grad_norm": 1.7843445539474487,
      "learning_rate": 1e-05,
      "loss": 0.3267,
      "step": 114900
    },
    {
      "epoch": 0.00115,
      "grad_norm": 1.8861100673675537,
      "learning_rate": 1e-05,
      "loss": 0.3318,
      "step": 115000
    },
    {
      "epoch": 0.001151,
      "grad_norm": 1.8112998008728027,
      "learning_rate": 1e-05,
      "loss": 0.3293,
      "step": 115100
    },
    {
      "epoch": 0.001152,
      "grad_norm": 1.7408936023712158,
      "learning_rate": 1e-05,
      "loss": 0.339,
      "step": 115200
    },
    {
      "epoch": 0.001153,
      "grad_norm": 1.5955983400344849,
      "learning_rate": 1e-05,
      "loss": 0.3326,
      "step": 115300
    },
    {
      "epoch": 0.001154,
      "grad_norm": 1.6836644411087036,
      "learning_rate": 1e-05,
      "loss": 0.3377,
      "step": 115400
    },
    {
      "epoch": 0.001155,
      "grad_norm": 1.7743850946426392,
      "learning_rate": 1e-05,
      "loss": 0.3278,
      "step": 115500
    },
    {
      "epoch": 0.001156,
      "grad_norm": 1.5382933616638184,
      "learning_rate": 1e-05,
      "loss": 0.3397,
      "step": 115600
    },
    {
      "epoch": 0.001157,
      "grad_norm": 1.7360892295837402,
      "learning_rate": 1e-05,
      "loss": 0.3207,
      "step": 115700
    },
    {
      "epoch": 0.001158,
      "grad_norm": 1.7574350833892822,
      "learning_rate": 1e-05,
      "loss": 0.3439,
      "step": 115800
    },
    {
      "epoch": 0.001159,
      "grad_norm": 1.8633227348327637,
      "learning_rate": 1e-05,
      "loss": 0.335,
      "step": 115900
    },
    {
      "epoch": 0.00116,
      "grad_norm": 1.4621351957321167,
      "learning_rate": 1e-05,
      "loss": 0.3358,
      "step": 116000
    },
    {
      "epoch": 0.001161,
      "grad_norm": 1.9157224893569946,
      "learning_rate": 1e-05,
      "loss": 0.3406,
      "step": 116100
    },
    {
      "epoch": 0.001162,
      "grad_norm": 1.6284751892089844,
      "learning_rate": 1e-05,
      "loss": 0.3277,
      "step": 116200
    },
    {
      "epoch": 0.001163,
      "grad_norm": 2.2173221111297607,
      "learning_rate": 1e-05,
      "loss": 0.3259,
      "step": 116300
    },
    {
      "epoch": 0.001164,
      "grad_norm": 1.8805922269821167,
      "learning_rate": 1e-05,
      "loss": 0.3304,
      "step": 116400
    },
    {
      "epoch": 0.001165,
      "grad_norm": 1.5072230100631714,
      "learning_rate": 1e-05,
      "loss": 0.3329,
      "step": 116500
    },
    {
      "epoch": 0.001166,
      "grad_norm": 1.7337315082550049,
      "learning_rate": 1e-05,
      "loss": 0.3358,
      "step": 116600
    },
    {
      "epoch": 0.001167,
      "grad_norm": 1.8346338272094727,
      "learning_rate": 1e-05,
      "loss": 0.335,
      "step": 116700
    },
    {
      "epoch": 0.001168,
      "grad_norm": 2.003572940826416,
      "learning_rate": 1e-05,
      "loss": 0.333,
      "step": 116800
    },
    {
      "epoch": 0.001169,
      "grad_norm": 1.6946192979812622,
      "learning_rate": 1e-05,
      "loss": 0.3274,
      "step": 116900
    },
    {
      "epoch": 0.00117,
      "grad_norm": 1.7123721837997437,
      "learning_rate": 1e-05,
      "loss": 0.3346,
      "step": 117000
    },
    {
      "epoch": 0.001171,
      "grad_norm": 1.8998627662658691,
      "learning_rate": 1e-05,
      "loss": 0.3355,
      "step": 117100
    },
    {
      "epoch": 0.001172,
      "grad_norm": 1.5401489734649658,
      "learning_rate": 1e-05,
      "loss": 0.3365,
      "step": 117200
    },
    {
      "epoch": 0.001173,
      "grad_norm": 1.7201097011566162,
      "learning_rate": 1e-05,
      "loss": 0.3382,
      "step": 117300
    },
    {
      "epoch": 0.001174,
      "grad_norm": 1.8772022724151611,
      "learning_rate": 1e-05,
      "loss": 0.3323,
      "step": 117400
    },
    {
      "epoch": 0.001175,
      "grad_norm": 1.5749614238739014,
      "learning_rate": 1e-05,
      "loss": 0.3247,
      "step": 117500
    },
    {
      "epoch": 0.001176,
      "grad_norm": 1.753891944885254,
      "learning_rate": 1e-05,
      "loss": 0.325,
      "step": 117600
    },
    {
      "epoch": 0.001177,
      "grad_norm": 1.6012095212936401,
      "learning_rate": 1e-05,
      "loss": 0.3353,
      "step": 117700
    },
    {
      "epoch": 0.001178,
      "grad_norm": 1.8432629108428955,
      "learning_rate": 1e-05,
      "loss": 0.3342,
      "step": 117800
    },
    {
      "epoch": 0.001179,
      "grad_norm": 1.583196997642517,
      "learning_rate": 1e-05,
      "loss": 0.33,
      "step": 117900
    },
    {
      "epoch": 0.00118,
      "grad_norm": 1.9006246328353882,
      "learning_rate": 1e-05,
      "loss": 0.3282,
      "step": 118000
    },
    {
      "epoch": 0.001181,
      "grad_norm": 1.6398696899414062,
      "learning_rate": 1e-05,
      "loss": 0.3326,
      "step": 118100
    },
    {
      "epoch": 0.001182,
      "grad_norm": 1.7854382991790771,
      "learning_rate": 1e-05,
      "loss": 0.3273,
      "step": 118200
    },
    {
      "epoch": 0.001183,
      "grad_norm": 2.0176942348480225,
      "learning_rate": 1e-05,
      "loss": 0.3296,
      "step": 118300
    },
    {
      "epoch": 0.001184,
      "grad_norm": 2.0091938972473145,
      "learning_rate": 1e-05,
      "loss": 0.3291,
      "step": 118400
    },
    {
      "epoch": 0.001185,
      "grad_norm": 1.906575322151184,
      "learning_rate": 1e-05,
      "loss": 0.3354,
      "step": 118500
    },
    {
      "epoch": 0.001186,
      "grad_norm": 1.7136719226837158,
      "learning_rate": 1e-05,
      "loss": 0.3314,
      "step": 118600
    },
    {
      "epoch": 0.001187,
      "grad_norm": 1.7901870012283325,
      "learning_rate": 1e-05,
      "loss": 0.3236,
      "step": 118700
    },
    {
      "epoch": 0.001188,
      "grad_norm": 1.7387175559997559,
      "learning_rate": 1e-05,
      "loss": 0.3227,
      "step": 118800
    },
    {
      "epoch": 0.001189,
      "grad_norm": 1.7231628894805908,
      "learning_rate": 1e-05,
      "loss": 0.324,
      "step": 118900
    },
    {
      "epoch": 0.00119,
      "grad_norm": 1.516570806503296,
      "learning_rate": 1e-05,
      "loss": 0.3332,
      "step": 119000
    },
    {
      "epoch": 0.001191,
      "grad_norm": 1.7026876211166382,
      "learning_rate": 1e-05,
      "loss": 0.3272,
      "step": 119100
    },
    {
      "epoch": 0.001192,
      "grad_norm": 1.3457015752792358,
      "learning_rate": 1e-05,
      "loss": 0.3306,
      "step": 119200
    },
    {
      "epoch": 0.001193,
      "grad_norm": 1.9337682723999023,
      "learning_rate": 1e-05,
      "loss": 0.3237,
      "step": 119300
    },
    {
      "epoch": 0.001194,
      "grad_norm": 1.6353681087493896,
      "learning_rate": 1e-05,
      "loss": 0.3287,
      "step": 119400
    },
    {
      "epoch": 0.001195,
      "grad_norm": 1.875755786895752,
      "learning_rate": 1e-05,
      "loss": 0.3245,
      "step": 119500
    },
    {
      "epoch": 0.001196,
      "grad_norm": 2.4236490726470947,
      "learning_rate": 1e-05,
      "loss": 0.3274,
      "step": 119600
    },
    {
      "epoch": 0.001197,
      "grad_norm": 1.7631841897964478,
      "learning_rate": 1e-05,
      "loss": 0.3237,
      "step": 119700
    },
    {
      "epoch": 0.001198,
      "grad_norm": 1.536399245262146,
      "learning_rate": 1e-05,
      "loss": 0.3285,
      "step": 119800
    },
    {
      "epoch": 0.001199,
      "grad_norm": 1.7705007791519165,
      "learning_rate": 1e-05,
      "loss": 0.3291,
      "step": 119900
    },
    {
      "epoch": 0.0012,
      "grad_norm": 1.8081552982330322,
      "learning_rate": 1e-05,
      "loss": 0.3251,
      "step": 120000
    },
    {
      "epoch": 0.0012,
      "eval_loss": 0.301513671875,
      "eval_runtime": 114.8107,
      "eval_samples_per_second": 435.499,
      "eval_steps_per_second": 27.219,
      "step": 120000
    },
    {
      "epoch": 0.001201,
      "grad_norm": 1.8171324729919434,
      "learning_rate": 1e-05,
      "loss": 0.3342,
      "step": 120100
    },
    {
      "epoch": 0.001202,
      "grad_norm": 1.6765378713607788,
      "learning_rate": 1e-05,
      "loss": 0.3276,
      "step": 120200
    },
    {
      "epoch": 0.001203,
      "grad_norm": 1.8498486280441284,
      "learning_rate": 1e-05,
      "loss": 0.3314,
      "step": 120300
    },
    {
      "epoch": 0.001204,
      "grad_norm": 1.536418080329895,
      "learning_rate": 1e-05,
      "loss": 0.3237,
      "step": 120400
    },
    {
      "epoch": 0.001205,
      "grad_norm": 1.5730416774749756,
      "learning_rate": 1e-05,
      "loss": 0.3263,
      "step": 120500
    },
    {
      "epoch": 0.001206,
      "grad_norm": 1.7513419389724731,
      "learning_rate": 1e-05,
      "loss": 0.3355,
      "step": 120600
    },
    {
      "epoch": 0.001207,
      "grad_norm": 1.7094500064849854,
      "learning_rate": 1e-05,
      "loss": 0.3243,
      "step": 120700
    },
    {
      "epoch": 0.001208,
      "grad_norm": 1.8792259693145752,
      "learning_rate": 1e-05,
      "loss": 0.3283,
      "step": 120800
    },
    {
      "epoch": 0.001209,
      "grad_norm": 1.7763640880584717,
      "learning_rate": 1e-05,
      "loss": 0.3325,
      "step": 120900
    },
    {
      "epoch": 0.00121,
      "grad_norm": 1.8400176763534546,
      "learning_rate": 1e-05,
      "loss": 0.3342,
      "step": 121000
    },
    {
      "epoch": 0.001211,
      "grad_norm": 1.9107251167297363,
      "learning_rate": 1e-05,
      "loss": 0.3314,
      "step": 121100
    },
    {
      "epoch": 0.001212,
      "grad_norm": 1.4877129793167114,
      "learning_rate": 1e-05,
      "loss": 0.3317,
      "step": 121200
    },
    {
      "epoch": 0.001213,
      "grad_norm": 1.6370038986206055,
      "learning_rate": 1e-05,
      "loss": 0.3253,
      "step": 121300
    },
    {
      "epoch": 0.001214,
      "grad_norm": 1.9504472017288208,
      "learning_rate": 1e-05,
      "loss": 0.3242,
      "step": 121400
    },
    {
      "epoch": 0.001215,
      "grad_norm": 2.077836036682129,
      "learning_rate": 1e-05,
      "loss": 0.3302,
      "step": 121500
    },
    {
      "epoch": 0.001216,
      "grad_norm": 1.4920196533203125,
      "learning_rate": 1e-05,
      "loss": 0.332,
      "step": 121600
    },
    {
      "epoch": 0.001217,
      "grad_norm": 1.7362091541290283,
      "learning_rate": 1e-05,
      "loss": 0.3313,
      "step": 121700
    },
    {
      "epoch": 0.001218,
      "grad_norm": 1.662297248840332,
      "learning_rate": 1e-05,
      "loss": 0.3287,
      "step": 121800
    },
    {
      "epoch": 0.001219,
      "grad_norm": 1.6544877290725708,
      "learning_rate": 1e-05,
      "loss": 0.3266,
      "step": 121900
    },
    {
      "epoch": 0.00122,
      "grad_norm": 1.5775099992752075,
      "learning_rate": 1e-05,
      "loss": 0.3263,
      "step": 122000
    },
    {
      "epoch": 0.001221,
      "grad_norm": 1.704094648361206,
      "learning_rate": 1e-05,
      "loss": 0.325,
      "step": 122100
    },
    {
      "epoch": 0.001222,
      "grad_norm": 1.9005615711212158,
      "learning_rate": 1e-05,
      "loss": 0.3268,
      "step": 122200
    },
    {
      "epoch": 0.001223,
      "grad_norm": 2.027251958847046,
      "learning_rate": 1e-05,
      "loss": 0.3266,
      "step": 122300
    },
    {
      "epoch": 0.001224,
      "grad_norm": 1.8168870210647583,
      "learning_rate": 1e-05,
      "loss": 0.3298,
      "step": 122400
    },
    {
      "epoch": 0.001225,
      "grad_norm": 1.709088921546936,
      "learning_rate": 1e-05,
      "loss": 0.322,
      "step": 122500
    },
    {
      "epoch": 0.001226,
      "grad_norm": 1.7680180072784424,
      "learning_rate": 1e-05,
      "loss": 0.3362,
      "step": 122600
    },
    {
      "epoch": 0.001227,
      "grad_norm": 1.681280493736267,
      "learning_rate": 1e-05,
      "loss": 0.3236,
      "step": 122700
    },
    {
      "epoch": 0.001228,
      "grad_norm": 1.7112354040145874,
      "learning_rate": 1e-05,
      "loss": 0.3308,
      "step": 122800
    },
    {
      "epoch": 0.001229,
      "grad_norm": 1.623887062072754,
      "learning_rate": 1e-05,
      "loss": 0.3255,
      "step": 122900
    },
    {
      "epoch": 0.00123,
      "grad_norm": 1.880348801612854,
      "learning_rate": 1e-05,
      "loss": 0.3348,
      "step": 123000
    },
    {
      "epoch": 0.001231,
      "grad_norm": 1.8015272617340088,
      "learning_rate": 1e-05,
      "loss": 0.3263,
      "step": 123100
    },
    {
      "epoch": 0.001232,
      "grad_norm": 1.794119954109192,
      "learning_rate": 1e-05,
      "loss": 0.3215,
      "step": 123200
    },
    {
      "epoch": 0.001233,
      "grad_norm": 1.3672672510147095,
      "learning_rate": 1e-05,
      "loss": 0.3246,
      "step": 123300
    },
    {
      "epoch": 0.001234,
      "grad_norm": 1.702120304107666,
      "learning_rate": 1e-05,
      "loss": 0.3277,
      "step": 123400
    },
    {
      "epoch": 0.001235,
      "grad_norm": 1.6856110095977783,
      "learning_rate": 1e-05,
      "loss": 0.3308,
      "step": 123500
    },
    {
      "epoch": 0.001236,
      "grad_norm": 1.3940743207931519,
      "learning_rate": 1e-05,
      "loss": 0.3224,
      "step": 123600
    },
    {
      "epoch": 0.001237,
      "grad_norm": 1.4997862577438354,
      "learning_rate": 1e-05,
      "loss": 0.3209,
      "step": 123700
    },
    {
      "epoch": 0.001238,
      "grad_norm": 1.6719286441802979,
      "learning_rate": 1e-05,
      "loss": 0.3287,
      "step": 123800
    },
    {
      "epoch": 0.001239,
      "grad_norm": 1.4933640956878662,
      "learning_rate": 1e-05,
      "loss": 0.3275,
      "step": 123900
    },
    {
      "epoch": 0.00124,
      "grad_norm": 1.6647841930389404,
      "learning_rate": 1e-05,
      "loss": 0.3217,
      "step": 124000
    },
    {
      "epoch": 0.001241,
      "grad_norm": 1.656747817993164,
      "learning_rate": 1e-05,
      "loss": 0.3185,
      "step": 124100
    },
    {
      "epoch": 0.001242,
      "grad_norm": 1.7526649236679077,
      "learning_rate": 1e-05,
      "loss": 0.3238,
      "step": 124200
    },
    {
      "epoch": 0.001243,
      "grad_norm": 1.5294679403305054,
      "learning_rate": 1e-05,
      "loss": 0.3282,
      "step": 124300
    },
    {
      "epoch": 0.001244,
      "grad_norm": 1.5839802026748657,
      "learning_rate": 1e-05,
      "loss": 0.3191,
      "step": 124400
    },
    {
      "epoch": 0.001245,
      "grad_norm": 2.2223639488220215,
      "learning_rate": 1e-05,
      "loss": 0.3296,
      "step": 124500
    },
    {
      "epoch": 0.001246,
      "grad_norm": 1.7450740337371826,
      "learning_rate": 1e-05,
      "loss": 0.3275,
      "step": 124600
    },
    {
      "epoch": 0.001247,
      "grad_norm": 1.591633677482605,
      "learning_rate": 1e-05,
      "loss": 0.3152,
      "step": 124700
    },
    {
      "epoch": 0.001248,
      "grad_norm": 1.8334294557571411,
      "learning_rate": 1e-05,
      "loss": 0.325,
      "step": 124800
    },
    {
      "epoch": 0.001249,
      "grad_norm": 1.8498133420944214,
      "learning_rate": 1e-05,
      "loss": 0.3267,
      "step": 124900
    },
    {
      "epoch": 0.00125,
      "grad_norm": 1.8751991987228394,
      "learning_rate": 1e-05,
      "loss": 0.329,
      "step": 125000
    },
    {
      "epoch": 0.001251,
      "grad_norm": 1.7075512409210205,
      "learning_rate": 1e-05,
      "loss": 0.3233,
      "step": 125100
    },
    {
      "epoch": 0.001252,
      "grad_norm": 1.7127974033355713,
      "learning_rate": 1e-05,
      "loss": 0.3256,
      "step": 125200
    },
    {
      "epoch": 0.001253,
      "grad_norm": 1.6013959646224976,
      "learning_rate": 1e-05,
      "loss": 0.3294,
      "step": 125300
    },
    {
      "epoch": 0.001254,
      "grad_norm": 1.616365671157837,
      "learning_rate": 1e-05,
      "loss": 0.3256,
      "step": 125400
    },
    {
      "epoch": 0.001255,
      "grad_norm": 1.9536734819412231,
      "learning_rate": 1e-05,
      "loss": 0.3268,
      "step": 125500
    },
    {
      "epoch": 0.001256,
      "grad_norm": 1.725591778755188,
      "learning_rate": 1e-05,
      "loss": 0.329,
      "step": 125600
    },
    {
      "epoch": 0.001257,
      "grad_norm": 1.4925453662872314,
      "learning_rate": 1e-05,
      "loss": 0.3284,
      "step": 125700
    },
    {
      "epoch": 0.001258,
      "grad_norm": 1.718002438545227,
      "learning_rate": 1e-05,
      "loss": 0.3269,
      "step": 125800
    },
    {
      "epoch": 0.001259,
      "grad_norm": 1.6296608448028564,
      "learning_rate": 1e-05,
      "loss": 0.3187,
      "step": 125900
    },
    {
      "epoch": 0.00126,
      "grad_norm": 1.584074854850769,
      "learning_rate": 1e-05,
      "loss": 0.3239,
      "step": 126000
    },
    {
      "epoch": 0.001261,
      "grad_norm": 1.7755796909332275,
      "learning_rate": 1e-05,
      "loss": 0.3231,
      "step": 126100
    },
    {
      "epoch": 0.001262,
      "grad_norm": 1.5260752439498901,
      "learning_rate": 1e-05,
      "loss": 0.3227,
      "step": 126200
    },
    {
      "epoch": 0.001263,
      "grad_norm": 1.7237966060638428,
      "learning_rate": 1e-05,
      "loss": 0.3281,
      "step": 126300
    },
    {
      "epoch": 0.001264,
      "grad_norm": 1.6392747163772583,
      "learning_rate": 1e-05,
      "loss": 0.3211,
      "step": 126400
    },
    {
      "epoch": 0.001265,
      "grad_norm": 1.8122193813323975,
      "learning_rate": 1e-05,
      "loss": 0.3264,
      "step": 126500
    },
    {
      "epoch": 0.001266,
      "grad_norm": 1.8250197172164917,
      "learning_rate": 1e-05,
      "loss": 0.32,
      "step": 126600
    },
    {
      "epoch": 0.001267,
      "grad_norm": 1.7046698331832886,
      "learning_rate": 1e-05,
      "loss": 0.3254,
      "step": 126700
    },
    {
      "epoch": 0.001268,
      "grad_norm": 1.74703049659729,
      "learning_rate": 1e-05,
      "loss": 0.3239,
      "step": 126800
    },
    {
      "epoch": 0.001269,
      "grad_norm": 1.683023452758789,
      "learning_rate": 1e-05,
      "loss": 0.3318,
      "step": 126900
    },
    {
      "epoch": 0.00127,
      "grad_norm": 1.9313528537750244,
      "learning_rate": 1e-05,
      "loss": 0.3203,
      "step": 127000
    },
    {
      "epoch": 0.001271,
      "grad_norm": 1.3307342529296875,
      "learning_rate": 1e-05,
      "loss": 0.327,
      "step": 127100
    },
    {
      "epoch": 0.001272,
      "grad_norm": 1.817285418510437,
      "learning_rate": 1e-05,
      "loss": 0.3352,
      "step": 127200
    },
    {
      "epoch": 0.001273,
      "grad_norm": 1.7280235290527344,
      "learning_rate": 1e-05,
      "loss": 0.3207,
      "step": 127300
    },
    {
      "epoch": 0.001274,
      "grad_norm": 1.7646243572235107,
      "learning_rate": 1e-05,
      "loss": 0.3303,
      "step": 127400
    },
    {
      "epoch": 0.001275,
      "grad_norm": 1.6660072803497314,
      "learning_rate": 1e-05,
      "loss": 0.3262,
      "step": 127500
    },
    {
      "epoch": 0.001276,
      "grad_norm": 1.8076339960098267,
      "learning_rate": 1e-05,
      "loss": 0.3195,
      "step": 127600
    },
    {
      "epoch": 0.001277,
      "grad_norm": 1.6329010725021362,
      "learning_rate": 1e-05,
      "loss": 0.3166,
      "step": 127700
    },
    {
      "epoch": 0.001278,
      "grad_norm": 1.8912221193313599,
      "learning_rate": 1e-05,
      "loss": 0.3253,
      "step": 127800
    },
    {
      "epoch": 0.001279,
      "grad_norm": 1.6600183248519897,
      "learning_rate": 1e-05,
      "loss": 0.3328,
      "step": 127900
    },
    {
      "epoch": 0.00128,
      "grad_norm": 2.2682251930236816,
      "learning_rate": 1e-05,
      "loss": 0.322,
      "step": 128000
    },
    {
      "epoch": 0.001281,
      "grad_norm": 1.9191845655441284,
      "learning_rate": 1e-05,
      "loss": 0.3236,
      "step": 128100
    },
    {
      "epoch": 0.001282,
      "grad_norm": 1.8286381959915161,
      "learning_rate": 1e-05,
      "loss": 0.3161,
      "step": 128200
    },
    {
      "epoch": 0.001283,
      "grad_norm": 1.5892844200134277,
      "learning_rate": 1e-05,
      "loss": 0.3176,
      "step": 128300
    },
    {
      "epoch": 0.001284,
      "grad_norm": 1.5269801616668701,
      "learning_rate": 1e-05,
      "loss": 0.3213,
      "step": 128400
    },
    {
      "epoch": 0.001285,
      "grad_norm": 1.9540048837661743,
      "learning_rate": 1e-05,
      "loss": 0.3254,
      "step": 128500
    },
    {
      "epoch": 0.001286,
      "grad_norm": 1.669771671295166,
      "learning_rate": 1e-05,
      "loss": 0.3218,
      "step": 128600
    },
    {
      "epoch": 0.001287,
      "grad_norm": 1.8254411220550537,
      "learning_rate": 1e-05,
      "loss": 0.3241,
      "step": 128700
    },
    {
      "epoch": 0.001288,
      "grad_norm": 1.9790143966674805,
      "learning_rate": 1e-05,
      "loss": 0.323,
      "step": 128800
    },
    {
      "epoch": 0.001289,
      "grad_norm": 1.7570897340774536,
      "learning_rate": 1e-05,
      "loss": 0.3233,
      "step": 128900
    },
    {
      "epoch": 0.00129,
      "grad_norm": 1.662227988243103,
      "learning_rate": 1e-05,
      "loss": 0.3232,
      "step": 129000
    },
    {
      "epoch": 0.001291,
      "grad_norm": 1.6427834033966064,
      "learning_rate": 1e-05,
      "loss": 0.3163,
      "step": 129100
    },
    {
      "epoch": 0.001292,
      "grad_norm": 1.74818754196167,
      "learning_rate": 1e-05,
      "loss": 0.3221,
      "step": 129200
    },
    {
      "epoch": 0.001293,
      "grad_norm": 1.4715895652770996,
      "learning_rate": 1e-05,
      "loss": 0.3199,
      "step": 129300
    },
    {
      "epoch": 0.001294,
      "grad_norm": 1.5854099988937378,
      "learning_rate": 1e-05,
      "loss": 0.3276,
      "step": 129400
    },
    {
      "epoch": 0.001295,
      "grad_norm": 1.6624189615249634,
      "learning_rate": 1e-05,
      "loss": 0.3206,
      "step": 129500
    },
    {
      "epoch": 0.001296,
      "grad_norm": 1.5426676273345947,
      "learning_rate": 1e-05,
      "loss": 0.3215,
      "step": 129600
    },
    {
      "epoch": 0.001297,
      "grad_norm": 1.7038050889968872,
      "learning_rate": 1e-05,
      "loss": 0.3225,
      "step": 129700
    },
    {
      "epoch": 0.001298,
      "grad_norm": 1.4283004999160767,
      "learning_rate": 1e-05,
      "loss": 0.3216,
      "step": 129800
    },
    {
      "epoch": 0.001299,
      "grad_norm": 1.5062975883483887,
      "learning_rate": 1e-05,
      "loss": 0.3221,
      "step": 129900
    },
    {
      "epoch": 0.0013,
      "grad_norm": 1.7706925868988037,
      "learning_rate": 1e-05,
      "loss": 0.3205,
      "step": 130000
    },
    {
      "epoch": 0.001301,
      "grad_norm": 1.6793439388275146,
      "learning_rate": 1e-05,
      "loss": 0.3186,
      "step": 130100
    },
    {
      "epoch": 0.001302,
      "grad_norm": 1.5680670738220215,
      "learning_rate": 1e-05,
      "loss": 0.3203,
      "step": 130200
    },
    {
      "epoch": 0.001303,
      "grad_norm": 1.8509889841079712,
      "learning_rate": 1e-05,
      "loss": 0.3293,
      "step": 130300
    },
    {
      "epoch": 0.001304,
      "grad_norm": 1.6747349500656128,
      "learning_rate": 1e-05,
      "loss": 0.3208,
      "step": 130400
    },
    {
      "epoch": 0.001305,
      "grad_norm": 1.7840492725372314,
      "learning_rate": 1e-05,
      "loss": 0.3176,
      "step": 130500
    },
    {
      "epoch": 0.001306,
      "grad_norm": 1.4201829433441162,
      "learning_rate": 1e-05,
      "loss": 0.3198,
      "step": 130600
    },
    {
      "epoch": 0.001307,
      "grad_norm": 1.5987930297851562,
      "learning_rate": 1e-05,
      "loss": 0.3276,
      "step": 130700
    },
    {
      "epoch": 0.001308,
      "grad_norm": 1.5990506410598755,
      "learning_rate": 1e-05,
      "loss": 0.3216,
      "step": 130800
    },
    {
      "epoch": 0.001309,
      "grad_norm": 1.563931941986084,
      "learning_rate": 1e-05,
      "loss": 0.3202,
      "step": 130900
    },
    {
      "epoch": 0.00131,
      "grad_norm": 1.9966181516647339,
      "learning_rate": 1e-05,
      "loss": 0.3173,
      "step": 131000
    },
    {
      "epoch": 0.001311,
      "grad_norm": 1.6617968082427979,
      "learning_rate": 1e-05,
      "loss": 0.3241,
      "step": 131100
    },
    {
      "epoch": 0.001312,
      "grad_norm": 1.903935432434082,
      "learning_rate": 1e-05,
      "loss": 0.3215,
      "step": 131200
    },
    {
      "epoch": 0.001313,
      "grad_norm": 1.9104382991790771,
      "learning_rate": 1e-05,
      "loss": 0.3183,
      "step": 131300
    },
    {
      "epoch": 0.001314,
      "grad_norm": 1.5082734823226929,
      "learning_rate": 1e-05,
      "loss": 0.3251,
      "step": 131400
    },
    {
      "epoch": 0.001315,
      "grad_norm": 1.869626522064209,
      "learning_rate": 1e-05,
      "loss": 0.3206,
      "step": 131500
    },
    {
      "epoch": 0.001316,
      "grad_norm": 1.8665653467178345,
      "learning_rate": 1e-05,
      "loss": 0.3178,
      "step": 131600
    },
    {
      "epoch": 0.001317,
      "grad_norm": 1.52765691280365,
      "learning_rate": 1e-05,
      "loss": 0.321,
      "step": 131700
    },
    {
      "epoch": 0.001318,
      "grad_norm": 1.4063327312469482,
      "learning_rate": 1e-05,
      "loss": 0.3158,
      "step": 131800
    },
    {
      "epoch": 0.001319,
      "grad_norm": 1.8856468200683594,
      "learning_rate": 1e-05,
      "loss": 0.3144,
      "step": 131900
    },
    {
      "epoch": 0.00132,
      "grad_norm": 1.7623271942138672,
      "learning_rate": 1e-05,
      "loss": 0.3224,
      "step": 132000
    },
    {
      "epoch": 0.001321,
      "grad_norm": 1.6237845420837402,
      "learning_rate": 1e-05,
      "loss": 0.3207,
      "step": 132100
    },
    {
      "epoch": 0.001322,
      "grad_norm": 1.55039644241333,
      "learning_rate": 1e-05,
      "loss": 0.3188,
      "step": 132200
    },
    {
      "epoch": 0.001323,
      "grad_norm": 1.6823863983154297,
      "learning_rate": 1e-05,
      "loss": 0.3204,
      "step": 132300
    },
    {
      "epoch": 0.001324,
      "grad_norm": 1.6407947540283203,
      "learning_rate": 1e-05,
      "loss": 0.3129,
      "step": 132400
    },
    {
      "epoch": 0.001325,
      "grad_norm": 1.6221039295196533,
      "learning_rate": 1e-05,
      "loss": 0.3184,
      "step": 132500
    },
    {
      "epoch": 0.001326,
      "grad_norm": 1.6350358724594116,
      "learning_rate": 1e-05,
      "loss": 0.312,
      "step": 132600
    },
    {
      "epoch": 0.001327,
      "grad_norm": 1.5436922311782837,
      "learning_rate": 1e-05,
      "loss": 0.3194,
      "step": 132700
    },
    {
      "epoch": 0.001328,
      "grad_norm": 1.4995625019073486,
      "learning_rate": 1e-05,
      "loss": 0.3202,
      "step": 132800
    },
    {
      "epoch": 0.001329,
      "grad_norm": 1.4857275485992432,
      "learning_rate": 1e-05,
      "loss": 0.3263,
      "step": 132900
    },
    {
      "epoch": 0.00133,
      "grad_norm": 1.6003996133804321,
      "learning_rate": 1e-05,
      "loss": 0.3185,
      "step": 133000
    },
    {
      "epoch": 0.001331,
      "grad_norm": 1.6696702241897583,
      "learning_rate": 1e-05,
      "loss": 0.3217,
      "step": 133100
    },
    {
      "epoch": 0.001332,
      "grad_norm": 1.5747042894363403,
      "learning_rate": 1e-05,
      "loss": 0.3193,
      "step": 133200
    },
    {
      "epoch": 0.001333,
      "grad_norm": 1.5583536624908447,
      "learning_rate": 1e-05,
      "loss": 0.3216,
      "step": 133300
    },
    {
      "epoch": 0.001334,
      "grad_norm": 1.7547342777252197,
      "learning_rate": 1e-05,
      "loss": 0.3156,
      "step": 133400
    },
    {
      "epoch": 0.001335,
      "grad_norm": 1.4350770711898804,
      "learning_rate": 1e-05,
      "loss": 0.3104,
      "step": 133500
    },
    {
      "epoch": 0.001336,
      "grad_norm": 1.7420467138290405,
      "learning_rate": 1e-05,
      "loss": 0.3268,
      "step": 133600
    },
    {
      "epoch": 0.001337,
      "grad_norm": 1.7873330116271973,
      "learning_rate": 1e-05,
      "loss": 0.3225,
      "step": 133700
    },
    {
      "epoch": 0.001338,
      "grad_norm": 1.7741400003433228,
      "learning_rate": 1e-05,
      "loss": 0.322,
      "step": 133800
    },
    {
      "epoch": 0.001339,
      "grad_norm": 1.6877092123031616,
      "learning_rate": 1e-05,
      "loss": 0.3209,
      "step": 133900
    },
    {
      "epoch": 0.00134,
      "grad_norm": 1.7802987098693848,
      "learning_rate": 1e-05,
      "loss": 0.3206,
      "step": 134000
    },
    {
      "epoch": 0.001341,
      "grad_norm": 1.8283404111862183,
      "learning_rate": 1e-05,
      "loss": 0.3256,
      "step": 134100
    },
    {
      "epoch": 0.001342,
      "grad_norm": 1.5722459554672241,
      "learning_rate": 1e-05,
      "loss": 0.3163,
      "step": 134200
    },
    {
      "epoch": 0.001343,
      "grad_norm": 1.7147879600524902,
      "learning_rate": 1e-05,
      "loss": 0.3161,
      "step": 134300
    },
    {
      "epoch": 0.001344,
      "grad_norm": 1.8033101558685303,
      "learning_rate": 1e-05,
      "loss": 0.3146,
      "step": 134400
    },
    {
      "epoch": 0.001345,
      "grad_norm": 1.5288691520690918,
      "learning_rate": 1e-05,
      "loss": 0.3225,
      "step": 134500
    },
    {
      "epoch": 0.001346,
      "grad_norm": 1.6482738256454468,
      "learning_rate": 1e-05,
      "loss": 0.3136,
      "step": 134600
    },
    {
      "epoch": 0.001347,
      "grad_norm": 1.5121197700500488,
      "learning_rate": 1e-05,
      "loss": 0.3229,
      "step": 134700
    },
    {
      "epoch": 0.001348,
      "grad_norm": 1.602310061454773,
      "learning_rate": 1e-05,
      "loss": 0.3183,
      "step": 134800
    },
    {
      "epoch": 0.001349,
      "grad_norm": 1.596355676651001,
      "learning_rate": 1e-05,
      "loss": 0.3099,
      "step": 134900
    },
    {
      "epoch": 0.00135,
      "grad_norm": 2.1672589778900146,
      "learning_rate": 1e-05,
      "loss": 0.3183,
      "step": 135000
    },
    {
      "epoch": 0.001351,
      "grad_norm": 1.529840350151062,
      "learning_rate": 1e-05,
      "loss": 0.322,
      "step": 135100
    },
    {
      "epoch": 0.001352,
      "grad_norm": 1.48147714138031,
      "learning_rate": 1e-05,
      "loss": 0.3178,
      "step": 135200
    },
    {
      "epoch": 0.001353,
      "grad_norm": 1.488888144493103,
      "learning_rate": 1e-05,
      "loss": 0.3121,
      "step": 135300
    },
    {
      "epoch": 0.001354,
      "grad_norm": 1.3818961381912231,
      "learning_rate": 1e-05,
      "loss": 0.325,
      "step": 135400
    },
    {
      "epoch": 0.001355,
      "grad_norm": 1.6328446865081787,
      "learning_rate": 1e-05,
      "loss": 0.3189,
      "step": 135500
    },
    {
      "epoch": 0.001356,
      "grad_norm": 1.7009713649749756,
      "learning_rate": 1e-05,
      "loss": 0.3188,
      "step": 135600
    },
    {
      "epoch": 0.001357,
      "grad_norm": 1.5784467458724976,
      "learning_rate": 1e-05,
      "loss": 0.3221,
      "step": 135700
    },
    {
      "epoch": 0.001358,
      "grad_norm": 1.5164488554000854,
      "learning_rate": 1e-05,
      "loss": 0.3105,
      "step": 135800
    },
    {
      "epoch": 0.001359,
      "grad_norm": 1.668848991394043,
      "learning_rate": 1e-05,
      "loss": 0.3104,
      "step": 135900
    },
    {
      "epoch": 0.00136,
      "grad_norm": 1.978113055229187,
      "learning_rate": 1e-05,
      "loss": 0.3151,
      "step": 136000
    },
    {
      "epoch": 0.001361,
      "grad_norm": 1.6246618032455444,
      "learning_rate": 1e-05,
      "loss": 0.3165,
      "step": 136100
    },
    {
      "epoch": 0.001362,
      "grad_norm": 1.5396642684936523,
      "learning_rate": 1e-05,
      "loss": 0.3154,
      "step": 136200
    },
    {
      "epoch": 0.001363,
      "grad_norm": 1.660476565361023,
      "learning_rate": 1e-05,
      "loss": 0.313,
      "step": 136300
    },
    {
      "epoch": 0.001364,
      "grad_norm": 1.4768214225769043,
      "learning_rate": 1e-05,
      "loss": 0.3171,
      "step": 136400
    },
    {
      "epoch": 0.001365,
      "grad_norm": 1.7115814685821533,
      "learning_rate": 1e-05,
      "loss": 0.3133,
      "step": 136500
    },
    {
      "epoch": 0.001366,
      "grad_norm": 1.6259845495224,
      "learning_rate": 1e-05,
      "loss": 0.3153,
      "step": 136600
    },
    {
      "epoch": 0.001367,
      "grad_norm": 1.5243622064590454,
      "learning_rate": 1e-05,
      "loss": 0.3151,
      "step": 136700
    },
    {
      "epoch": 0.001368,
      "grad_norm": 1.8191895484924316,
      "learning_rate": 1e-05,
      "loss": 0.3229,
      "step": 136800
    },
    {
      "epoch": 0.001369,
      "grad_norm": 1.5113294124603271,
      "learning_rate": 1e-05,
      "loss": 0.317,
      "step": 136900
    },
    {
      "epoch": 0.00137,
      "grad_norm": 1.46476149559021,
      "learning_rate": 1e-05,
      "loss": 0.3156,
      "step": 137000
    },
    {
      "epoch": 0.001371,
      "grad_norm": 1.4856374263763428,
      "learning_rate": 1e-05,
      "loss": 0.3116,
      "step": 137100
    },
    {
      "epoch": 0.001372,
      "grad_norm": 1.6964186429977417,
      "learning_rate": 1e-05,
      "loss": 0.3155,
      "step": 137200
    },
    {
      "epoch": 0.001373,
      "grad_norm": 1.8333369493484497,
      "learning_rate": 1e-05,
      "loss": 0.3142,
      "step": 137300
    },
    {
      "epoch": 0.001374,
      "grad_norm": 1.6640217304229736,
      "learning_rate": 1e-05,
      "loss": 0.3177,
      "step": 137400
    },
    {
      "epoch": 0.001375,
      "grad_norm": 1.566697597503662,
      "learning_rate": 1e-05,
      "loss": 0.3114,
      "step": 137500
    },
    {
      "epoch": 0.001376,
      "grad_norm": 1.9087796211242676,
      "learning_rate": 1e-05,
      "loss": 0.3106,
      "step": 137600
    },
    {
      "epoch": 0.001377,
      "grad_norm": 1.7321326732635498,
      "learning_rate": 1e-05,
      "loss": 0.3171,
      "step": 137700
    },
    {
      "epoch": 0.001378,
      "grad_norm": 1.8865094184875488,
      "learning_rate": 1e-05,
      "loss": 0.3154,
      "step": 137800
    },
    {
      "epoch": 0.001379,
      "grad_norm": 1.7816592454910278,
      "learning_rate": 1e-05,
      "loss": 0.3079,
      "step": 137900
    },
    {
      "epoch": 0.00138,
      "grad_norm": 1.442036509513855,
      "learning_rate": 1e-05,
      "loss": 0.3102,
      "step": 138000
    },
    {
      "epoch": 0.001381,
      "grad_norm": 1.6105217933654785,
      "learning_rate": 1e-05,
      "loss": 0.3145,
      "step": 138100
    },
    {
      "epoch": 0.001382,
      "grad_norm": 1.597864031791687,
      "learning_rate": 1e-05,
      "loss": 0.3116,
      "step": 138200
    },
    {
      "epoch": 0.001383,
      "grad_norm": 1.888089656829834,
      "learning_rate": 1e-05,
      "loss": 0.3215,
      "step": 138300
    },
    {
      "epoch": 0.001384,
      "grad_norm": 1.372367262840271,
      "learning_rate": 1e-05,
      "loss": 0.3121,
      "step": 138400
    },
    {
      "epoch": 0.001385,
      "grad_norm": 1.7237030267715454,
      "learning_rate": 1e-05,
      "loss": 0.3066,
      "step": 138500
    },
    {
      "epoch": 0.001386,
      "grad_norm": 1.8450541496276855,
      "learning_rate": 1e-05,
      "loss": 0.3197,
      "step": 138600
    },
    {
      "epoch": 0.001387,
      "grad_norm": 1.6042810678482056,
      "learning_rate": 1e-05,
      "loss": 0.3143,
      "step": 138700
    },
    {
      "epoch": 0.001388,
      "grad_norm": 1.914825677871704,
      "learning_rate": 1e-05,
      "loss": 0.3109,
      "step": 138800
    },
    {
      "epoch": 0.001389,
      "grad_norm": 1.810681939125061,
      "learning_rate": 1e-05,
      "loss": 0.3207,
      "step": 138900
    },
    {
      "epoch": 0.00139,
      "grad_norm": 1.6505619287490845,
      "learning_rate": 1e-05,
      "loss": 0.3112,
      "step": 139000
    },
    {
      "epoch": 0.001391,
      "grad_norm": 1.5920408964157104,
      "learning_rate": 1e-05,
      "loss": 0.3143,
      "step": 139100
    },
    {
      "epoch": 0.001392,
      "grad_norm": 1.603387713432312,
      "learning_rate": 1e-05,
      "loss": 0.3121,
      "step": 139200
    },
    {
      "epoch": 0.001393,
      "grad_norm": 1.7222926616668701,
      "learning_rate": 1e-05,
      "loss": 0.3161,
      "step": 139300
    },
    {
      "epoch": 0.001394,
      "grad_norm": 1.6358146667480469,
      "learning_rate": 1e-05,
      "loss": 0.3083,
      "step": 139400
    },
    {
      "epoch": 0.001395,
      "grad_norm": 2.0886547565460205,
      "learning_rate": 1e-05,
      "loss": 0.3215,
      "step": 139500
    },
    {
      "epoch": 0.001396,
      "grad_norm": 1.68631911277771,
      "learning_rate": 1e-05,
      "loss": 0.3082,
      "step": 139600
    },
    {
      "epoch": 0.001397,
      "grad_norm": 1.5744168758392334,
      "learning_rate": 1e-05,
      "loss": 0.315,
      "step": 139700
    },
    {
      "epoch": 0.001398,
      "grad_norm": 1.7238872051239014,
      "learning_rate": 1e-05,
      "loss": 0.3151,
      "step": 139800
    },
    {
      "epoch": 0.001399,
      "grad_norm": 1.6450138092041016,
      "learning_rate": 1e-05,
      "loss": 0.3129,
      "step": 139900
    },
    {
      "epoch": 0.0014,
      "grad_norm": 1.4830751419067383,
      "learning_rate": 1e-05,
      "loss": 0.3164,
      "step": 140000
    },
    {
      "epoch": 0.0014,
      "eval_loss": 0.287109375,
      "eval_runtime": 109.9314,
      "eval_samples_per_second": 454.829,
      "eval_steps_per_second": 28.427,
      "step": 140000
    },
    {
      "epoch": 0.001401,
      "grad_norm": 1.7286059856414795,
      "learning_rate": 1e-05,
      "loss": 0.311,
      "step": 140100
    },
    {
      "epoch": 0.001402,
      "grad_norm": 1.7702032327651978,
      "learning_rate": 1e-05,
      "loss": 0.326,
      "step": 140200
    },
    {
      "epoch": 0.001403,
      "grad_norm": 1.6898577213287354,
      "learning_rate": 1e-05,
      "loss": 0.3216,
      "step": 140300
    },
    {
      "epoch": 0.001404,
      "grad_norm": 1.6577975749969482,
      "learning_rate": 1e-05,
      "loss": 0.3145,
      "step": 140400
    },
    {
      "epoch": 0.001405,
      "grad_norm": 1.444854736328125,
      "learning_rate": 1e-05,
      "loss": 0.3216,
      "step": 140500
    },
    {
      "epoch": 0.001406,
      "grad_norm": 1.6251206398010254,
      "learning_rate": 1e-05,
      "loss": 0.3108,
      "step": 140600
    },
    {
      "epoch": 0.001407,
      "grad_norm": 1.4880505800247192,
      "learning_rate": 1e-05,
      "loss": 0.3131,
      "step": 140700
    },
    {
      "epoch": 0.001408,
      "grad_norm": 1.8925511837005615,
      "learning_rate": 1e-05,
      "loss": 0.3117,
      "step": 140800
    },
    {
      "epoch": 0.001409,
      "grad_norm": 1.6990015506744385,
      "learning_rate": 1e-05,
      "loss": 0.3101,
      "step": 140900
    },
    {
      "epoch": 0.00141,
      "grad_norm": 1.498661756515503,
      "learning_rate": 1e-05,
      "loss": 0.3082,
      "step": 141000
    },
    {
      "epoch": 0.001411,
      "grad_norm": 1.7527713775634766,
      "learning_rate": 1e-05,
      "loss": 0.3178,
      "step": 141100
    },
    {
      "epoch": 0.001412,
      "grad_norm": 1.6200438737869263,
      "learning_rate": 1e-05,
      "loss": 0.3228,
      "step": 141200
    },
    {
      "epoch": 0.001413,
      "grad_norm": 1.3735147714614868,
      "learning_rate": 1e-05,
      "loss": 0.3124,
      "step": 141300
    },
    {
      "epoch": 0.001414,
      "grad_norm": 2.0076656341552734,
      "learning_rate": 1e-05,
      "loss": 0.3119,
      "step": 141400
    },
    {
      "epoch": 0.001415,
      "grad_norm": 1.5308282375335693,
      "learning_rate": 1e-05,
      "loss": 0.321,
      "step": 141500
    },
    {
      "epoch": 0.001416,
      "grad_norm": 1.5367777347564697,
      "learning_rate": 1e-05,
      "loss": 0.3064,
      "step": 141600
    },
    {
      "epoch": 0.001417,
      "grad_norm": 1.64597749710083,
      "learning_rate": 1e-05,
      "loss": 0.3116,
      "step": 141700
    },
    {
      "epoch": 0.001418,
      "grad_norm": 1.976902723312378,
      "learning_rate": 1e-05,
      "loss": 0.3187,
      "step": 141800
    },
    {
      "epoch": 0.001419,
      "grad_norm": 1.8003846406936646,
      "learning_rate": 1e-05,
      "loss": 0.3106,
      "step": 141900
    },
    {
      "epoch": 0.00142,
      "grad_norm": 2.8026585578918457,
      "learning_rate": 1e-05,
      "loss": 0.3186,
      "step": 142000
    },
    {
      "epoch": 0.001421,
      "grad_norm": 3.536267042160034,
      "learning_rate": 1e-05,
      "loss": 0.3104,
      "step": 142100
    },
    {
      "epoch": 0.001422,
      "grad_norm": 1.5805248022079468,
      "learning_rate": 1e-05,
      "loss": 0.3136,
      "step": 142200
    },
    {
      "epoch": 0.001423,
      "grad_norm": 1.5467828512191772,
      "learning_rate": 1e-05,
      "loss": 0.3102,
      "step": 142300
    },
    {
      "epoch": 0.001424,
      "grad_norm": 1.5788090229034424,
      "learning_rate": 1e-05,
      "loss": 0.3153,
      "step": 142400
    },
    {
      "epoch": 0.001425,
      "grad_norm": 1.8822991847991943,
      "learning_rate": 1e-05,
      "loss": 0.3124,
      "step": 142500
    },
    {
      "epoch": 0.001426,
      "grad_norm": 1.547636866569519,
      "learning_rate": 1e-05,
      "loss": 0.3141,
      "step": 142600
    },
    {
      "epoch": 0.001427,
      "grad_norm": 1.628211498260498,
      "learning_rate": 1e-05,
      "loss": 0.3182,
      "step": 142700
    },
    {
      "epoch": 0.001428,
      "grad_norm": 1.645572304725647,
      "learning_rate": 1e-05,
      "loss": 0.3106,
      "step": 142800
    },
    {
      "epoch": 0.001429,
      "grad_norm": 1.5614272356033325,
      "learning_rate": 1e-05,
      "loss": 0.3113,
      "step": 142900
    },
    {
      "epoch": 0.00143,
      "grad_norm": 1.6089304685592651,
      "learning_rate": 1e-05,
      "loss": 0.3115,
      "step": 143000
    },
    {
      "epoch": 0.001431,
      "grad_norm": 1.7061288356781006,
      "learning_rate": 1e-05,
      "loss": 0.3097,
      "step": 143100
    },
    {
      "epoch": 0.001432,
      "grad_norm": 1.629626989364624,
      "learning_rate": 1e-05,
      "loss": 0.3127,
      "step": 143200
    },
    {
      "epoch": 0.001433,
      "grad_norm": 1.5930266380310059,
      "learning_rate": 1e-05,
      "loss": 0.3136,
      "step": 143300
    },
    {
      "epoch": 0.001434,
      "grad_norm": 1.6816060543060303,
      "learning_rate": 1e-05,
      "loss": 0.3171,
      "step": 143400
    },
    {
      "epoch": 0.001435,
      "grad_norm": 1.6803059577941895,
      "learning_rate": 1e-05,
      "loss": 0.3162,
      "step": 143500
    },
    {
      "epoch": 0.001436,
      "grad_norm": 1.4301313161849976,
      "learning_rate": 1e-05,
      "loss": 0.3009,
      "step": 143600
    },
    {
      "epoch": 0.001437,
      "grad_norm": 1.407421588897705,
      "learning_rate": 1e-05,
      "loss": 0.3078,
      "step": 143700
    },
    {
      "epoch": 0.001438,
      "grad_norm": 1.6475402116775513,
      "learning_rate": 1e-05,
      "loss": 0.3088,
      "step": 143800
    },
    {
      "epoch": 0.001439,
      "grad_norm": 1.5251747369766235,
      "learning_rate": 1e-05,
      "loss": 0.3086,
      "step": 143900
    },
    {
      "epoch": 0.00144,
      "grad_norm": 1.5712449550628662,
      "learning_rate": 1e-05,
      "loss": 0.3117,
      "step": 144000
    },
    {
      "epoch": 0.001441,
      "grad_norm": 1.6450409889221191,
      "learning_rate": 1e-05,
      "loss": 0.3131,
      "step": 144100
    },
    {
      "epoch": 0.001442,
      "grad_norm": 1.451005458831787,
      "learning_rate": 1e-05,
      "loss": 0.3117,
      "step": 144200
    },
    {
      "epoch": 0.001443,
      "grad_norm": 1.5392875671386719,
      "learning_rate": 1e-05,
      "loss": 0.3186,
      "step": 144300
    },
    {
      "epoch": 0.001444,
      "grad_norm": 1.8175650835037231,
      "learning_rate": 1e-05,
      "loss": 0.3183,
      "step": 144400
    },
    {
      "epoch": 0.001445,
      "grad_norm": 1.506216287612915,
      "learning_rate": 1e-05,
      "loss": 0.318,
      "step": 144500
    },
    {
      "epoch": 0.001446,
      "grad_norm": 1.4329332113265991,
      "learning_rate": 1e-05,
      "loss": 0.3107,
      "step": 144600
    },
    {
      "epoch": 0.001447,
      "grad_norm": 1.6957765817642212,
      "learning_rate": 1e-05,
      "loss": 0.309,
      "step": 144700
    },
    {
      "epoch": 0.001448,
      "grad_norm": 1.3159312009811401,
      "learning_rate": 1e-05,
      "loss": 0.313,
      "step": 144800
    },
    {
      "epoch": 0.001449,
      "grad_norm": 1.6114338636398315,
      "learning_rate": 1e-05,
      "loss": 0.3096,
      "step": 144900
    },
    {
      "epoch": 0.00145,
      "grad_norm": 1.8013079166412354,
      "learning_rate": 1e-05,
      "loss": 0.3154,
      "step": 145000
    },
    {
      "epoch": 0.001451,
      "grad_norm": 1.7098653316497803,
      "learning_rate": 1e-05,
      "loss": 0.3091,
      "step": 145100
    },
    {
      "epoch": 0.001452,
      "grad_norm": 1.5512733459472656,
      "learning_rate": 1e-05,
      "loss": 0.3109,
      "step": 145200
    },
    {
      "epoch": 0.001453,
      "grad_norm": 1.725237488746643,
      "learning_rate": 1e-05,
      "loss": 0.3048,
      "step": 145300
    },
    {
      "epoch": 0.001454,
      "grad_norm": 4.254234313964844,
      "learning_rate": 1e-05,
      "loss": 0.3154,
      "step": 145400
    },
    {
      "epoch": 0.001455,
      "grad_norm": 1.7910503149032593,
      "learning_rate": 1e-05,
      "loss": 0.3085,
      "step": 145500
    },
    {
      "epoch": 0.001456,
      "grad_norm": 1.4521243572235107,
      "learning_rate": 1e-05,
      "loss": 0.316,
      "step": 145600
    },
    {
      "epoch": 0.001457,
      "grad_norm": 1.7298940420150757,
      "learning_rate": 1e-05,
      "loss": 0.3052,
      "step": 145700
    },
    {
      "epoch": 0.001458,
      "grad_norm": 1.7451497316360474,
      "learning_rate": 1e-05,
      "loss": 0.3111,
      "step": 145800
    },
    {
      "epoch": 0.001459,
      "grad_norm": 1.7721543312072754,
      "learning_rate": 1e-05,
      "loss": 0.3124,
      "step": 145900
    },
    {
      "epoch": 0.00146,
      "grad_norm": 2.228154182434082,
      "learning_rate": 1e-05,
      "loss": 0.3114,
      "step": 146000
    },
    {
      "epoch": 0.001461,
      "grad_norm": 1.6171804666519165,
      "learning_rate": 1e-05,
      "loss": 0.3028,
      "step": 146100
    },
    {
      "epoch": 0.001462,
      "grad_norm": 1.8199244737625122,
      "learning_rate": 1e-05,
      "loss": 0.3165,
      "step": 146200
    },
    {
      "epoch": 0.001463,
      "grad_norm": 1.5894031524658203,
      "learning_rate": 1e-05,
      "loss": 0.3069,
      "step": 146300
    },
    {
      "epoch": 0.001464,
      "grad_norm": 1.5978094339370728,
      "learning_rate": 1e-05,
      "loss": 0.3092,
      "step": 146400
    },
    {
      "epoch": 0.001465,
      "grad_norm": 1.4421255588531494,
      "learning_rate": 1e-05,
      "loss": 0.312,
      "step": 146500
    },
    {
      "epoch": 0.001466,
      "grad_norm": 1.6096898317337036,
      "learning_rate": 1e-05,
      "loss": 0.3049,
      "step": 146600
    },
    {
      "epoch": 0.001467,
      "grad_norm": 1.5457234382629395,
      "learning_rate": 1e-05,
      "loss": 0.3096,
      "step": 146700
    },
    {
      "epoch": 0.001468,
      "grad_norm": 1.6169909238815308,
      "learning_rate": 1e-05,
      "loss": 0.3086,
      "step": 146800
    },
    {
      "epoch": 0.001469,
      "grad_norm": 1.7878210544586182,
      "learning_rate": 1e-05,
      "loss": 0.315,
      "step": 146900
    },
    {
      "epoch": 0.00147,
      "grad_norm": 6.46168327331543,
      "learning_rate": 1e-05,
      "loss": 0.305,
      "step": 147000
    },
    {
      "epoch": 0.001471,
      "grad_norm": 1.6731548309326172,
      "learning_rate": 1e-05,
      "loss": 0.3102,
      "step": 147100
    },
    {
      "epoch": 0.001472,
      "grad_norm": 1.9476267099380493,
      "learning_rate": 1e-05,
      "loss": 0.3122,
      "step": 147200
    },
    {
      "epoch": 0.001473,
      "grad_norm": 1.6077353954315186,
      "learning_rate": 1e-05,
      "loss": 0.3084,
      "step": 147300
    },
    {
      "epoch": 0.001474,
      "grad_norm": 1.7394909858703613,
      "learning_rate": 1e-05,
      "loss": 0.3063,
      "step": 147400
    },
    {
      "epoch": 0.001475,
      "grad_norm": 1.4782484769821167,
      "learning_rate": 1e-05,
      "loss": 0.3105,
      "step": 147500
    },
    {
      "epoch": 0.001476,
      "grad_norm": 1.6981760263442993,
      "learning_rate": 1e-05,
      "loss": 0.3119,
      "step": 147600
    },
    {
      "epoch": 0.001477,
      "grad_norm": 1.697596549987793,
      "learning_rate": 1e-05,
      "loss": 0.3132,
      "step": 147700
    },
    {
      "epoch": 0.001478,
      "grad_norm": 1.630706787109375,
      "learning_rate": 1e-05,
      "loss": 0.3147,
      "step": 147800
    },
    {
      "epoch": 0.001479,
      "grad_norm": 1.685257077217102,
      "learning_rate": 1e-05,
      "loss": 0.3096,
      "step": 147900
    },
    {
      "epoch": 0.00148,
      "grad_norm": 1.8165167570114136,
      "learning_rate": 1e-05,
      "loss": 0.307,
      "step": 148000
    },
    {
      "epoch": 0.001481,
      "grad_norm": 1.6408799886703491,
      "learning_rate": 1e-05,
      "loss": 0.3111,
      "step": 148100
    },
    {
      "epoch": 0.001482,
      "grad_norm": 1.6128547191619873,
      "learning_rate": 1e-05,
      "loss": 0.3083,
      "step": 148200
    },
    {
      "epoch": 0.001483,
      "grad_norm": 1.8511683940887451,
      "learning_rate": 1e-05,
      "loss": 0.3014,
      "step": 148300
    },
    {
      "epoch": 0.001484,
      "grad_norm": 1.576206922531128,
      "learning_rate": 1e-05,
      "loss": 0.309,
      "step": 148400
    },
    {
      "epoch": 0.001485,
      "grad_norm": 1.543514370918274,
      "learning_rate": 1e-05,
      "loss": 0.3093,
      "step": 148500
    },
    {
      "epoch": 0.001486,
      "grad_norm": 1.5939360857009888,
      "learning_rate": 1e-05,
      "loss": 0.3105,
      "step": 148600
    },
    {
      "epoch": 0.001487,
      "grad_norm": 1.4022550582885742,
      "learning_rate": 1e-05,
      "loss": 0.311,
      "step": 148700
    },
    {
      "epoch": 0.001488,
      "grad_norm": 1.8196625709533691,
      "learning_rate": 1e-05,
      "loss": 0.3119,
      "step": 148800
    },
    {
      "epoch": 0.001489,
      "grad_norm": 1.6308430433273315,
      "learning_rate": 1e-05,
      "loss": 0.3042,
      "step": 148900
    },
    {
      "epoch": 0.00149,
      "grad_norm": 1.6367475986480713,
      "learning_rate": 1e-05,
      "loss": 0.3142,
      "step": 149000
    },
    {
      "epoch": 0.001491,
      "grad_norm": 1.5516581535339355,
      "learning_rate": 1e-05,
      "loss": 0.3069,
      "step": 149100
    },
    {
      "epoch": 0.001492,
      "grad_norm": 1.524357557296753,
      "learning_rate": 1e-05,
      "loss": 0.3031,
      "step": 149200
    },
    {
      "epoch": 0.001493,
      "grad_norm": 1.6461905241012573,
      "learning_rate": 1e-05,
      "loss": 0.3083,
      "step": 149300
    },
    {
      "epoch": 0.001494,
      "grad_norm": 1.8886070251464844,
      "learning_rate": 1e-05,
      "loss": 0.3137,
      "step": 149400
    },
    {
      "epoch": 0.001495,
      "grad_norm": 1.7399191856384277,
      "learning_rate": 1e-05,
      "loss": 0.3098,
      "step": 149500
    },
    {
      "epoch": 0.001496,
      "grad_norm": 1.7540628910064697,
      "learning_rate": 1e-05,
      "loss": 0.3155,
      "step": 149600
    },
    {
      "epoch": 0.001497,
      "grad_norm": 1.6544119119644165,
      "learning_rate": 1e-05,
      "loss": 0.3034,
      "step": 149700
    },
    {
      "epoch": 0.001498,
      "grad_norm": 1.6985324621200562,
      "learning_rate": 1e-05,
      "loss": 0.3041,
      "step": 149800
    },
    {
      "epoch": 0.001499,
      "grad_norm": 1.6984387636184692,
      "learning_rate": 1e-05,
      "loss": 0.3041,
      "step": 149900
    },
    {
      "epoch": 0.0015,
      "grad_norm": 1.5459750890731812,
      "learning_rate": 1e-05,
      "loss": 0.3055,
      "step": 150000
    },
    {
      "epoch": 0.001501,
      "grad_norm": 1.6217613220214844,
      "learning_rate": 1e-05,
      "loss": 0.3083,
      "step": 150100
    },
    {
      "epoch": 0.001502,
      "grad_norm": 1.6481753587722778,
      "learning_rate": 1e-05,
      "loss": 0.3045,
      "step": 150200
    },
    {
      "epoch": 0.001503,
      "grad_norm": 1.5565217733383179,
      "learning_rate": 1e-05,
      "loss": 0.3099,
      "step": 150300
    },
    {
      "epoch": 0.001504,
      "grad_norm": 1.678059458732605,
      "learning_rate": 1e-05,
      "loss": 0.3116,
      "step": 150400
    },
    {
      "epoch": 0.001505,
      "grad_norm": 1.6894927024841309,
      "learning_rate": 1e-05,
      "loss": 0.3101,
      "step": 150500
    },
    {
      "epoch": 0.001506,
      "grad_norm": 1.5071243047714233,
      "learning_rate": 1e-05,
      "loss": 0.3049,
      "step": 150600
    },
    {
      "epoch": 0.001507,
      "grad_norm": 1.6531084775924683,
      "learning_rate": 1e-05,
      "loss": 0.31,
      "step": 150700
    },
    {
      "epoch": 0.001508,
      "grad_norm": 1.5029364824295044,
      "learning_rate": 1e-05,
      "loss": 0.2991,
      "step": 150800
    },
    {
      "epoch": 0.001509,
      "grad_norm": 1.687752366065979,
      "learning_rate": 1e-05,
      "loss": 0.3079,
      "step": 150900
    },
    {
      "epoch": 0.00151,
      "grad_norm": 1.485236406326294,
      "learning_rate": 1e-05,
      "loss": 0.2977,
      "step": 151000
    },
    {
      "epoch": 0.001511,
      "grad_norm": 1.5481332540512085,
      "learning_rate": 1e-05,
      "loss": 0.3076,
      "step": 151100
    },
    {
      "epoch": 0.001512,
      "grad_norm": 1.5191718339920044,
      "learning_rate": 1e-05,
      "loss": 0.302,
      "step": 151200
    },
    {
      "epoch": 0.001513,
      "grad_norm": 1.6339939832687378,
      "learning_rate": 1e-05,
      "loss": 0.3012,
      "step": 151300
    },
    {
      "epoch": 0.001514,
      "grad_norm": 1.5735626220703125,
      "learning_rate": 1e-05,
      "loss": 0.3124,
      "step": 151400
    },
    {
      "epoch": 0.001515,
      "grad_norm": 1.6196351051330566,
      "learning_rate": 1e-05,
      "loss": 0.3135,
      "step": 151500
    },
    {
      "epoch": 0.001516,
      "grad_norm": 1.6257820129394531,
      "learning_rate": 1e-05,
      "loss": 0.3034,
      "step": 151600
    },
    {
      "epoch": 0.001517,
      "grad_norm": 1.5410822629928589,
      "learning_rate": 1e-05,
      "loss": 0.3014,
      "step": 151700
    },
    {
      "epoch": 0.001518,
      "grad_norm": 1.7772650718688965,
      "learning_rate": 1e-05,
      "loss": 0.3023,
      "step": 151800
    },
    {
      "epoch": 0.001519,
      "grad_norm": 1.683762788772583,
      "learning_rate": 1e-05,
      "loss": 0.3003,
      "step": 151900
    },
    {
      "epoch": 0.00152,
      "grad_norm": 2.5986571311950684,
      "learning_rate": 1e-05,
      "loss": 0.3066,
      "step": 152000
    },
    {
      "epoch": 0.001521,
      "grad_norm": 1.802271842956543,
      "learning_rate": 1e-05,
      "loss": 0.3066,
      "step": 152100
    },
    {
      "epoch": 0.001522,
      "grad_norm": 1.8014322519302368,
      "learning_rate": 1e-05,
      "loss": 0.3039,
      "step": 152200
    },
    {
      "epoch": 0.001523,
      "grad_norm": 3.4917097091674805,
      "learning_rate": 1e-05,
      "loss": 0.3077,
      "step": 152300
    },
    {
      "epoch": 0.001524,
      "grad_norm": 1.3952269554138184,
      "learning_rate": 1e-05,
      "loss": 0.3063,
      "step": 152400
    },
    {
      "epoch": 0.001525,
      "grad_norm": 1.4687715768814087,
      "learning_rate": 1e-05,
      "loss": 0.3133,
      "step": 152500
    },
    {
      "epoch": 0.001526,
      "grad_norm": 1.571907877922058,
      "learning_rate": 1e-05,
      "loss": 0.3053,
      "step": 152600
    },
    {
      "epoch": 0.001527,
      "grad_norm": 1.6538187265396118,
      "learning_rate": 1e-05,
      "loss": 0.3025,
      "step": 152700
    },
    {
      "epoch": 0.001528,
      "grad_norm": 1.332327961921692,
      "learning_rate": 1e-05,
      "loss": 0.3036,
      "step": 152800
    },
    {
      "epoch": 0.001529,
      "grad_norm": 1.9874423742294312,
      "learning_rate": 1e-05,
      "loss": 0.304,
      "step": 152900
    },
    {
      "epoch": 0.00153,
      "grad_norm": 1.5692553520202637,
      "learning_rate": 1e-05,
      "loss": 0.3054,
      "step": 153000
    },
    {
      "epoch": 0.001531,
      "grad_norm": 1.6490308046340942,
      "learning_rate": 1e-05,
      "loss": 0.3094,
      "step": 153100
    },
    {
      "epoch": 0.001532,
      "grad_norm": 1.4249401092529297,
      "learning_rate": 1e-05,
      "loss": 0.3078,
      "step": 153200
    },
    {
      "epoch": 0.001533,
      "grad_norm": 1.432947039604187,
      "learning_rate": 1e-05,
      "loss": 0.3029,
      "step": 153300
    },
    {
      "epoch": 0.001534,
      "grad_norm": 1.6225773096084595,
      "learning_rate": 1e-05,
      "loss": 0.2992,
      "step": 153400
    },
    {
      "epoch": 0.001535,
      "grad_norm": 1.6100537776947021,
      "learning_rate": 1e-05,
      "loss": 0.3072,
      "step": 153500
    },
    {
      "epoch": 0.001536,
      "grad_norm": 1.6616079807281494,
      "learning_rate": 1e-05,
      "loss": 0.3066,
      "step": 153600
    },
    {
      "epoch": 0.001537,
      "grad_norm": 1.8805843591690063,
      "learning_rate": 1e-05,
      "loss": 0.3005,
      "step": 153700
    },
    {
      "epoch": 0.001538,
      "grad_norm": 1.3726553916931152,
      "learning_rate": 1e-05,
      "loss": 0.3055,
      "step": 153800
    },
    {
      "epoch": 0.001539,
      "grad_norm": 1.6736137866973877,
      "learning_rate": 1e-05,
      "loss": 0.3053,
      "step": 153900
    },
    {
      "epoch": 0.00154,
      "grad_norm": 1.704338550567627,
      "learning_rate": 1e-05,
      "loss": 0.3085,
      "step": 154000
    },
    {
      "epoch": 0.001541,
      "grad_norm": 1.5360641479492188,
      "learning_rate": 1e-05,
      "loss": 0.3051,
      "step": 154100
    },
    {
      "epoch": 0.001542,
      "grad_norm": 1.5851362943649292,
      "learning_rate": 1e-05,
      "loss": 0.3098,
      "step": 154200
    },
    {
      "epoch": 0.001543,
      "grad_norm": 1.5866998434066772,
      "learning_rate": 1e-05,
      "loss": 0.3059,
      "step": 154300
    },
    {
      "epoch": 0.001544,
      "grad_norm": 1.4236769676208496,
      "learning_rate": 1e-05,
      "loss": 0.3052,
      "step": 154400
    },
    {
      "epoch": 0.001545,
      "grad_norm": 1.5114357471466064,
      "learning_rate": 1e-05,
      "loss": 0.3013,
      "step": 154500
    },
    {
      "epoch": 0.001546,
      "grad_norm": 1.5853383541107178,
      "learning_rate": 1e-05,
      "loss": 0.3032,
      "step": 154600
    },
    {
      "epoch": 0.001547,
      "grad_norm": 1.7222450971603394,
      "learning_rate": 1e-05,
      "loss": 0.3004,
      "step": 154700
    },
    {
      "epoch": 0.001548,
      "grad_norm": 1.832231879234314,
      "learning_rate": 1e-05,
      "loss": 0.3022,
      "step": 154800
    },
    {
      "epoch": 0.001549,
      "grad_norm": 1.5499014854431152,
      "learning_rate": 1e-05,
      "loss": 0.307,
      "step": 154900
    },
    {
      "epoch": 0.00155,
      "grad_norm": 1.6969635486602783,
      "learning_rate": 1e-05,
      "loss": 0.306,
      "step": 155000
    },
    {
      "epoch": 0.001551,
      "grad_norm": 1.4039770364761353,
      "learning_rate": 1e-05,
      "loss": 0.2953,
      "step": 155100
    },
    {
      "epoch": 0.001552,
      "grad_norm": 1.4944506883621216,
      "learning_rate": 1e-05,
      "loss": 0.3068,
      "step": 155200
    },
    {
      "epoch": 0.001553,
      "grad_norm": 1.6612184047698975,
      "learning_rate": 1e-05,
      "loss": 0.3011,
      "step": 155300
    },
    {
      "epoch": 0.001554,
      "grad_norm": 1.3576972484588623,
      "learning_rate": 1e-05,
      "loss": 0.2988,
      "step": 155400
    },
    {
      "epoch": 0.001555,
      "grad_norm": 1.514603853225708,
      "learning_rate": 1e-05,
      "loss": 0.3004,
      "step": 155500
    },
    {
      "epoch": 0.001556,
      "grad_norm": 1.7504587173461914,
      "learning_rate": 1e-05,
      "loss": 0.3031,
      "step": 155600
    },
    {
      "epoch": 0.001557,
      "grad_norm": 1.723211407661438,
      "learning_rate": 1e-05,
      "loss": 0.3057,
      "step": 155700
    },
    {
      "epoch": 0.001558,
      "grad_norm": 1.395255446434021,
      "learning_rate": 1e-05,
      "loss": 0.3081,
      "step": 155800
    },
    {
      "epoch": 0.001559,
      "grad_norm": 1.8021862506866455,
      "learning_rate": 1e-05,
      "loss": 0.3039,
      "step": 155900
    },
    {
      "epoch": 0.00156,
      "grad_norm": 1.6634401082992554,
      "learning_rate": 1e-05,
      "loss": 0.3064,
      "step": 156000
    },
    {
      "epoch": 0.001561,
      "grad_norm": 1.6071007251739502,
      "learning_rate": 1e-05,
      "loss": 0.3046,
      "step": 156100
    },
    {
      "epoch": 0.001562,
      "grad_norm": 1.7638030052185059,
      "learning_rate": 1e-05,
      "loss": 0.3073,
      "step": 156200
    },
    {
      "epoch": 0.001563,
      "grad_norm": 1.6403735876083374,
      "learning_rate": 1e-05,
      "loss": 0.304,
      "step": 156300
    },
    {
      "epoch": 0.001564,
      "grad_norm": 1.8615624904632568,
      "learning_rate": 1e-05,
      "loss": 0.2961,
      "step": 156400
    },
    {
      "epoch": 0.001565,
      "grad_norm": 1.501093864440918,
      "learning_rate": 1e-05,
      "loss": 0.2993,
      "step": 156500
    },
    {
      "epoch": 0.001566,
      "grad_norm": 1.7402315139770508,
      "learning_rate": 1e-05,
      "loss": 0.3033,
      "step": 156600
    },
    {
      "epoch": 0.001567,
      "grad_norm": 1.4010441303253174,
      "learning_rate": 1e-05,
      "loss": 0.3079,
      "step": 156700
    },
    {
      "epoch": 0.001568,
      "grad_norm": 1.894376516342163,
      "learning_rate": 1e-05,
      "loss": 0.3006,
      "step": 156800
    },
    {
      "epoch": 0.001569,
      "grad_norm": 1.381251335144043,
      "learning_rate": 1e-05,
      "loss": 0.3053,
      "step": 156900
    },
    {
      "epoch": 0.00157,
      "grad_norm": 1.8080320358276367,
      "learning_rate": 1e-05,
      "loss": 0.2998,
      "step": 157000
    },
    {
      "epoch": 0.001571,
      "grad_norm": 1.4762428998947144,
      "learning_rate": 1e-05,
      "loss": 0.304,
      "step": 157100
    },
    {
      "epoch": 0.001572,
      "grad_norm": 1.4068925380706787,
      "learning_rate": 1e-05,
      "loss": 0.3069,
      "step": 157200
    },
    {
      "epoch": 0.001573,
      "grad_norm": 1.5252565145492554,
      "learning_rate": 1e-05,
      "loss": 0.2932,
      "step": 157300
    },
    {
      "epoch": 0.001574,
      "grad_norm": 1.5361262559890747,
      "learning_rate": 1e-05,
      "loss": 0.3026,
      "step": 157400
    },
    {
      "epoch": 0.001575,
      "grad_norm": 1.4667400121688843,
      "learning_rate": 1e-05,
      "loss": 0.2973,
      "step": 157500
    },
    {
      "epoch": 0.001576,
      "grad_norm": 1.3710594177246094,
      "learning_rate": 1e-05,
      "loss": 0.2958,
      "step": 157600
    },
    {
      "epoch": 0.001577,
      "grad_norm": 1.4320799112319946,
      "learning_rate": 1e-05,
      "loss": 0.3053,
      "step": 157700
    },
    {
      "epoch": 0.001578,
      "grad_norm": 1.7390029430389404,
      "learning_rate": 1e-05,
      "loss": 0.3042,
      "step": 157800
    },
    {
      "epoch": 0.001579,
      "grad_norm": 1.7083619832992554,
      "learning_rate": 1e-05,
      "loss": 0.3051,
      "step": 157900
    },
    {
      "epoch": 0.00158,
      "grad_norm": 1.5926868915557861,
      "learning_rate": 1e-05,
      "loss": 0.3092,
      "step": 158000
    },
    {
      "epoch": 0.001581,
      "grad_norm": 1.8559361696243286,
      "learning_rate": 1e-05,
      "loss": 0.3046,
      "step": 158100
    },
    {
      "epoch": 0.001582,
      "grad_norm": 1.7098585367202759,
      "learning_rate": 1e-05,
      "loss": 0.3016,
      "step": 158200
    },
    {
      "epoch": 0.001583,
      "grad_norm": 1.5086554288864136,
      "learning_rate": 1e-05,
      "loss": 0.3099,
      "step": 158300
    },
    {
      "epoch": 0.001584,
      "grad_norm": 1.5402973890304565,
      "learning_rate": 1e-05,
      "loss": 0.3047,
      "step": 158400
    },
    {
      "epoch": 0.001585,
      "grad_norm": 1.8630613088607788,
      "learning_rate": 1e-05,
      "loss": 0.2992,
      "step": 158500
    },
    {
      "epoch": 0.001586,
      "grad_norm": 1.4310539960861206,
      "learning_rate": 1e-05,
      "loss": 0.3038,
      "step": 158600
    },
    {
      "epoch": 0.001587,
      "grad_norm": 1.7185118198394775,
      "learning_rate": 1e-05,
      "loss": 0.2946,
      "step": 158700
    },
    {
      "epoch": 0.001588,
      "grad_norm": 1.6692184209823608,
      "learning_rate": 1e-05,
      "loss": 0.3042,
      "step": 158800
    },
    {
      "epoch": 0.001589,
      "grad_norm": 1.501453161239624,
      "learning_rate": 1e-05,
      "loss": 0.2964,
      "step": 158900
    },
    {
      "epoch": 0.00159,
      "grad_norm": 1.6817126274108887,
      "learning_rate": 1e-05,
      "loss": 0.3005,
      "step": 159000
    },
    {
      "epoch": 0.001591,
      "grad_norm": 1.4650346040725708,
      "learning_rate": 1e-05,
      "loss": 0.3043,
      "step": 159100
    },
    {
      "epoch": 0.001592,
      "grad_norm": 1.7585411071777344,
      "learning_rate": 1e-05,
      "loss": 0.2997,
      "step": 159200
    },
    {
      "epoch": 0.001593,
      "grad_norm": 1.440165638923645,
      "learning_rate": 1e-05,
      "loss": 0.3077,
      "step": 159300
    },
    {
      "epoch": 0.001594,
      "grad_norm": 1.6657960414886475,
      "learning_rate": 1e-05,
      "loss": 0.301,
      "step": 159400
    },
    {
      "epoch": 0.001595,
      "grad_norm": 1.4905738830566406,
      "learning_rate": 1e-05,
      "loss": 0.3025,
      "step": 159500
    },
    {
      "epoch": 0.001596,
      "grad_norm": 1.744041085243225,
      "learning_rate": 1e-05,
      "loss": 0.3015,
      "step": 159600
    },
    {
      "epoch": 0.001597,
      "grad_norm": 1.3322395086288452,
      "learning_rate": 1e-05,
      "loss": 0.2977,
      "step": 159700
    },
    {
      "epoch": 0.001598,
      "grad_norm": 1.577319622039795,
      "learning_rate": 1e-05,
      "loss": 0.2962,
      "step": 159800
    },
    {
      "epoch": 0.001599,
      "grad_norm": 1.4697024822235107,
      "learning_rate": 1e-05,
      "loss": 0.2978,
      "step": 159900
    },
    {
      "epoch": 0.0016,
      "grad_norm": 1.4797179698944092,
      "learning_rate": 1e-05,
      "loss": 0.2991,
      "step": 160000
    },
    {
      "epoch": 0.0016,
      "eval_loss": 0.2783203125,
      "eval_runtime": 115.6285,
      "eval_samples_per_second": 432.419,
      "eval_steps_per_second": 27.026,
      "step": 160000
    },
    {
      "epoch": 0.001601,
      "grad_norm": 1.489996075630188,
      "learning_rate": 1e-05,
      "loss": 0.3065,
      "step": 160100
    },
    {
      "epoch": 0.001602,
      "grad_norm": 1.6529942750930786,
      "learning_rate": 1e-05,
      "loss": 0.2961,
      "step": 160200
    },
    {
      "epoch": 0.001603,
      "grad_norm": 1.6032297611236572,
      "learning_rate": 1e-05,
      "loss": 0.3036,
      "step": 160300
    },
    {
      "epoch": 0.001604,
      "grad_norm": 1.3672584295272827,
      "learning_rate": 1e-05,
      "loss": 0.3034,
      "step": 160400
    },
    {
      "epoch": 0.001605,
      "grad_norm": 1.5010960102081299,
      "learning_rate": 1e-05,
      "loss": 0.3026,
      "step": 160500
    },
    {
      "epoch": 0.001606,
      "grad_norm": 1.631774663925171,
      "learning_rate": 1e-05,
      "loss": 0.3042,
      "step": 160600
    },
    {
      "epoch": 0.001607,
      "grad_norm": 1.3571579456329346,
      "learning_rate": 1e-05,
      "loss": 0.3008,
      "step": 160700
    },
    {
      "epoch": 0.001608,
      "grad_norm": 1.9333149194717407,
      "learning_rate": 1e-05,
      "loss": 0.2979,
      "step": 160800
    },
    {
      "epoch": 0.001609,
      "grad_norm": 1.5662444829940796,
      "learning_rate": 1e-05,
      "loss": 0.2991,
      "step": 160900
    },
    {
      "epoch": 0.00161,
      "grad_norm": 1.4831576347351074,
      "learning_rate": 1e-05,
      "loss": 0.2976,
      "step": 161000
    },
    {
      "epoch": 0.001611,
      "grad_norm": 1.3023030757904053,
      "learning_rate": 1e-05,
      "loss": 0.3031,
      "step": 161100
    },
    {
      "epoch": 0.001612,
      "grad_norm": 1.3196107149124146,
      "learning_rate": 1e-05,
      "loss": 0.2966,
      "step": 161200
    },
    {
      "epoch": 0.001613,
      "grad_norm": 1.6283353567123413,
      "learning_rate": 1e-05,
      "loss": 0.3041,
      "step": 161300
    },
    {
      "epoch": 0.001614,
      "grad_norm": 1.5330151319503784,
      "learning_rate": 1e-05,
      "loss": 0.2953,
      "step": 161400
    },
    {
      "epoch": 0.001615,
      "grad_norm": 1.583950400352478,
      "learning_rate": 1e-05,
      "loss": 0.2976,
      "step": 161500
    },
    {
      "epoch": 0.001616,
      "grad_norm": 1.557778000831604,
      "learning_rate": 1e-05,
      "loss": 0.3016,
      "step": 161600
    },
    {
      "epoch": 0.001617,
      "grad_norm": 2.2620208263397217,
      "learning_rate": 1e-05,
      "loss": 0.3063,
      "step": 161700
    },
    {
      "epoch": 0.001618,
      "grad_norm": 1.6115883588790894,
      "learning_rate": 1e-05,
      "loss": 0.2965,
      "step": 161800
    },
    {
      "epoch": 0.001619,
      "grad_norm": 1.6807005405426025,
      "learning_rate": 1e-05,
      "loss": 0.302,
      "step": 161900
    },
    {
      "epoch": 0.00162,
      "grad_norm": 1.4891862869262695,
      "learning_rate": 1e-05,
      "loss": 0.293,
      "step": 162000
    },
    {
      "epoch": 0.001621,
      "grad_norm": 1.6026562452316284,
      "learning_rate": 1e-05,
      "loss": 0.3031,
      "step": 162100
    },
    {
      "epoch": 0.001622,
      "grad_norm": 1.4442458152770996,
      "learning_rate": 1e-05,
      "loss": 0.3084,
      "step": 162200
    },
    {
      "epoch": 0.001623,
      "grad_norm": 1.5560252666473389,
      "learning_rate": 1e-05,
      "loss": 0.3017,
      "step": 162300
    },
    {
      "epoch": 0.001624,
      "grad_norm": 1.6526131629943848,
      "learning_rate": 1e-05,
      "loss": 0.2969,
      "step": 162400
    },
    {
      "epoch": 0.001625,
      "grad_norm": 1.4917162656784058,
      "learning_rate": 1e-05,
      "loss": 0.3017,
      "step": 162500
    },
    {
      "epoch": 0.001626,
      "grad_norm": 1.526892066001892,
      "learning_rate": 1e-05,
      "loss": 0.3001,
      "step": 162600
    },
    {
      "epoch": 0.001627,
      "grad_norm": 1.3089638948440552,
      "learning_rate": 1e-05,
      "loss": 0.2965,
      "step": 162700
    },
    {
      "epoch": 0.001628,
      "grad_norm": 1.6630245447158813,
      "learning_rate": 1e-05,
      "loss": 0.2936,
      "step": 162800
    },
    {
      "epoch": 0.001629,
      "grad_norm": 1.420673131942749,
      "learning_rate": 1e-05,
      "loss": 0.3051,
      "step": 162900
    },
    {
      "epoch": 0.00163,
      "grad_norm": 1.4711486101150513,
      "learning_rate": 1e-05,
      "loss": 0.2923,
      "step": 163000
    },
    {
      "epoch": 0.001631,
      "grad_norm": 1.6381266117095947,
      "learning_rate": 1e-05,
      "loss": 0.294,
      "step": 163100
    },
    {
      "epoch": 0.001632,
      "grad_norm": 1.5917518138885498,
      "learning_rate": 1e-05,
      "loss": 0.3024,
      "step": 163200
    },
    {
      "epoch": 0.001633,
      "grad_norm": 1.6768611669540405,
      "learning_rate": 1e-05,
      "loss": 0.2992,
      "step": 163300
    },
    {
      "epoch": 0.001634,
      "grad_norm": 1.5716297626495361,
      "learning_rate": 1e-05,
      "loss": 0.3028,
      "step": 163400
    },
    {
      "epoch": 0.001635,
      "grad_norm": 1.5690321922302246,
      "learning_rate": 1e-05,
      "loss": 0.2989,
      "step": 163500
    },
    {
      "epoch": 0.001636,
      "grad_norm": 1.698068618774414,
      "learning_rate": 1e-05,
      "loss": 0.3009,
      "step": 163600
    },
    {
      "epoch": 0.001637,
      "grad_norm": 1.7230242490768433,
      "learning_rate": 1e-05,
      "loss": 0.304,
      "step": 163700
    },
    {
      "epoch": 0.001638,
      "grad_norm": 1.6072338819503784,
      "learning_rate": 1e-05,
      "loss": 0.2956,
      "step": 163800
    },
    {
      "epoch": 0.001639,
      "grad_norm": 1.611342430114746,
      "learning_rate": 1e-05,
      "loss": 0.304,
      "step": 163900
    },
    {
      "epoch": 0.00164,
      "grad_norm": 1.4601253271102905,
      "learning_rate": 1e-05,
      "loss": 0.3055,
      "step": 164000
    },
    {
      "epoch": 0.001641,
      "grad_norm": 1.567654013633728,
      "learning_rate": 1e-05,
      "loss": 0.2974,
      "step": 164100
    },
    {
      "epoch": 0.001642,
      "grad_norm": 2.222820281982422,
      "learning_rate": 1e-05,
      "loss": 0.2955,
      "step": 164200
    },
    {
      "epoch": 0.001643,
      "grad_norm": 4.182979106903076,
      "learning_rate": 1e-05,
      "loss": 0.3032,
      "step": 164300
    },
    {
      "epoch": 0.001644,
      "grad_norm": 1.8960726261138916,
      "learning_rate": 1e-05,
      "loss": 0.3026,
      "step": 164400
    },
    {
      "epoch": 0.001645,
      "grad_norm": 1.5564576387405396,
      "learning_rate": 1e-05,
      "loss": 0.2951,
      "step": 164500
    },
    {
      "epoch": 0.001646,
      "grad_norm": 1.519041657447815,
      "learning_rate": 1e-05,
      "loss": 0.2961,
      "step": 164600
    },
    {
      "epoch": 0.001647,
      "grad_norm": 1.6985987424850464,
      "learning_rate": 1e-05,
      "loss": 0.2964,
      "step": 164700
    },
    {
      "epoch": 0.001648,
      "grad_norm": 1.3167078495025635,
      "learning_rate": 1e-05,
      "loss": 0.2982,
      "step": 164800
    },
    {
      "epoch": 0.001649,
      "grad_norm": 1.5005210638046265,
      "learning_rate": 1e-05,
      "loss": 0.2936,
      "step": 164900
    },
    {
      "epoch": 0.00165,
      "grad_norm": 1.4177864789962769,
      "learning_rate": 1e-05,
      "loss": 0.2978,
      "step": 165000
    },
    {
      "epoch": 0.001651,
      "grad_norm": 1.4829902648925781,
      "learning_rate": 1e-05,
      "loss": 0.2986,
      "step": 165100
    },
    {
      "epoch": 0.001652,
      "grad_norm": 1.3919358253479004,
      "learning_rate": 1e-05,
      "loss": 0.2917,
      "step": 165200
    },
    {
      "epoch": 0.001653,
      "grad_norm": 1.3996176719665527,
      "learning_rate": 1e-05,
      "loss": 0.298,
      "step": 165300
    },
    {
      "epoch": 0.001654,
      "grad_norm": 2.0305674076080322,
      "learning_rate": 1e-05,
      "loss": 0.2974,
      "step": 165400
    },
    {
      "epoch": 0.001655,
      "grad_norm": 1.710474967956543,
      "learning_rate": 1e-05,
      "loss": 0.2989,
      "step": 165500
    },
    {
      "epoch": 0.001656,
      "grad_norm": 1.4588967561721802,
      "learning_rate": 1e-05,
      "loss": 0.2953,
      "step": 165600
    },
    {
      "epoch": 0.001657,
      "grad_norm": 1.4981319904327393,
      "learning_rate": 1e-05,
      "loss": 0.2997,
      "step": 165700
    },
    {
      "epoch": 0.001658,
      "grad_norm": 1.4303194284439087,
      "learning_rate": 1e-05,
      "loss": 0.2996,
      "step": 165800
    },
    {
      "epoch": 0.001659,
      "grad_norm": 1.3741976022720337,
      "learning_rate": 1e-05,
      "loss": 0.2921,
      "step": 165900
    },
    {
      "epoch": 0.00166,
      "grad_norm": 1.6370424032211304,
      "learning_rate": 1e-05,
      "loss": 0.302,
      "step": 166000
    },
    {
      "epoch": 0.001661,
      "grad_norm": 1.6333328485488892,
      "learning_rate": 1e-05,
      "loss": 0.2957,
      "step": 166100
    },
    {
      "epoch": 0.001662,
      "grad_norm": 1.5434244871139526,
      "learning_rate": 1e-05,
      "loss": 0.2877,
      "step": 166200
    },
    {
      "epoch": 0.001663,
      "grad_norm": 1.4523191452026367,
      "learning_rate": 1e-05,
      "loss": 0.2984,
      "step": 166300
    },
    {
      "epoch": 0.001664,
      "grad_norm": 1.4161934852600098,
      "learning_rate": 1e-05,
      "loss": 0.309,
      "step": 166400
    },
    {
      "epoch": 0.001665,
      "grad_norm": 1.5231043100357056,
      "learning_rate": 1e-05,
      "loss": 0.2942,
      "step": 166500
    },
    {
      "epoch": 0.001666,
      "grad_norm": 1.4963332414627075,
      "learning_rate": 1e-05,
      "loss": 0.2931,
      "step": 166600
    },
    {
      "epoch": 0.001667,
      "grad_norm": 1.5862305164337158,
      "learning_rate": 1e-05,
      "loss": 0.2876,
      "step": 166700
    },
    {
      "epoch": 0.001668,
      "grad_norm": 1.4414396286010742,
      "learning_rate": 1e-05,
      "loss": 0.2971,
      "step": 166800
    },
    {
      "epoch": 0.001669,
      "grad_norm": 1.5475411415100098,
      "learning_rate": 1e-05,
      "loss": 0.3065,
      "step": 166900
    },
    {
      "epoch": 0.00167,
      "grad_norm": 1.6189428567886353,
      "learning_rate": 1e-05,
      "loss": 0.2929,
      "step": 167000
    },
    {
      "epoch": 0.001671,
      "grad_norm": 3.0786776542663574,
      "learning_rate": 1e-05,
      "loss": 0.2959,
      "step": 167100
    },
    {
      "epoch": 0.001672,
      "grad_norm": 1.6727555990219116,
      "learning_rate": 1e-05,
      "loss": 0.302,
      "step": 167200
    },
    {
      "epoch": 0.001673,
      "grad_norm": 1.6792882680892944,
      "learning_rate": 1e-05,
      "loss": 0.2956,
      "step": 167300
    },
    {
      "epoch": 0.001674,
      "grad_norm": 1.4409688711166382,
      "learning_rate": 1e-05,
      "loss": 0.2931,
      "step": 167400
    },
    {
      "epoch": 0.001675,
      "grad_norm": 1.84697687625885,
      "learning_rate": 1e-05,
      "loss": 0.2942,
      "step": 167500
    },
    {
      "epoch": 0.001676,
      "grad_norm": 1.294731855392456,
      "learning_rate": 1e-05,
      "loss": 0.2908,
      "step": 167600
    },
    {
      "epoch": 0.001677,
      "grad_norm": 1.557396411895752,
      "learning_rate": 1e-05,
      "loss": 0.2906,
      "step": 167700
    },
    {
      "epoch": 0.001678,
      "grad_norm": 1.6044083833694458,
      "learning_rate": 1e-05,
      "loss": 0.2953,
      "step": 167800
    },
    {
      "epoch": 0.001679,
      "grad_norm": 1.5445910692214966,
      "learning_rate": 1e-05,
      "loss": 0.2985,
      "step": 167900
    },
    {
      "epoch": 0.00168,
      "grad_norm": 1.5258064270019531,
      "learning_rate": 1e-05,
      "loss": 0.292,
      "step": 168000
    },
    {
      "epoch": 0.001681,
      "grad_norm": 1.6000378131866455,
      "learning_rate": 1e-05,
      "loss": 0.2961,
      "step": 168100
    },
    {
      "epoch": 0.001682,
      "grad_norm": 1.4779728651046753,
      "learning_rate": 1e-05,
      "loss": 0.3035,
      "step": 168200
    },
    {
      "epoch": 0.001683,
      "grad_norm": 1.8927539587020874,
      "learning_rate": 1e-05,
      "loss": 0.2961,
      "step": 168300
    },
    {
      "epoch": 0.001684,
      "grad_norm": 1.5276615619659424,
      "learning_rate": 1e-05,
      "loss": 0.2941,
      "step": 168400
    },
    {
      "epoch": 0.001685,
      "grad_norm": 1.4742182493209839,
      "learning_rate": 1e-05,
      "loss": 0.2977,
      "step": 168500
    },
    {
      "epoch": 0.001686,
      "grad_norm": 1.3637776374816895,
      "learning_rate": 1e-05,
      "loss": 0.2919,
      "step": 168600
    },
    {
      "epoch": 0.001687,
      "grad_norm": 2.3156306743621826,
      "learning_rate": 1e-05,
      "loss": 0.3035,
      "step": 168700
    },
    {
      "epoch": 0.001688,
      "grad_norm": 1.5192285776138306,
      "learning_rate": 1e-05,
      "loss": 0.2936,
      "step": 168800
    },
    {
      "epoch": 0.001689,
      "grad_norm": 1.6717158555984497,
      "learning_rate": 1e-05,
      "loss": 0.2873,
      "step": 168900
    },
    {
      "epoch": 0.00169,
      "grad_norm": 1.9118905067443848,
      "learning_rate": 1e-05,
      "loss": 0.287,
      "step": 169000
    },
    {
      "epoch": 0.001691,
      "grad_norm": 1.750054955482483,
      "learning_rate": 1e-05,
      "loss": 0.2943,
      "step": 169100
    },
    {
      "epoch": 0.001692,
      "grad_norm": 1.545607566833496,
      "learning_rate": 1e-05,
      "loss": 0.2942,
      "step": 169200
    },
    {
      "epoch": 0.001693,
      "grad_norm": 1.576042652130127,
      "learning_rate": 1e-05,
      "loss": 0.2913,
      "step": 169300
    },
    {
      "epoch": 0.001694,
      "grad_norm": 1.5161222219467163,
      "learning_rate": 1e-05,
      "loss": 0.2922,
      "step": 169400
    },
    {
      "epoch": 0.001695,
      "grad_norm": 1.5191494226455688,
      "learning_rate": 1e-05,
      "loss": 0.3026,
      "step": 169500
    },
    {
      "epoch": 0.001696,
      "grad_norm": 1.807310700416565,
      "learning_rate": 1e-05,
      "loss": 0.2954,
      "step": 169600
    },
    {
      "epoch": 0.001697,
      "grad_norm": 1.5956357717514038,
      "learning_rate": 1e-05,
      "loss": 0.3014,
      "step": 169700
    },
    {
      "epoch": 0.001698,
      "grad_norm": 2.556617259979248,
      "learning_rate": 1e-05,
      "loss": 0.2993,
      "step": 169800
    },
    {
      "epoch": 0.001699,
      "grad_norm": 1.5786460638046265,
      "learning_rate": 1e-05,
      "loss": 0.2942,
      "step": 169900
    },
    {
      "epoch": 0.0017,
      "grad_norm": 1.6583482027053833,
      "learning_rate": 1e-05,
      "loss": 0.3006,
      "step": 170000
    },
    {
      "epoch": 0.001701,
      "grad_norm": 1.8018178939819336,
      "learning_rate": 1e-05,
      "loss": 0.296,
      "step": 170100
    },
    {
      "epoch": 0.001702,
      "grad_norm": 1.3693221807479858,
      "learning_rate": 1e-05,
      "loss": 0.2949,
      "step": 170200
    },
    {
      "epoch": 0.001703,
      "grad_norm": 1.5675960779190063,
      "learning_rate": 1e-05,
      "loss": 0.2949,
      "step": 170300
    },
    {
      "epoch": 0.001704,
      "grad_norm": 1.8481919765472412,
      "learning_rate": 1e-05,
      "loss": 0.3014,
      "step": 170400
    },
    {
      "epoch": 0.001705,
      "grad_norm": 1.6740190982818604,
      "learning_rate": 1e-05,
      "loss": 0.2991,
      "step": 170500
    },
    {
      "epoch": 0.001706,
      "grad_norm": 1.5895060300827026,
      "learning_rate": 1e-05,
      "loss": 0.2957,
      "step": 170600
    },
    {
      "epoch": 0.001707,
      "grad_norm": 1.3283610343933105,
      "learning_rate": 1e-05,
      "loss": 0.2902,
      "step": 170700
    },
    {
      "epoch": 0.001708,
      "grad_norm": 1.542960524559021,
      "learning_rate": 1e-05,
      "loss": 0.2922,
      "step": 170800
    },
    {
      "epoch": 0.001709,
      "grad_norm": 1.5971072912216187,
      "learning_rate": 1e-05,
      "loss": 0.2965,
      "step": 170900
    },
    {
      "epoch": 0.00171,
      "grad_norm": 1.559484601020813,
      "learning_rate": 1e-05,
      "loss": 0.2917,
      "step": 171000
    },
    {
      "epoch": 0.001711,
      "grad_norm": 1.4500508308410645,
      "learning_rate": 1e-05,
      "loss": 0.2948,
      "step": 171100
    },
    {
      "epoch": 0.001712,
      "grad_norm": 1.7252469062805176,
      "learning_rate": 1e-05,
      "loss": 0.2978,
      "step": 171200
    },
    {
      "epoch": 0.001713,
      "grad_norm": 1.3989806175231934,
      "learning_rate": 1e-05,
      "loss": 0.2918,
      "step": 171300
    },
    {
      "epoch": 0.001714,
      "grad_norm": 1.3513588905334473,
      "learning_rate": 1e-05,
      "loss": 0.2945,
      "step": 171400
    },
    {
      "epoch": 0.001715,
      "grad_norm": 1.7322951555252075,
      "learning_rate": 1e-05,
      "loss": 0.2937,
      "step": 171500
    },
    {
      "epoch": 0.001716,
      "grad_norm": 1.5518382787704468,
      "learning_rate": 1e-05,
      "loss": 0.2963,
      "step": 171600
    },
    {
      "epoch": 0.001717,
      "grad_norm": 1.6225837469100952,
      "learning_rate": 1e-05,
      "loss": 0.2996,
      "step": 171700
    },
    {
      "epoch": 0.001718,
      "grad_norm": 1.6591675281524658,
      "learning_rate": 1e-05,
      "loss": 0.3009,
      "step": 171800
    },
    {
      "epoch": 0.001719,
      "grad_norm": 1.6477521657943726,
      "learning_rate": 1e-05,
      "loss": 0.2986,
      "step": 171900
    },
    {
      "epoch": 0.00172,
      "grad_norm": 1.392760992050171,
      "learning_rate": 1e-05,
      "loss": 0.2988,
      "step": 172000
    },
    {
      "epoch": 0.001721,
      "grad_norm": 2.2496235370635986,
      "learning_rate": 1e-05,
      "loss": 0.2929,
      "step": 172100
    },
    {
      "epoch": 0.001722,
      "grad_norm": 1.5061190128326416,
      "learning_rate": 1e-05,
      "loss": 0.2886,
      "step": 172200
    },
    {
      "epoch": 0.001723,
      "grad_norm": 1.567452311515808,
      "learning_rate": 1e-05,
      "loss": 0.2995,
      "step": 172300
    },
    {
      "epoch": 0.001724,
      "grad_norm": 1.5281249284744263,
      "learning_rate": 1e-05,
      "loss": 0.3001,
      "step": 172400
    },
    {
      "epoch": 0.001725,
      "grad_norm": 1.5612703561782837,
      "learning_rate": 1e-05,
      "loss": 0.3001,
      "step": 172500
    },
    {
      "epoch": 0.001726,
      "grad_norm": 1.5059471130371094,
      "learning_rate": 1e-05,
      "loss": 0.2931,
      "step": 172600
    },
    {
      "epoch": 0.001727,
      "grad_norm": 1.598825216293335,
      "learning_rate": 1e-05,
      "loss": 0.2935,
      "step": 172700
    },
    {
      "epoch": 0.001728,
      "grad_norm": 1.605755090713501,
      "learning_rate": 1e-05,
      "loss": 0.2971,
      "step": 172800
    },
    {
      "epoch": 0.001729,
      "grad_norm": 1.6247023344039917,
      "learning_rate": 1e-05,
      "loss": 0.2902,
      "step": 172900
    },
    {
      "epoch": 0.00173,
      "grad_norm": 1.3344452381134033,
      "learning_rate": 1e-05,
      "loss": 0.3012,
      "step": 173000
    },
    {
      "epoch": 0.001731,
      "grad_norm": 1.5707899332046509,
      "learning_rate": 1e-05,
      "loss": 0.2924,
      "step": 173100
    },
    {
      "epoch": 0.001732,
      "grad_norm": 1.3291939496994019,
      "learning_rate": 1e-05,
      "loss": 0.2883,
      "step": 173200
    },
    {
      "epoch": 0.001733,
      "grad_norm": 1.429669737815857,
      "learning_rate": 1e-05,
      "loss": 0.2973,
      "step": 173300
    },
    {
      "epoch": 0.001734,
      "grad_norm": 1.5371525287628174,
      "learning_rate": 1e-05,
      "loss": 0.2962,
      "step": 173400
    },
    {
      "epoch": 0.001735,
      "grad_norm": 1.5190463066101074,
      "learning_rate": 1e-05,
      "loss": 0.3017,
      "step": 173500
    },
    {
      "epoch": 0.001736,
      "grad_norm": 1.4913296699523926,
      "learning_rate": 1e-05,
      "loss": 0.3002,
      "step": 173600
    },
    {
      "epoch": 0.001737,
      "grad_norm": 1.5345001220703125,
      "learning_rate": 1e-05,
      "loss": 0.2946,
      "step": 173700
    },
    {
      "epoch": 0.001738,
      "grad_norm": 1.5316048860549927,
      "learning_rate": 1e-05,
      "loss": 0.2978,
      "step": 173800
    },
    {
      "epoch": 0.001739,
      "grad_norm": 1.597626805305481,
      "learning_rate": 1e-05,
      "loss": 0.289,
      "step": 173900
    },
    {
      "epoch": 0.00174,
      "grad_norm": 1.2749770879745483,
      "learning_rate": 1e-05,
      "loss": 0.2933,
      "step": 174000
    },
    {
      "epoch": 0.001741,
      "grad_norm": 1.5808467864990234,
      "learning_rate": 1e-05,
      "loss": 0.2997,
      "step": 174100
    },
    {
      "epoch": 0.001742,
      "grad_norm": 1.2959426641464233,
      "learning_rate": 1e-05,
      "loss": 0.2873,
      "step": 174200
    },
    {
      "epoch": 0.001743,
      "grad_norm": 1.6337339878082275,
      "learning_rate": 1e-05,
      "loss": 0.2862,
      "step": 174300
    },
    {
      "epoch": 0.001744,
      "grad_norm": 1.528238296508789,
      "learning_rate": 1e-05,
      "loss": 0.2847,
      "step": 174400
    },
    {
      "epoch": 0.001745,
      "grad_norm": 1.4361398220062256,
      "learning_rate": 1e-05,
      "loss": 0.2953,
      "step": 174500
    },
    {
      "epoch": 0.001746,
      "grad_norm": 1.6236249208450317,
      "learning_rate": 1e-05,
      "loss": 0.2961,
      "step": 174600
    },
    {
      "epoch": 0.001747,
      "grad_norm": 1.4904263019561768,
      "learning_rate": 1e-05,
      "loss": 0.287,
      "step": 174700
    },
    {
      "epoch": 0.001748,
      "grad_norm": 1.9077177047729492,
      "learning_rate": 1e-05,
      "loss": 0.292,
      "step": 174800
    },
    {
      "epoch": 0.001749,
      "grad_norm": 1.7309199571609497,
      "learning_rate": 1e-05,
      "loss": 0.2952,
      "step": 174900
    },
    {
      "epoch": 0.00175,
      "grad_norm": 1.6516369581222534,
      "learning_rate": 1e-05,
      "loss": 0.2935,
      "step": 175000
    },
    {
      "epoch": 0.001751,
      "grad_norm": 1.6567866802215576,
      "learning_rate": 1e-05,
      "loss": 0.2913,
      "step": 175100
    },
    {
      "epoch": 0.001752,
      "grad_norm": 1.4897688627243042,
      "learning_rate": 1e-05,
      "loss": 0.2883,
      "step": 175200
    },
    {
      "epoch": 0.001753,
      "grad_norm": 1.3633755445480347,
      "learning_rate": 1e-05,
      "loss": 0.2985,
      "step": 175300
    },
    {
      "epoch": 0.001754,
      "grad_norm": 1.6736537218093872,
      "learning_rate": 1e-05,
      "loss": 0.2924,
      "step": 175400
    },
    {
      "epoch": 0.001755,
      "grad_norm": 1.5742863416671753,
      "learning_rate": 1e-05,
      "loss": 0.285,
      "step": 175500
    },
    {
      "epoch": 0.001756,
      "grad_norm": 1.5823429822921753,
      "learning_rate": 1e-05,
      "loss": 0.2865,
      "step": 175600
    },
    {
      "epoch": 0.001757,
      "grad_norm": 1.4199198484420776,
      "learning_rate": 1e-05,
      "loss": 0.2969,
      "step": 175700
    },
    {
      "epoch": 0.001758,
      "grad_norm": 1.4227279424667358,
      "learning_rate": 1e-05,
      "loss": 0.2889,
      "step": 175800
    },
    {
      "epoch": 0.001759,
      "grad_norm": 1.3790444135665894,
      "learning_rate": 1e-05,
      "loss": 0.2947,
      "step": 175900
    },
    {
      "epoch": 0.00176,
      "grad_norm": 1.717409372329712,
      "learning_rate": 1e-05,
      "loss": 0.2934,
      "step": 176000
    },
    {
      "epoch": 0.001761,
      "grad_norm": 1.5897624492645264,
      "learning_rate": 1e-05,
      "loss": 0.3003,
      "step": 176100
    },
    {
      "epoch": 0.001762,
      "grad_norm": 1.5490648746490479,
      "learning_rate": 1e-05,
      "loss": 0.294,
      "step": 176200
    },
    {
      "epoch": 0.001763,
      "grad_norm": 1.5469624996185303,
      "learning_rate": 1e-05,
      "loss": 0.2915,
      "step": 176300
    },
    {
      "epoch": 0.001764,
      "grad_norm": 1.6679738759994507,
      "learning_rate": 1e-05,
      "loss": 0.2966,
      "step": 176400
    },
    {
      "epoch": 0.001765,
      "grad_norm": 1.3103872537612915,
      "learning_rate": 1e-05,
      "loss": 0.2943,
      "step": 176500
    },
    {
      "epoch": 0.001766,
      "grad_norm": 1.6076604127883911,
      "learning_rate": 1e-05,
      "loss": 0.293,
      "step": 176600
    },
    {
      "epoch": 0.001767,
      "grad_norm": 1.7157478332519531,
      "learning_rate": 1e-05,
      "loss": 0.2908,
      "step": 176700
    },
    {
      "epoch": 0.001768,
      "grad_norm": 1.6932576894760132,
      "learning_rate": 1e-05,
      "loss": 0.2931,
      "step": 176800
    },
    {
      "epoch": 0.001769,
      "grad_norm": 1.3828284740447998,
      "learning_rate": 1e-05,
      "loss": 0.2886,
      "step": 176900
    },
    {
      "epoch": 0.00177,
      "grad_norm": 1.5653789043426514,
      "learning_rate": 1e-05,
      "loss": 0.2952,
      "step": 177000
    },
    {
      "epoch": 0.001771,
      "grad_norm": 1.7369993925094604,
      "learning_rate": 1e-05,
      "loss": 0.287,
      "step": 177100
    },
    {
      "epoch": 0.001772,
      "grad_norm": 1.541108250617981,
      "learning_rate": 1e-05,
      "loss": 0.2885,
      "step": 177200
    },
    {
      "epoch": 0.001773,
      "grad_norm": 1.5713878870010376,
      "learning_rate": 1e-05,
      "loss": 0.2951,
      "step": 177300
    },
    {
      "epoch": 0.001774,
      "grad_norm": 1.6048654317855835,
      "learning_rate": 1e-05,
      "loss": 0.2922,
      "step": 177400
    },
    {
      "epoch": 0.001775,
      "grad_norm": 1.6117722988128662,
      "learning_rate": 1e-05,
      "loss": 0.2909,
      "step": 177500
    },
    {
      "epoch": 0.001776,
      "grad_norm": 1.759687900543213,
      "learning_rate": 1e-05,
      "loss": 0.2967,
      "step": 177600
    },
    {
      "epoch": 0.001777,
      "grad_norm": 1.4271762371063232,
      "learning_rate": 1e-05,
      "loss": 0.2868,
      "step": 177700
    },
    {
      "epoch": 0.001778,
      "grad_norm": 1.4909316301345825,
      "learning_rate": 1e-05,
      "loss": 0.2918,
      "step": 177800
    },
    {
      "epoch": 0.001779,
      "grad_norm": 1.498526692390442,
      "learning_rate": 1e-05,
      "loss": 0.2973,
      "step": 177900
    },
    {
      "epoch": 0.00178,
      "grad_norm": 1.373579502105713,
      "learning_rate": 1e-05,
      "loss": 0.2857,
      "step": 178000
    },
    {
      "epoch": 0.001781,
      "grad_norm": 1.5439717769622803,
      "learning_rate": 1e-05,
      "loss": 0.2907,
      "step": 178100
    },
    {
      "epoch": 0.001782,
      "grad_norm": 1.6108523607254028,
      "learning_rate": 1e-05,
      "loss": 0.294,
      "step": 178200
    },
    {
      "epoch": 0.001783,
      "grad_norm": 1.3739374876022339,
      "learning_rate": 1e-05,
      "loss": 0.2959,
      "step": 178300
    },
    {
      "epoch": 0.001784,
      "grad_norm": 2.883863925933838,
      "learning_rate": 1e-05,
      "loss": 0.2891,
      "step": 178400
    },
    {
      "epoch": 0.001785,
      "grad_norm": 1.3920929431915283,
      "learning_rate": 1e-05,
      "loss": 0.291,
      "step": 178500
    },
    {
      "epoch": 0.001786,
      "grad_norm": 1.4327913522720337,
      "learning_rate": 1e-05,
      "loss": 0.2892,
      "step": 178600
    },
    {
      "epoch": 0.001787,
      "grad_norm": 1.6962852478027344,
      "learning_rate": 1e-05,
      "loss": 0.2852,
      "step": 178700
    },
    {
      "epoch": 0.001788,
      "grad_norm": 1.5259137153625488,
      "learning_rate": 1e-05,
      "loss": 0.2974,
      "step": 178800
    },
    {
      "epoch": 0.001789,
      "grad_norm": 1.4931055307388306,
      "learning_rate": 1e-05,
      "loss": 0.2888,
      "step": 178900
    },
    {
      "epoch": 0.00179,
      "grad_norm": 1.5622413158416748,
      "learning_rate": 1e-05,
      "loss": 0.2997,
      "step": 179000
    },
    {
      "epoch": 0.001791,
      "grad_norm": 1.5523242950439453,
      "learning_rate": 1e-05,
      "loss": 0.2929,
      "step": 179100
    },
    {
      "epoch": 0.001792,
      "grad_norm": 1.4353841543197632,
      "learning_rate": 1e-05,
      "loss": 0.285,
      "step": 179200
    },
    {
      "epoch": 0.001793,
      "grad_norm": 1.4831793308258057,
      "learning_rate": 1e-05,
      "loss": 0.2938,
      "step": 179300
    },
    {
      "epoch": 0.001794,
      "grad_norm": 1.483508825302124,
      "learning_rate": 1e-05,
      "loss": 0.2919,
      "step": 179400
    },
    {
      "epoch": 0.001795,
      "grad_norm": 1.4768630266189575,
      "learning_rate": 1e-05,
      "loss": 0.2892,
      "step": 179500
    },
    {
      "epoch": 0.001796,
      "grad_norm": 1.329671859741211,
      "learning_rate": 1e-05,
      "loss": 0.2902,
      "step": 179600
    },
    {
      "epoch": 0.001797,
      "grad_norm": 1.4865089654922485,
      "learning_rate": 1e-05,
      "loss": 0.2906,
      "step": 179700
    },
    {
      "epoch": 0.001798,
      "grad_norm": 1.6199544668197632,
      "learning_rate": 1e-05,
      "loss": 0.3013,
      "step": 179800
    },
    {
      "epoch": 0.001799,
      "grad_norm": 1.633739948272705,
      "learning_rate": 1e-05,
      "loss": 0.29,
      "step": 179900
    },
    {
      "epoch": 0.0018,
      "grad_norm": 1.3728777170181274,
      "learning_rate": 1e-05,
      "loss": 0.2875,
      "step": 180000
    },
    {
      "epoch": 0.0018,
      "eval_loss": 0.264892578125,
      "eval_runtime": 113.9893,
      "eval_samples_per_second": 438.638,
      "eval_steps_per_second": 27.415,
      "step": 180000
    },
    {
      "epoch": 0.001801,
      "grad_norm": 2.310349702835083,
      "learning_rate": 1e-05,
      "loss": 0.2925,
      "step": 180100
    },
    {
      "epoch": 0.001802,
      "grad_norm": 1.3453627824783325,
      "learning_rate": 1e-05,
      "loss": 0.292,
      "step": 180200
    },
    {
      "epoch": 0.001803,
      "grad_norm": 1.8540631532669067,
      "learning_rate": 1e-05,
      "loss": 0.2946,
      "step": 180300
    },
    {
      "epoch": 0.001804,
      "grad_norm": 1.594420075416565,
      "learning_rate": 1e-05,
      "loss": 0.2923,
      "step": 180400
    },
    {
      "epoch": 0.001805,
      "grad_norm": 1.5511283874511719,
      "learning_rate": 1e-05,
      "loss": 0.2924,
      "step": 180500
    },
    {
      "epoch": 0.001806,
      "grad_norm": 1.8114066123962402,
      "learning_rate": 1e-05,
      "loss": 0.2868,
      "step": 180600
    },
    {
      "epoch": 0.001807,
      "grad_norm": 1.5278881788253784,
      "learning_rate": 1e-05,
      "loss": 0.2896,
      "step": 180700
    },
    {
      "epoch": 0.001808,
      "grad_norm": 1.4767954349517822,
      "learning_rate": 1e-05,
      "loss": 0.2928,
      "step": 180800
    },
    {
      "epoch": 0.001809,
      "grad_norm": 1.3067213296890259,
      "learning_rate": 1e-05,
      "loss": 0.2918,
      "step": 180900
    },
    {
      "epoch": 0.00181,
      "grad_norm": 1.7097564935684204,
      "learning_rate": 1e-05,
      "loss": 0.2913,
      "step": 181000
    },
    {
      "epoch": 0.001811,
      "grad_norm": 1.6690146923065186,
      "learning_rate": 1e-05,
      "loss": 0.2808,
      "step": 181100
    },
    {
      "epoch": 0.001812,
      "grad_norm": 1.6829502582550049,
      "learning_rate": 1e-05,
      "loss": 0.2874,
      "step": 181200
    },
    {
      "epoch": 0.001813,
      "grad_norm": 1.2836750745773315,
      "learning_rate": 1e-05,
      "loss": 0.2914,
      "step": 181300
    },
    {
      "epoch": 0.001814,
      "grad_norm": 1.5141675472259521,
      "learning_rate": 1e-05,
      "loss": 0.2869,
      "step": 181400
    },
    {
      "epoch": 0.001815,
      "grad_norm": 1.571880578994751,
      "learning_rate": 1e-05,
      "loss": 0.2908,
      "step": 181500
    },
    {
      "epoch": 0.001816,
      "grad_norm": 1.5643311738967896,
      "learning_rate": 1e-05,
      "loss": 0.2905,
      "step": 181600
    },
    {
      "epoch": 0.001817,
      "grad_norm": 1.9679372310638428,
      "learning_rate": 1e-05,
      "loss": 0.2866,
      "step": 181700
    },
    {
      "epoch": 0.001818,
      "grad_norm": 1.5207774639129639,
      "learning_rate": 1e-05,
      "loss": 0.2876,
      "step": 181800
    },
    {
      "epoch": 0.001819,
      "grad_norm": 1.4971661567687988,
      "learning_rate": 1e-05,
      "loss": 0.2837,
      "step": 181900
    },
    {
      "epoch": 0.00182,
      "grad_norm": 1.3630481958389282,
      "learning_rate": 1e-05,
      "loss": 0.2901,
      "step": 182000
    },
    {
      "epoch": 0.001821,
      "grad_norm": 1.7479013204574585,
      "learning_rate": 1e-05,
      "loss": 0.2809,
      "step": 182100
    },
    {
      "epoch": 0.001822,
      "grad_norm": 1.6308436393737793,
      "learning_rate": 1e-05,
      "loss": 0.2869,
      "step": 182200
    },
    {
      "epoch": 0.001823,
      "grad_norm": 1.6583669185638428,
      "learning_rate": 1e-05,
      "loss": 0.2828,
      "step": 182300
    },
    {
      "epoch": 0.001824,
      "grad_norm": 1.7341161966323853,
      "learning_rate": 1e-05,
      "loss": 0.296,
      "step": 182400
    },
    {
      "epoch": 0.001825,
      "grad_norm": 1.2434451580047607,
      "learning_rate": 1e-05,
      "loss": 0.2903,
      "step": 182500
    },
    {
      "epoch": 0.001826,
      "grad_norm": 1.4031060934066772,
      "learning_rate": 1e-05,
      "loss": 0.2871,
      "step": 182600
    },
    {
      "epoch": 0.001827,
      "grad_norm": 1.417802095413208,
      "learning_rate": 1e-05,
      "loss": 0.2849,
      "step": 182700
    },
    {
      "epoch": 0.001828,
      "grad_norm": 1.6376116275787354,
      "learning_rate": 1e-05,
      "loss": 0.2888,
      "step": 182800
    },
    {
      "epoch": 0.001829,
      "grad_norm": 1.5004040002822876,
      "learning_rate": 1e-05,
      "loss": 0.2889,
      "step": 182900
    },
    {
      "epoch": 0.00183,
      "grad_norm": 1.3705480098724365,
      "learning_rate": 1e-05,
      "loss": 0.2916,
      "step": 183000
    },
    {
      "epoch": 0.001831,
      "grad_norm": 1.4046076536178589,
      "learning_rate": 1e-05,
      "loss": 0.2871,
      "step": 183100
    },
    {
      "epoch": 0.001832,
      "grad_norm": 1.460054874420166,
      "learning_rate": 1e-05,
      "loss": 0.2915,
      "step": 183200
    },
    {
      "epoch": 0.001833,
      "grad_norm": 2.7054927349090576,
      "learning_rate": 1e-05,
      "loss": 0.2913,
      "step": 183300
    },
    {
      "epoch": 0.001834,
      "grad_norm": 1.5564157962799072,
      "learning_rate": 1e-05,
      "loss": 0.2803,
      "step": 183400
    },
    {
      "epoch": 0.001835,
      "grad_norm": 1.4496382474899292,
      "learning_rate": 1e-05,
      "loss": 0.2936,
      "step": 183500
    },
    {
      "epoch": 0.001836,
      "grad_norm": 1.3869458436965942,
      "learning_rate": 1e-05,
      "loss": 0.2927,
      "step": 183600
    },
    {
      "epoch": 0.001837,
      "grad_norm": 1.5351581573486328,
      "learning_rate": 1e-05,
      "loss": 0.294,
      "step": 183700
    },
    {
      "epoch": 0.001838,
      "grad_norm": 1.3545173406600952,
      "learning_rate": 1e-05,
      "loss": 0.289,
      "step": 183800
    },
    {
      "epoch": 0.001839,
      "grad_norm": 2.300602674484253,
      "learning_rate": 1e-05,
      "loss": 0.2814,
      "step": 183900
    },
    {
      "epoch": 0.00184,
      "grad_norm": 1.4842824935913086,
      "learning_rate": 1e-05,
      "loss": 0.2895,
      "step": 184000
    },
    {
      "epoch": 0.001841,
      "grad_norm": 1.6287872791290283,
      "learning_rate": 1e-05,
      "loss": 0.2941,
      "step": 184100
    },
    {
      "epoch": 0.001842,
      "grad_norm": 1.3800750970840454,
      "learning_rate": 1e-05,
      "loss": 0.2834,
      "step": 184200
    },
    {
      "epoch": 0.001843,
      "grad_norm": 1.4979010820388794,
      "learning_rate": 1e-05,
      "loss": 0.2903,
      "step": 184300
    },
    {
      "epoch": 0.001844,
      "grad_norm": 1.5063962936401367,
      "learning_rate": 1e-05,
      "loss": 0.2857,
      "step": 184400
    },
    {
      "epoch": 0.001845,
      "grad_norm": 1.2965285778045654,
      "learning_rate": 1e-05,
      "loss": 0.2856,
      "step": 184500
    },
    {
      "epoch": 0.001846,
      "grad_norm": 1.467738389968872,
      "learning_rate": 1e-05,
      "loss": 0.2826,
      "step": 184600
    },
    {
      "epoch": 0.001847,
      "grad_norm": 1.4696054458618164,
      "learning_rate": 1e-05,
      "loss": 0.2938,
      "step": 184700
    },
    {
      "epoch": 0.001848,
      "grad_norm": 1.6094032526016235,
      "learning_rate": 1e-05,
      "loss": 0.2852,
      "step": 184800
    },
    {
      "epoch": 0.001849,
      "grad_norm": 1.5092664957046509,
      "learning_rate": 1e-05,
      "loss": 0.2951,
      "step": 184900
    },
    {
      "epoch": 0.00185,
      "grad_norm": 1.482487678527832,
      "learning_rate": 1e-05,
      "loss": 0.2905,
      "step": 185000
    },
    {
      "epoch": 0.001851,
      "grad_norm": 1.4276769161224365,
      "learning_rate": 1e-05,
      "loss": 0.2852,
      "step": 185100
    },
    {
      "epoch": 0.001852,
      "grad_norm": 1.756006121635437,
      "learning_rate": 1e-05,
      "loss": 0.2904,
      "step": 185200
    },
    {
      "epoch": 0.001853,
      "grad_norm": 1.7595645189285278,
      "learning_rate": 1e-05,
      "loss": 0.2808,
      "step": 185300
    },
    {
      "epoch": 0.001854,
      "grad_norm": 1.4511970281600952,
      "learning_rate": 1e-05,
      "loss": 0.2854,
      "step": 185400
    },
    {
      "epoch": 0.001855,
      "grad_norm": 1.4957120418548584,
      "learning_rate": 1e-05,
      "loss": 0.287,
      "step": 185500
    },
    {
      "epoch": 0.001856,
      "grad_norm": 1.5508650541305542,
      "learning_rate": 1e-05,
      "loss": 0.2868,
      "step": 185600
    },
    {
      "epoch": 0.001857,
      "grad_norm": 1.4669588804244995,
      "learning_rate": 1e-05,
      "loss": 0.2955,
      "step": 185700
    },
    {
      "epoch": 0.001858,
      "grad_norm": 1.456214189529419,
      "learning_rate": 1e-05,
      "loss": 0.2887,
      "step": 185800
    },
    {
      "epoch": 0.001859,
      "grad_norm": 1.5572723150253296,
      "learning_rate": 1e-05,
      "loss": 0.2878,
      "step": 185900
    },
    {
      "epoch": 0.00186,
      "grad_norm": 1.1924431324005127,
      "learning_rate": 1e-05,
      "loss": 0.2872,
      "step": 186000
    },
    {
      "epoch": 0.001861,
      "grad_norm": 1.5211331844329834,
      "learning_rate": 1e-05,
      "loss": 0.2831,
      "step": 186100
    },
    {
      "epoch": 0.001862,
      "grad_norm": 3.843024969100952,
      "learning_rate": 1e-05,
      "loss": 0.2831,
      "step": 186200
    },
    {
      "epoch": 0.001863,
      "grad_norm": 1.5179247856140137,
      "learning_rate": 1e-05,
      "loss": 0.285,
      "step": 186300
    },
    {
      "epoch": 0.001864,
      "grad_norm": 1.3682702779769897,
      "learning_rate": 1e-05,
      "loss": 0.2882,
      "step": 186400
    },
    {
      "epoch": 0.001865,
      "grad_norm": 1.5251023769378662,
      "learning_rate": 1e-05,
      "loss": 0.2915,
      "step": 186500
    },
    {
      "epoch": 0.001866,
      "grad_norm": 1.627709984779358,
      "learning_rate": 1e-05,
      "loss": 0.2847,
      "step": 186600
    },
    {
      "epoch": 0.001867,
      "grad_norm": 1.5060405731201172,
      "learning_rate": 1e-05,
      "loss": 0.2845,
      "step": 186700
    },
    {
      "epoch": 0.001868,
      "grad_norm": 1.6034202575683594,
      "learning_rate": 1e-05,
      "loss": 0.287,
      "step": 186800
    },
    {
      "epoch": 0.001869,
      "grad_norm": 1.4927774667739868,
      "learning_rate": 1e-05,
      "loss": 0.2857,
      "step": 186900
    },
    {
      "epoch": 0.00187,
      "grad_norm": 1.4778563976287842,
      "learning_rate": 1e-05,
      "loss": 0.2844,
      "step": 187000
    },
    {
      "epoch": 0.001871,
      "grad_norm": 1.2116749286651611,
      "learning_rate": 1e-05,
      "loss": 0.2805,
      "step": 187100
    },
    {
      "epoch": 0.001872,
      "grad_norm": 1.4425694942474365,
      "learning_rate": 1e-05,
      "loss": 0.2824,
      "step": 187200
    },
    {
      "epoch": 0.001873,
      "grad_norm": 1.2100266218185425,
      "learning_rate": 1e-05,
      "loss": 0.2836,
      "step": 187300
    },
    {
      "epoch": 0.001874,
      "grad_norm": 1.548009991645813,
      "learning_rate": 1e-05,
      "loss": 0.2854,
      "step": 187400
    },
    {
      "epoch": 0.001875,
      "grad_norm": 1.5316060781478882,
      "learning_rate": 1e-05,
      "loss": 0.288,
      "step": 187500
    },
    {
      "epoch": 0.001876,
      "grad_norm": 1.5404126644134521,
      "learning_rate": 1e-05,
      "loss": 0.2892,
      "step": 187600
    },
    {
      "epoch": 0.001877,
      "grad_norm": 1.592418909072876,
      "learning_rate": 1e-05,
      "loss": 0.2818,
      "step": 187700
    },
    {
      "epoch": 0.001878,
      "grad_norm": 1.5386697053909302,
      "learning_rate": 1e-05,
      "loss": 0.2812,
      "step": 187800
    },
    {
      "epoch": 0.001879,
      "grad_norm": 1.7977592945098877,
      "learning_rate": 1e-05,
      "loss": 0.2883,
      "step": 187900
    },
    {
      "epoch": 0.00188,
      "grad_norm": 1.4943695068359375,
      "learning_rate": 1e-05,
      "loss": 0.2924,
      "step": 188000
    },
    {
      "epoch": 0.001881,
      "grad_norm": 1.7356852293014526,
      "learning_rate": 1e-05,
      "loss": 0.2861,
      "step": 188100
    },
    {
      "epoch": 0.001882,
      "grad_norm": 1.300524353981018,
      "learning_rate": 1e-05,
      "loss": 0.2871,
      "step": 188200
    },
    {
      "epoch": 0.001883,
      "grad_norm": 1.5252922773361206,
      "learning_rate": 1e-05,
      "loss": 0.293,
      "step": 188300
    },
    {
      "epoch": 0.001884,
      "grad_norm": 1.6379549503326416,
      "learning_rate": 1e-05,
      "loss": 0.2928,
      "step": 188400
    },
    {
      "epoch": 0.001885,
      "grad_norm": 1.485962152481079,
      "learning_rate": 1e-05,
      "loss": 0.2797,
      "step": 188500
    },
    {
      "epoch": 0.001886,
      "grad_norm": 1.5201560258865356,
      "learning_rate": 1e-05,
      "loss": 0.2906,
      "step": 188600
    },
    {
      "epoch": 0.001887,
      "grad_norm": 1.5450036525726318,
      "learning_rate": 1e-05,
      "loss": 0.2903,
      "step": 188700
    },
    {
      "epoch": 0.001888,
      "grad_norm": 1.6390937566757202,
      "learning_rate": 1e-05,
      "loss": 0.2824,
      "step": 188800
    },
    {
      "epoch": 0.001889,
      "grad_norm": 1.4087883234024048,
      "learning_rate": 1e-05,
      "loss": 0.2825,
      "step": 188900
    },
    {
      "epoch": 0.00189,
      "grad_norm": 1.5856794118881226,
      "learning_rate": 1e-05,
      "loss": 0.2836,
      "step": 189000
    },
    {
      "epoch": 0.001891,
      "grad_norm": 1.4078369140625,
      "learning_rate": 1e-05,
      "loss": 0.2774,
      "step": 189100
    },
    {
      "epoch": 0.001892,
      "grad_norm": 1.7400150299072266,
      "learning_rate": 1e-05,
      "loss": 0.2856,
      "step": 189200
    },
    {
      "epoch": 0.001893,
      "grad_norm": 1.580092191696167,
      "learning_rate": 1e-05,
      "loss": 0.2798,
      "step": 189300
    },
    {
      "epoch": 0.001894,
      "grad_norm": 1.6314994096755981,
      "learning_rate": 1e-05,
      "loss": 0.2876,
      "step": 189400
    },
    {
      "epoch": 0.001895,
      "grad_norm": 1.33416748046875,
      "learning_rate": 1e-05,
      "loss": 0.2809,
      "step": 189500
    },
    {
      "epoch": 0.001896,
      "grad_norm": 1.512454628944397,
      "learning_rate": 1e-05,
      "loss": 0.2784,
      "step": 189600
    },
    {
      "epoch": 0.001897,
      "grad_norm": 1.4232842922210693,
      "learning_rate": 1e-05,
      "loss": 0.2867,
      "step": 189700
    },
    {
      "epoch": 0.001898,
      "grad_norm": 1.3176745176315308,
      "learning_rate": 1e-05,
      "loss": 0.284,
      "step": 189800
    },
    {
      "epoch": 0.001899,
      "grad_norm": 1.394158959388733,
      "learning_rate": 1e-05,
      "loss": 0.2833,
      "step": 189900
    },
    {
      "epoch": 0.0019,
      "grad_norm": 1.4825711250305176,
      "learning_rate": 1e-05,
      "loss": 0.287,
      "step": 190000
    },
    {
      "epoch": 0.001901,
      "grad_norm": 1.6661640405654907,
      "learning_rate": 1e-05,
      "loss": 0.2821,
      "step": 190100
    },
    {
      "epoch": 0.001902,
      "grad_norm": 1.7882932424545288,
      "learning_rate": 1e-05,
      "loss": 0.2793,
      "step": 190200
    },
    {
      "epoch": 0.001903,
      "grad_norm": 1.774411678314209,
      "learning_rate": 1e-05,
      "loss": 0.2914,
      "step": 190300
    },
    {
      "epoch": 0.001904,
      "grad_norm": 1.4190601110458374,
      "learning_rate": 1e-05,
      "loss": 0.285,
      "step": 190400
    },
    {
      "epoch": 0.001905,
      "grad_norm": 1.6410194635391235,
      "learning_rate": 1e-05,
      "loss": 0.2836,
      "step": 190500
    },
    {
      "epoch": 0.001906,
      "grad_norm": 1.3235634565353394,
      "learning_rate": 1e-05,
      "loss": 0.2843,
      "step": 190600
    },
    {
      "epoch": 0.001907,
      "grad_norm": 1.5994374752044678,
      "learning_rate": 1e-05,
      "loss": 0.2911,
      "step": 190700
    },
    {
      "epoch": 0.001908,
      "grad_norm": 1.479262113571167,
      "learning_rate": 1e-05,
      "loss": 0.2851,
      "step": 190800
    },
    {
      "epoch": 0.001909,
      "grad_norm": 1.8383204936981201,
      "learning_rate": 1e-05,
      "loss": 0.2861,
      "step": 190900
    },
    {
      "epoch": 0.00191,
      "grad_norm": 1.7182331085205078,
      "learning_rate": 1e-05,
      "loss": 0.2892,
      "step": 191000
    },
    {
      "epoch": 0.001911,
      "grad_norm": 1.2352073192596436,
      "learning_rate": 1e-05,
      "loss": 0.2882,
      "step": 191100
    },
    {
      "epoch": 0.001912,
      "grad_norm": 1.6550449132919312,
      "learning_rate": 1e-05,
      "loss": 0.2857,
      "step": 191200
    },
    {
      "epoch": 0.001913,
      "grad_norm": 1.4845826625823975,
      "learning_rate": 1e-05,
      "loss": 0.2815,
      "step": 191300
    },
    {
      "epoch": 0.001914,
      "grad_norm": 1.3901734352111816,
      "learning_rate": 1e-05,
      "loss": 0.2775,
      "step": 191400
    },
    {
      "epoch": 0.001915,
      "grad_norm": 1.410360336303711,
      "learning_rate": 1e-05,
      "loss": 0.2842,
      "step": 191500
    },
    {
      "epoch": 0.001916,
      "grad_norm": 1.6031688451766968,
      "learning_rate": 1e-05,
      "loss": 0.2809,
      "step": 191600
    },
    {
      "epoch": 0.001917,
      "grad_norm": 1.6306028366088867,
      "learning_rate": 1e-05,
      "loss": 0.2844,
      "step": 191700
    },
    {
      "epoch": 0.001918,
      "grad_norm": 1.450675129890442,
      "learning_rate": 1e-05,
      "loss": 0.2808,
      "step": 191800
    },
    {
      "epoch": 0.001919,
      "grad_norm": 1.4375156164169312,
      "learning_rate": 1e-05,
      "loss": 0.288,
      "step": 191900
    },
    {
      "epoch": 0.00192,
      "grad_norm": 1.5712558031082153,
      "learning_rate": 1e-05,
      "loss": 0.2827,
      "step": 192000
    },
    {
      "epoch": 0.001921,
      "grad_norm": 1.6321333646774292,
      "learning_rate": 1e-05,
      "loss": 0.2853,
      "step": 192100
    },
    {
      "epoch": 0.001922,
      "grad_norm": 1.3940213918685913,
      "learning_rate": 1e-05,
      "loss": 0.2868,
      "step": 192200
    },
    {
      "epoch": 0.001923,
      "grad_norm": 1.506698489189148,
      "learning_rate": 1e-05,
      "loss": 0.2892,
      "step": 192300
    },
    {
      "epoch": 0.001924,
      "grad_norm": 1.4711730480194092,
      "learning_rate": 1e-05,
      "loss": 0.2865,
      "step": 192400
    },
    {
      "epoch": 0.001925,
      "grad_norm": 1.4492762088775635,
      "learning_rate": 1e-05,
      "loss": 0.2858,
      "step": 192500
    },
    {
      "epoch": 0.001926,
      "grad_norm": 1.5630435943603516,
      "learning_rate": 1e-05,
      "loss": 0.285,
      "step": 192600
    },
    {
      "epoch": 0.001927,
      "grad_norm": 1.4810446500778198,
      "learning_rate": 1e-05,
      "loss": 0.283,
      "step": 192700
    },
    {
      "epoch": 0.001928,
      "grad_norm": 1.5246132612228394,
      "learning_rate": 1e-05,
      "loss": 0.2844,
      "step": 192800
    },
    {
      "epoch": 0.001929,
      "grad_norm": 1.7029883861541748,
      "learning_rate": 1e-05,
      "loss": 0.28,
      "step": 192900
    },
    {
      "epoch": 0.00193,
      "grad_norm": 1.930101990699768,
      "learning_rate": 1e-05,
      "loss": 0.28,
      "step": 193000
    },
    {
      "epoch": 0.001931,
      "grad_norm": 2.439939260482788,
      "learning_rate": 1e-05,
      "loss": 0.284,
      "step": 193100
    },
    {
      "epoch": 0.001932,
      "grad_norm": 1.4808944463729858,
      "learning_rate": 1e-05,
      "loss": 0.2805,
      "step": 193200
    },
    {
      "epoch": 0.001933,
      "grad_norm": 1.3932912349700928,
      "learning_rate": 1e-05,
      "loss": 0.2787,
      "step": 193300
    },
    {
      "epoch": 0.001934,
      "grad_norm": 1.4781297445297241,
      "learning_rate": 1e-05,
      "loss": 0.2774,
      "step": 193400
    },
    {
      "epoch": 0.001935,
      "grad_norm": 1.4157606363296509,
      "learning_rate": 1e-05,
      "loss": 0.2855,
      "step": 193500
    },
    {
      "epoch": 0.001936,
      "grad_norm": 1.5318036079406738,
      "learning_rate": 1e-05,
      "loss": 0.281,
      "step": 193600
    },
    {
      "epoch": 0.001937,
      "grad_norm": 1.4803863763809204,
      "learning_rate": 1e-05,
      "loss": 0.2792,
      "step": 193700
    },
    {
      "epoch": 0.001938,
      "grad_norm": 1.4421052932739258,
      "learning_rate": 1e-05,
      "loss": 0.2888,
      "step": 193800
    },
    {
      "epoch": 0.001939,
      "grad_norm": 1.336422085762024,
      "learning_rate": 1e-05,
      "loss": 0.2798,
      "step": 193900
    },
    {
      "epoch": 0.00194,
      "grad_norm": 1.4723069667816162,
      "learning_rate": 1e-05,
      "loss": 0.2826,
      "step": 194000
    },
    {
      "epoch": 0.001941,
      "grad_norm": 1.549203634262085,
      "learning_rate": 1e-05,
      "loss": 0.2804,
      "step": 194100
    },
    {
      "epoch": 0.001942,
      "grad_norm": 1.5718244314193726,
      "learning_rate": 1e-05,
      "loss": 0.2773,
      "step": 194200
    },
    {
      "epoch": 0.001943,
      "grad_norm": 1.4759576320648193,
      "learning_rate": 1e-05,
      "loss": 0.2773,
      "step": 194300
    },
    {
      "epoch": 0.001944,
      "grad_norm": 1.3965938091278076,
      "learning_rate": 1e-05,
      "loss": 0.2771,
      "step": 194400
    },
    {
      "epoch": 0.001945,
      "grad_norm": 1.3787996768951416,
      "learning_rate": 1e-05,
      "loss": 0.2806,
      "step": 194500
    },
    {
      "epoch": 0.001946,
      "grad_norm": 1.3691178560256958,
      "learning_rate": 1e-05,
      "loss": 0.2892,
      "step": 194600
    },
    {
      "epoch": 0.001947,
      "grad_norm": 1.4109934568405151,
      "learning_rate": 1e-05,
      "loss": 0.2902,
      "step": 194700
    },
    {
      "epoch": 0.001948,
      "grad_norm": 1.4349749088287354,
      "learning_rate": 1e-05,
      "loss": 0.2852,
      "step": 194800
    },
    {
      "epoch": 0.001949,
      "grad_norm": 1.3615974187850952,
      "learning_rate": 1e-05,
      "loss": 0.2808,
      "step": 194900
    },
    {
      "epoch": 0.00195,
      "grad_norm": 1.5208380222320557,
      "learning_rate": 1e-05,
      "loss": 0.2804,
      "step": 195000
    },
    {
      "epoch": 0.001951,
      "grad_norm": 1.5616921186447144,
      "learning_rate": 1e-05,
      "loss": 0.2793,
      "step": 195100
    },
    {
      "epoch": 0.001952,
      "grad_norm": 1.490240454673767,
      "learning_rate": 1e-05,
      "loss": 0.2835,
      "step": 195200
    },
    {
      "epoch": 0.001953,
      "grad_norm": 1.4141552448272705,
      "learning_rate": 1e-05,
      "loss": 0.2841,
      "step": 195300
    },
    {
      "epoch": 0.001954,
      "grad_norm": 1.4841254949569702,
      "learning_rate": 1e-05,
      "loss": 0.2899,
      "step": 195400
    },
    {
      "epoch": 0.001955,
      "grad_norm": 1.3822132349014282,
      "learning_rate": 1e-05,
      "loss": 0.2873,
      "step": 195500
    },
    {
      "epoch": 0.001956,
      "grad_norm": 1.5400711297988892,
      "learning_rate": 1e-05,
      "loss": 0.2816,
      "step": 195600
    },
    {
      "epoch": 0.001957,
      "grad_norm": 3.026294708251953,
      "learning_rate": 1e-05,
      "loss": 0.2818,
      "step": 195700
    },
    {
      "epoch": 0.001958,
      "grad_norm": 1.5581517219543457,
      "learning_rate": 1e-05,
      "loss": 0.2866,
      "step": 195800
    },
    {
      "epoch": 0.001959,
      "grad_norm": 1.292336106300354,
      "learning_rate": 1e-05,
      "loss": 0.2807,
      "step": 195900
    },
    {
      "epoch": 0.00196,
      "grad_norm": 1.3840731382369995,
      "learning_rate": 1e-05,
      "loss": 0.2799,
      "step": 196000
    },
    {
      "epoch": 0.001961,
      "grad_norm": 1.82817542552948,
      "learning_rate": 1e-05,
      "loss": 0.2871,
      "step": 196100
    },
    {
      "epoch": 0.001962,
      "grad_norm": 2.580714702606201,
      "learning_rate": 1e-05,
      "loss": 0.2752,
      "step": 196200
    },
    {
      "epoch": 0.001963,
      "grad_norm": 1.8855743408203125,
      "learning_rate": 1e-05,
      "loss": 0.2822,
      "step": 196300
    },
    {
      "epoch": 0.001964,
      "grad_norm": 1.3740893602371216,
      "learning_rate": 1e-05,
      "loss": 0.2811,
      "step": 196400
    },
    {
      "epoch": 0.001965,
      "grad_norm": 1.2532157897949219,
      "learning_rate": 1e-05,
      "loss": 0.2889,
      "step": 196500
    },
    {
      "epoch": 0.001966,
      "grad_norm": 1.313609004020691,
      "learning_rate": 1e-05,
      "loss": 0.2818,
      "step": 196600
    },
    {
      "epoch": 0.001967,
      "grad_norm": 1.326478123664856,
      "learning_rate": 1e-05,
      "loss": 0.287,
      "step": 196700
    },
    {
      "epoch": 0.001968,
      "grad_norm": 1.5435999631881714,
      "learning_rate": 1e-05,
      "loss": 0.2848,
      "step": 196800
    },
    {
      "epoch": 0.001969,
      "grad_norm": 1.5859767198562622,
      "learning_rate": 1e-05,
      "loss": 0.282,
      "step": 196900
    },
    {
      "epoch": 0.00197,
      "grad_norm": 1.3006385564804077,
      "learning_rate": 1e-05,
      "loss": 0.2793,
      "step": 197000
    },
    {
      "epoch": 0.001971,
      "grad_norm": 1.2609304189682007,
      "learning_rate": 1e-05,
      "loss": 0.2791,
      "step": 197100
    },
    {
      "epoch": 0.001972,
      "grad_norm": 1.4160981178283691,
      "learning_rate": 1e-05,
      "loss": 0.2832,
      "step": 197200
    },
    {
      "epoch": 0.001973,
      "grad_norm": 2.234437942504883,
      "learning_rate": 1e-05,
      "loss": 0.2852,
      "step": 197300
    },
    {
      "epoch": 0.001974,
      "grad_norm": 1.295417070388794,
      "learning_rate": 1e-05,
      "loss": 0.2825,
      "step": 197400
    },
    {
      "epoch": 0.001975,
      "grad_norm": 1.227103352546692,
      "learning_rate": 1e-05,
      "loss": 0.2787,
      "step": 197500
    },
    {
      "epoch": 0.001976,
      "grad_norm": 1.5311518907546997,
      "learning_rate": 1e-05,
      "loss": 0.2807,
      "step": 197600
    },
    {
      "epoch": 0.001977,
      "grad_norm": 1.3764680624008179,
      "learning_rate": 1e-05,
      "loss": 0.2772,
      "step": 197700
    },
    {
      "epoch": 0.001978,
      "grad_norm": 1.43855619430542,
      "learning_rate": 1e-05,
      "loss": 0.2806,
      "step": 197800
    },
    {
      "epoch": 0.001979,
      "grad_norm": 1.5360887050628662,
      "learning_rate": 1e-05,
      "loss": 0.2774,
      "step": 197900
    },
    {
      "epoch": 0.00198,
      "grad_norm": 1.2137932777404785,
      "learning_rate": 1e-05,
      "loss": 0.2833,
      "step": 198000
    },
    {
      "epoch": 0.001981,
      "grad_norm": 1.5310431718826294,
      "learning_rate": 1e-05,
      "loss": 0.2795,
      "step": 198100
    },
    {
      "epoch": 0.001982,
      "grad_norm": 1.6440815925598145,
      "learning_rate": 1e-05,
      "loss": 0.2814,
      "step": 198200
    },
    {
      "epoch": 0.001983,
      "grad_norm": 1.8036118745803833,
      "learning_rate": 1e-05,
      "loss": 0.2835,
      "step": 198300
    },
    {
      "epoch": 0.001984,
      "grad_norm": 1.3439860343933105,
      "learning_rate": 1e-05,
      "loss": 0.2829,
      "step": 198400
    },
    {
      "epoch": 0.001985,
      "grad_norm": 1.4137046337127686,
      "learning_rate": 1e-05,
      "loss": 0.2814,
      "step": 198500
    },
    {
      "epoch": 0.001986,
      "grad_norm": 1.4916012287139893,
      "learning_rate": 1e-05,
      "loss": 0.277,
      "step": 198600
    },
    {
      "epoch": 0.001987,
      "grad_norm": 1.4562140703201294,
      "learning_rate": 1e-05,
      "loss": 0.2866,
      "step": 198700
    },
    {
      "epoch": 0.001988,
      "grad_norm": 1.6629130840301514,
      "learning_rate": 1e-05,
      "loss": 0.2775,
      "step": 198800
    },
    {
      "epoch": 0.001989,
      "grad_norm": 1.524603247642517,
      "learning_rate": 1e-05,
      "loss": 0.2808,
      "step": 198900
    },
    {
      "epoch": 0.00199,
      "grad_norm": 2.1545159816741943,
      "learning_rate": 1e-05,
      "loss": 0.2791,
      "step": 199000
    },
    {
      "epoch": 0.001991,
      "grad_norm": 1.3595882654190063,
      "learning_rate": 1e-05,
      "loss": 0.2845,
      "step": 199100
    },
    {
      "epoch": 0.001992,
      "grad_norm": 1.4160555601119995,
      "learning_rate": 1e-05,
      "loss": 0.2779,
      "step": 199200
    },
    {
      "epoch": 0.001993,
      "grad_norm": 1.612523078918457,
      "learning_rate": 1e-05,
      "loss": 0.2878,
      "step": 199300
    },
    {
      "epoch": 0.001994,
      "grad_norm": 1.4364423751831055,
      "learning_rate": 1e-05,
      "loss": 0.2813,
      "step": 199400
    },
    {
      "epoch": 0.001995,
      "grad_norm": 1.47967529296875,
      "learning_rate": 1e-05,
      "loss": 0.2841,
      "step": 199500
    },
    {
      "epoch": 0.001996,
      "grad_norm": 1.621627688407898,
      "learning_rate": 1e-05,
      "loss": 0.2765,
      "step": 199600
    },
    {
      "epoch": 0.001997,
      "grad_norm": 1.3066225051879883,
      "learning_rate": 1e-05,
      "loss": 0.2828,
      "step": 199700
    },
    {
      "epoch": 0.001998,
      "grad_norm": 1.614108681678772,
      "learning_rate": 1e-05,
      "loss": 0.2808,
      "step": 199800
    },
    {
      "epoch": 0.001999,
      "grad_norm": 1.457350730895996,
      "learning_rate": 1e-05,
      "loss": 0.2815,
      "step": 199900
    },
    {
      "epoch": 0.002,
      "grad_norm": 1.6253563165664673,
      "learning_rate": 1e-05,
      "loss": 0.2792,
      "step": 200000
    },
    {
      "epoch": 0.002,
      "eval_loss": 0.256103515625,
      "eval_runtime": 115.1679,
      "eval_samples_per_second": 434.149,
      "eval_steps_per_second": 27.134,
      "step": 200000
    },
    {
      "epoch": 0.002001,
      "grad_norm": 1.5363661050796509,
      "learning_rate": 1e-05,
      "loss": 0.2754,
      "step": 200100
    },
    {
      "epoch": 0.002002,
      "grad_norm": 1.5348174571990967,
      "learning_rate": 1e-05,
      "loss": 0.2851,
      "step": 200200
    },
    {
      "epoch": 0.002003,
      "grad_norm": 1.4772671461105347,
      "learning_rate": 1e-05,
      "loss": 0.279,
      "step": 200300
    },
    {
      "epoch": 0.002004,
      "grad_norm": 1.5636744499206543,
      "learning_rate": 1e-05,
      "loss": 0.2862,
      "step": 200400
    },
    {
      "epoch": 0.002005,
      "grad_norm": 1.8320351839065552,
      "learning_rate": 1e-05,
      "loss": 0.279,
      "step": 200500
    },
    {
      "epoch": 0.002006,
      "grad_norm": 1.7066452503204346,
      "learning_rate": 1e-05,
      "loss": 0.2757,
      "step": 200600
    },
    {
      "epoch": 0.002007,
      "grad_norm": 1.3273617029190063,
      "learning_rate": 1e-05,
      "loss": 0.2815,
      "step": 200700
    },
    {
      "epoch": 0.002008,
      "grad_norm": 1.492756724357605,
      "learning_rate": 1e-05,
      "loss": 0.2779,
      "step": 200800
    },
    {
      "epoch": 0.002009,
      "grad_norm": 1.5056836605072021,
      "learning_rate": 1e-05,
      "loss": 0.2805,
      "step": 200900
    },
    {
      "epoch": 0.00201,
      "grad_norm": 1.504055142402649,
      "learning_rate": 1e-05,
      "loss": 0.2725,
      "step": 201000
    },
    {
      "epoch": 0.002011,
      "grad_norm": 1.424229621887207,
      "learning_rate": 1e-05,
      "loss": 0.2768,
      "step": 201100
    },
    {
      "epoch": 0.002012,
      "grad_norm": 1.6440435647964478,
      "learning_rate": 1e-05,
      "loss": 0.2776,
      "step": 201200
    },
    {
      "epoch": 0.002013,
      "grad_norm": 1.4276853799819946,
      "learning_rate": 1e-05,
      "loss": 0.2737,
      "step": 201300
    },
    {
      "epoch": 0.002014,
      "grad_norm": 1.479280948638916,
      "learning_rate": 1e-05,
      "loss": 0.2772,
      "step": 201400
    },
    {
      "epoch": 0.002015,
      "grad_norm": 1.7645676136016846,
      "learning_rate": 1e-05,
      "loss": 0.2779,
      "step": 201500
    },
    {
      "epoch": 0.002016,
      "grad_norm": 1.7327306270599365,
      "learning_rate": 1e-05,
      "loss": 0.2809,
      "step": 201600
    },
    {
      "epoch": 0.002017,
      "grad_norm": 1.4628037214279175,
      "learning_rate": 1e-05,
      "loss": 0.2815,
      "step": 201700
    },
    {
      "epoch": 0.002018,
      "grad_norm": 1.2225489616394043,
      "learning_rate": 1e-05,
      "loss": 0.286,
      "step": 201800
    },
    {
      "epoch": 0.002019,
      "grad_norm": 1.4493396282196045,
      "learning_rate": 1e-05,
      "loss": 0.2781,
      "step": 201900
    },
    {
      "epoch": 0.00202,
      "grad_norm": 1.1203590631484985,
      "learning_rate": 1e-05,
      "loss": 0.2782,
      "step": 202000
    },
    {
      "epoch": 0.002021,
      "grad_norm": 1.4502707719802856,
      "learning_rate": 1e-05,
      "loss": 0.277,
      "step": 202100
    },
    {
      "epoch": 0.002022,
      "grad_norm": 1.7395519018173218,
      "learning_rate": 1e-05,
      "loss": 0.286,
      "step": 202200
    },
    {
      "epoch": 0.002023,
      "grad_norm": 1.9898625612258911,
      "learning_rate": 1e-05,
      "loss": 0.2721,
      "step": 202300
    },
    {
      "epoch": 0.002024,
      "grad_norm": 1.6557375192642212,
      "learning_rate": 1e-05,
      "loss": 0.2752,
      "step": 202400
    },
    {
      "epoch": 0.002025,
      "grad_norm": 1.4543732404708862,
      "learning_rate": 1e-05,
      "loss": 0.2872,
      "step": 202500
    },
    {
      "epoch": 0.002026,
      "grad_norm": 1.44635009765625,
      "learning_rate": 1e-05,
      "loss": 0.2745,
      "step": 202600
    },
    {
      "epoch": 0.002027,
      "grad_norm": 1.3892033100128174,
      "learning_rate": 1e-05,
      "loss": 0.2804,
      "step": 202700
    },
    {
      "epoch": 0.002028,
      "grad_norm": 1.2940878868103027,
      "learning_rate": 1e-05,
      "loss": 0.2778,
      "step": 202800
    },
    {
      "epoch": 0.002029,
      "grad_norm": 2.028639554977417,
      "learning_rate": 1e-05,
      "loss": 0.2741,
      "step": 202900
    },
    {
      "epoch": 0.00203,
      "grad_norm": 1.462415099143982,
      "learning_rate": 1e-05,
      "loss": 0.2775,
      "step": 203000
    },
    {
      "epoch": 0.002031,
      "grad_norm": 1.60899019241333,
      "learning_rate": 1e-05,
      "loss": 0.2764,
      "step": 203100
    },
    {
      "epoch": 0.002032,
      "grad_norm": 1.5385807752609253,
      "learning_rate": 1e-05,
      "loss": 0.2794,
      "step": 203200
    },
    {
      "epoch": 0.002033,
      "grad_norm": 1.3882631063461304,
      "learning_rate": 1e-05,
      "loss": 0.2736,
      "step": 203300
    },
    {
      "epoch": 0.002034,
      "grad_norm": 1.2531319856643677,
      "learning_rate": 1e-05,
      "loss": 0.2848,
      "step": 203400
    },
    {
      "epoch": 0.002035,
      "grad_norm": 1.621891736984253,
      "learning_rate": 1e-05,
      "loss": 0.2773,
      "step": 203500
    },
    {
      "epoch": 0.002036,
      "grad_norm": 1.3949509859085083,
      "learning_rate": 1e-05,
      "loss": 0.2804,
      "step": 203600
    },
    {
      "epoch": 0.002037,
      "grad_norm": 1.3484961986541748,
      "learning_rate": 1e-05,
      "loss": 0.2785,
      "step": 203700
    },
    {
      "epoch": 0.002038,
      "grad_norm": 1.524052381515503,
      "learning_rate": 1e-05,
      "loss": 0.2817,
      "step": 203800
    },
    {
      "epoch": 0.002039,
      "grad_norm": 1.4757261276245117,
      "learning_rate": 1e-05,
      "loss": 0.2786,
      "step": 203900
    },
    {
      "epoch": 0.00204,
      "grad_norm": 1.7308425903320312,
      "learning_rate": 1e-05,
      "loss": 0.2765,
      "step": 204000
    },
    {
      "epoch": 0.002041,
      "grad_norm": 1.433393120765686,
      "learning_rate": 1e-05,
      "loss": 0.2745,
      "step": 204100
    },
    {
      "epoch": 0.002042,
      "grad_norm": 1.398602843284607,
      "learning_rate": 1e-05,
      "loss": 0.278,
      "step": 204200
    },
    {
      "epoch": 0.002043,
      "grad_norm": 1.6153137683868408,
      "learning_rate": 1e-05,
      "loss": 0.2814,
      "step": 204300
    },
    {
      "epoch": 0.002044,
      "grad_norm": 1.516076683998108,
      "learning_rate": 1e-05,
      "loss": 0.2709,
      "step": 204400
    },
    {
      "epoch": 0.002045,
      "grad_norm": 1.6530369520187378,
      "learning_rate": 1e-05,
      "loss": 0.2824,
      "step": 204500
    },
    {
      "epoch": 0.002046,
      "grad_norm": 1.4678118228912354,
      "learning_rate": 1e-05,
      "loss": 0.2754,
      "step": 204600
    },
    {
      "epoch": 0.002047,
      "grad_norm": 1.64580237865448,
      "learning_rate": 1e-05,
      "loss": 0.2796,
      "step": 204700
    },
    {
      "epoch": 0.002048,
      "grad_norm": 1.6422346830368042,
      "learning_rate": 1e-05,
      "loss": 0.2742,
      "step": 204800
    },
    {
      "epoch": 0.002049,
      "grad_norm": 1.5554474592208862,
      "learning_rate": 1e-05,
      "loss": 0.2838,
      "step": 204900
    },
    {
      "epoch": 0.00205,
      "grad_norm": 1.4247395992279053,
      "learning_rate": 1e-05,
      "loss": 0.2788,
      "step": 205000
    },
    {
      "epoch": 0.002051,
      "grad_norm": 3.377981662750244,
      "learning_rate": 1e-05,
      "loss": 0.2739,
      "step": 205100
    },
    {
      "epoch": 0.002052,
      "grad_norm": 1.668649435043335,
      "learning_rate": 1e-05,
      "loss": 0.2817,
      "step": 205200
    },
    {
      "epoch": 0.002053,
      "grad_norm": 1.3109833002090454,
      "learning_rate": 1e-05,
      "loss": 0.281,
      "step": 205300
    },
    {
      "epoch": 0.002054,
      "grad_norm": 1.3675384521484375,
      "learning_rate": 1e-05,
      "loss": 0.2686,
      "step": 205400
    },
    {
      "epoch": 0.002055,
      "grad_norm": 1.542527437210083,
      "learning_rate": 1e-05,
      "loss": 0.2709,
      "step": 205500
    },
    {
      "epoch": 0.002056,
      "grad_norm": 1.3900882005691528,
      "learning_rate": 1e-05,
      "loss": 0.2841,
      "step": 205600
    },
    {
      "epoch": 0.002057,
      "grad_norm": 1.5033669471740723,
      "learning_rate": 1e-05,
      "loss": 0.2798,
      "step": 205700
    },
    {
      "epoch": 0.002058,
      "grad_norm": 1.638098120689392,
      "learning_rate": 1e-05,
      "loss": 0.275,
      "step": 205800
    },
    {
      "epoch": 0.002059,
      "grad_norm": 1.4195244312286377,
      "learning_rate": 1e-05,
      "loss": 0.2781,
      "step": 205900
    },
    {
      "epoch": 0.00206,
      "grad_norm": 1.5741572380065918,
      "learning_rate": 1e-05,
      "loss": 0.2811,
      "step": 206000
    },
    {
      "epoch": 0.002061,
      "grad_norm": 1.4734532833099365,
      "learning_rate": 1e-05,
      "loss": 0.2757,
      "step": 206100
    },
    {
      "epoch": 0.002062,
      "grad_norm": 1.4549283981323242,
      "learning_rate": 1e-05,
      "loss": 0.2797,
      "step": 206200
    },
    {
      "epoch": 0.002063,
      "grad_norm": 1.3052780628204346,
      "learning_rate": 1e-05,
      "loss": 0.2819,
      "step": 206300
    },
    {
      "epoch": 0.002064,
      "grad_norm": 1.349663257598877,
      "learning_rate": 1e-05,
      "loss": 0.2726,
      "step": 206400
    },
    {
      "epoch": 0.002065,
      "grad_norm": 1.477702021598816,
      "learning_rate": 1e-05,
      "loss": 0.278,
      "step": 206500
    },
    {
      "epoch": 0.002066,
      "grad_norm": 1.4143027067184448,
      "learning_rate": 1e-05,
      "loss": 0.2789,
      "step": 206600
    },
    {
      "epoch": 0.002067,
      "grad_norm": 1.456390380859375,
      "learning_rate": 1e-05,
      "loss": 0.2764,
      "step": 206700
    },
    {
      "epoch": 0.002068,
      "grad_norm": 1.343540906906128,
      "learning_rate": 1e-05,
      "loss": 0.2773,
      "step": 206800
    },
    {
      "epoch": 0.002069,
      "grad_norm": 1.3482036590576172,
      "learning_rate": 1e-05,
      "loss": 0.2727,
      "step": 206900
    },
    {
      "epoch": 0.00207,
      "grad_norm": 1.5749857425689697,
      "learning_rate": 1e-05,
      "loss": 0.2752,
      "step": 207000
    },
    {
      "epoch": 0.002071,
      "grad_norm": 1.458197832107544,
      "learning_rate": 1e-05,
      "loss": 0.2803,
      "step": 207100
    },
    {
      "epoch": 0.002072,
      "grad_norm": 1.5353132486343384,
      "learning_rate": 1e-05,
      "loss": 0.2698,
      "step": 207200
    },
    {
      "epoch": 0.002073,
      "grad_norm": 1.5403767824172974,
      "learning_rate": 1e-05,
      "loss": 0.2768,
      "step": 207300
    },
    {
      "epoch": 0.002074,
      "grad_norm": 1.4076712131500244,
      "learning_rate": 1e-05,
      "loss": 0.2783,
      "step": 207400
    },
    {
      "epoch": 0.002075,
      "grad_norm": 1.5805490016937256,
      "learning_rate": 1e-05,
      "loss": 0.2782,
      "step": 207500
    },
    {
      "epoch": 0.002076,
      "grad_norm": 1.342297911643982,
      "learning_rate": 1e-05,
      "loss": 0.2791,
      "step": 207600
    },
    {
      "epoch": 0.002077,
      "grad_norm": 1.5067929029464722,
      "learning_rate": 1e-05,
      "loss": 0.2768,
      "step": 207700
    },
    {
      "epoch": 0.002078,
      "grad_norm": 1.2661635875701904,
      "learning_rate": 1e-05,
      "loss": 0.2719,
      "step": 207800
    },
    {
      "epoch": 0.002079,
      "grad_norm": 1.2550525665283203,
      "learning_rate": 1e-05,
      "loss": 0.2734,
      "step": 207900
    },
    {
      "epoch": 0.00208,
      "grad_norm": 1.478445053100586,
      "learning_rate": 1e-05,
      "loss": 0.2722,
      "step": 208000
    },
    {
      "epoch": 0.002081,
      "grad_norm": 1.4284790754318237,
      "learning_rate": 1e-05,
      "loss": 0.2698,
      "step": 208100
    },
    {
      "epoch": 0.002082,
      "grad_norm": 1.6922401189804077,
      "learning_rate": 1e-05,
      "loss": 0.2813,
      "step": 208200
    },
    {
      "epoch": 0.002083,
      "grad_norm": 1.50126314163208,
      "learning_rate": 1e-05,
      "loss": 0.2797,
      "step": 208300
    },
    {
      "epoch": 0.002084,
      "grad_norm": 1.4059314727783203,
      "learning_rate": 1e-05,
      "loss": 0.2759,
      "step": 208400
    },
    {
      "epoch": 0.002085,
      "grad_norm": 1.5741177797317505,
      "learning_rate": 1e-05,
      "loss": 0.2758,
      "step": 208500
    },
    {
      "epoch": 0.002086,
      "grad_norm": 1.491023063659668,
      "learning_rate": 1e-05,
      "loss": 0.2769,
      "step": 208600
    },
    {
      "epoch": 0.002087,
      "grad_norm": 1.4003244638442993,
      "learning_rate": 1e-05,
      "loss": 0.2726,
      "step": 208700
    },
    {
      "epoch": 0.002088,
      "grad_norm": 1.695166826248169,
      "learning_rate": 1e-05,
      "loss": 0.2795,
      "step": 208800
    },
    {
      "epoch": 0.002089,
      "grad_norm": 1.3839768171310425,
      "learning_rate": 1e-05,
      "loss": 0.2716,
      "step": 208900
    },
    {
      "epoch": 0.00209,
      "grad_norm": 1.2503012418746948,
      "learning_rate": 1e-05,
      "loss": 0.2749,
      "step": 209000
    },
    {
      "epoch": 0.002091,
      "grad_norm": 1.3623175621032715,
      "learning_rate": 1e-05,
      "loss": 0.2772,
      "step": 209100
    },
    {
      "epoch": 0.002092,
      "grad_norm": 1.3409157991409302,
      "learning_rate": 1e-05,
      "loss": 0.2735,
      "step": 209200
    },
    {
      "epoch": 0.002093,
      "grad_norm": 1.5793631076812744,
      "learning_rate": 1e-05,
      "loss": 0.2813,
      "step": 209300
    },
    {
      "epoch": 0.002094,
      "grad_norm": 1.5260995626449585,
      "learning_rate": 1e-05,
      "loss": 0.2626,
      "step": 209400
    },
    {
      "epoch": 0.002095,
      "grad_norm": 1.5600001811981201,
      "learning_rate": 1e-05,
      "loss": 0.2792,
      "step": 209500
    },
    {
      "epoch": 0.002096,
      "grad_norm": 1.5654038190841675,
      "learning_rate": 1e-05,
      "loss": 0.2796,
      "step": 209600
    },
    {
      "epoch": 0.002097,
      "grad_norm": 1.3999748229980469,
      "learning_rate": 1e-05,
      "loss": 0.2745,
      "step": 209700
    },
    {
      "epoch": 0.002098,
      "grad_norm": 1.4651896953582764,
      "learning_rate": 1e-05,
      "loss": 0.2809,
      "step": 209800
    },
    {
      "epoch": 0.002099,
      "grad_norm": 2.7390918731689453,
      "learning_rate": 1e-05,
      "loss": 0.2699,
      "step": 209900
    },
    {
      "epoch": 0.0021,
      "grad_norm": 1.6523025035858154,
      "learning_rate": 1e-05,
      "loss": 0.2718,
      "step": 210000
    },
    {
      "epoch": 0.002101,
      "grad_norm": 1.4938172101974487,
      "learning_rate": 1e-05,
      "loss": 0.278,
      "step": 210100
    },
    {
      "epoch": 0.002102,
      "grad_norm": 1.3947250843048096,
      "learning_rate": 1e-05,
      "loss": 0.2638,
      "step": 210200
    },
    {
      "epoch": 0.002103,
      "grad_norm": 1.4332685470581055,
      "learning_rate": 1e-05,
      "loss": 0.2776,
      "step": 210300
    },
    {
      "epoch": 0.002104,
      "grad_norm": 1.5522576570510864,
      "learning_rate": 1e-05,
      "loss": 0.2715,
      "step": 210400
    },
    {
      "epoch": 0.002105,
      "grad_norm": 1.2607855796813965,
      "learning_rate": 1e-05,
      "loss": 0.2801,
      "step": 210500
    },
    {
      "epoch": 0.002106,
      "grad_norm": 1.489503026008606,
      "learning_rate": 1e-05,
      "loss": 0.2772,
      "step": 210600
    },
    {
      "epoch": 0.002107,
      "grad_norm": 1.315061092376709,
      "learning_rate": 1e-05,
      "loss": 0.2785,
      "step": 210700
    },
    {
      "epoch": 0.002108,
      "grad_norm": 1.56254243850708,
      "learning_rate": 1e-05,
      "loss": 0.2724,
      "step": 210800
    },
    {
      "epoch": 0.002109,
      "grad_norm": 1.4691557884216309,
      "learning_rate": 1e-05,
      "loss": 0.2741,
      "step": 210900
    },
    {
      "epoch": 0.00211,
      "grad_norm": 1.551344633102417,
      "learning_rate": 1e-05,
      "loss": 0.2802,
      "step": 211000
    },
    {
      "epoch": 0.002111,
      "grad_norm": 1.3702019453048706,
      "learning_rate": 1e-05,
      "loss": 0.275,
      "step": 211100
    },
    {
      "epoch": 0.002112,
      "grad_norm": 1.563392996788025,
      "learning_rate": 1e-05,
      "loss": 0.2778,
      "step": 211200
    },
    {
      "epoch": 0.002113,
      "grad_norm": 1.403418779373169,
      "learning_rate": 1e-05,
      "loss": 0.2751,
      "step": 211300
    },
    {
      "epoch": 0.002114,
      "grad_norm": 1.391108512878418,
      "learning_rate": 1e-05,
      "loss": 0.2762,
      "step": 211400
    },
    {
      "epoch": 0.002115,
      "grad_norm": 1.3116768598556519,
      "learning_rate": 1e-05,
      "loss": 0.2719,
      "step": 211500
    },
    {
      "epoch": 0.002116,
      "grad_norm": 1.449575424194336,
      "learning_rate": 1e-05,
      "loss": 0.2787,
      "step": 211600
    },
    {
      "epoch": 0.002117,
      "grad_norm": 1.7022771835327148,
      "learning_rate": 1e-05,
      "loss": 0.2728,
      "step": 211700
    },
    {
      "epoch": 0.002118,
      "grad_norm": 1.4799153804779053,
      "learning_rate": 1e-05,
      "loss": 0.2734,
      "step": 211800
    },
    {
      "epoch": 0.002119,
      "grad_norm": 1.2987920045852661,
      "learning_rate": 1e-05,
      "loss": 0.2739,
      "step": 211900
    },
    {
      "epoch": 0.00212,
      "grad_norm": 1.3577724695205688,
      "learning_rate": 1e-05,
      "loss": 0.2772,
      "step": 212000
    },
    {
      "epoch": 0.002121,
      "grad_norm": 1.3344751596450806,
      "learning_rate": 1e-05,
      "loss": 0.2714,
      "step": 212100
    },
    {
      "epoch": 0.002122,
      "grad_norm": 1.6037973165512085,
      "learning_rate": 1e-05,
      "loss": 0.2761,
      "step": 212200
    },
    {
      "epoch": 0.002123,
      "grad_norm": 1.3923399448394775,
      "learning_rate": 1e-05,
      "loss": 0.2758,
      "step": 212300
    },
    {
      "epoch": 0.002124,
      "grad_norm": 8.393562316894531,
      "learning_rate": 1e-05,
      "loss": 0.2752,
      "step": 212400
    },
    {
      "epoch": 0.002125,
      "grad_norm": 1.412778615951538,
      "learning_rate": 1e-05,
      "loss": 0.2771,
      "step": 212500
    },
    {
      "epoch": 0.002126,
      "grad_norm": 1.3804876804351807,
      "learning_rate": 1e-05,
      "loss": 0.2745,
      "step": 212600
    },
    {
      "epoch": 0.002127,
      "grad_norm": 1.4782977104187012,
      "learning_rate": 1e-05,
      "loss": 0.2773,
      "step": 212700
    },
    {
      "epoch": 0.002128,
      "grad_norm": 1.4774922132492065,
      "learning_rate": 1e-05,
      "loss": 0.2782,
      "step": 212800
    },
    {
      "epoch": 0.002129,
      "grad_norm": 1.4704676866531372,
      "learning_rate": 1e-05,
      "loss": 0.2703,
      "step": 212900
    },
    {
      "epoch": 0.00213,
      "grad_norm": 1.2644281387329102,
      "learning_rate": 1e-05,
      "loss": 0.2716,
      "step": 213000
    },
    {
      "epoch": 0.002131,
      "grad_norm": 1.6625502109527588,
      "learning_rate": 1e-05,
      "loss": 0.277,
      "step": 213100
    },
    {
      "epoch": 0.002132,
      "grad_norm": 1.3630250692367554,
      "learning_rate": 1e-05,
      "loss": 0.2773,
      "step": 213200
    },
    {
      "epoch": 0.002133,
      "grad_norm": 1.6373766660690308,
      "learning_rate": 1e-05,
      "loss": 0.2701,
      "step": 213300
    },
    {
      "epoch": 0.002134,
      "grad_norm": 1.4623923301696777,
      "learning_rate": 1e-05,
      "loss": 0.2655,
      "step": 213400
    },
    {
      "epoch": 0.002135,
      "grad_norm": 1.6090341806411743,
      "learning_rate": 1e-05,
      "loss": 0.2739,
      "step": 213500
    },
    {
      "epoch": 0.002136,
      "grad_norm": 1.5013991594314575,
      "learning_rate": 1e-05,
      "loss": 0.2708,
      "step": 213600
    },
    {
      "epoch": 0.002137,
      "grad_norm": 1.26161527633667,
      "learning_rate": 1e-05,
      "loss": 0.2757,
      "step": 213700
    },
    {
      "epoch": 0.002138,
      "grad_norm": 1.418388843536377,
      "learning_rate": 1e-05,
      "loss": 0.2761,
      "step": 213800
    },
    {
      "epoch": 0.002139,
      "grad_norm": 1.293133020401001,
      "learning_rate": 1e-05,
      "loss": 0.2855,
      "step": 213900
    },
    {
      "epoch": 0.00214,
      "grad_norm": 1.9040838479995728,
      "learning_rate": 1e-05,
      "loss": 0.2761,
      "step": 214000
    },
    {
      "epoch": 0.002141,
      "grad_norm": 1.5187069177627563,
      "learning_rate": 1e-05,
      "loss": 0.2758,
      "step": 214100
    },
    {
      "epoch": 0.002142,
      "grad_norm": 1.3588346242904663,
      "learning_rate": 1e-05,
      "loss": 0.2709,
      "step": 214200
    },
    {
      "epoch": 0.002143,
      "grad_norm": 1.5500917434692383,
      "learning_rate": 1e-05,
      "loss": 0.2794,
      "step": 214300
    },
    {
      "epoch": 0.002144,
      "grad_norm": 1.6189924478530884,
      "learning_rate": 1e-05,
      "loss": 0.273,
      "step": 214400
    },
    {
      "epoch": 0.002145,
      "grad_norm": 1.7019140720367432,
      "learning_rate": 1e-05,
      "loss": 0.2763,
      "step": 214500
    },
    {
      "epoch": 0.002146,
      "grad_norm": 1.259387493133545,
      "learning_rate": 1e-05,
      "loss": 0.2776,
      "step": 214600
    },
    {
      "epoch": 0.002147,
      "grad_norm": 2.1989023685455322,
      "learning_rate": 1e-05,
      "loss": 0.2803,
      "step": 214700
    },
    {
      "epoch": 0.002148,
      "grad_norm": 2.2808122634887695,
      "learning_rate": 1e-05,
      "loss": 0.2789,
      "step": 214800
    },
    {
      "epoch": 0.002149,
      "grad_norm": 1.5754348039627075,
      "learning_rate": 1e-05,
      "loss": 0.2754,
      "step": 214900
    },
    {
      "epoch": 0.00215,
      "grad_norm": 1.4869142770767212,
      "learning_rate": 1e-05,
      "loss": 0.2714,
      "step": 215000
    },
    {
      "epoch": 0.002151,
      "grad_norm": 1.4373691082000732,
      "learning_rate": 1e-05,
      "loss": 0.2723,
      "step": 215100
    },
    {
      "epoch": 0.002152,
      "grad_norm": 1.5293453931808472,
      "learning_rate": 1e-05,
      "loss": 0.2758,
      "step": 215200
    },
    {
      "epoch": 0.002153,
      "grad_norm": 1.556516408920288,
      "learning_rate": 1e-05,
      "loss": 0.2691,
      "step": 215300
    },
    {
      "epoch": 0.002154,
      "grad_norm": 1.6215890645980835,
      "learning_rate": 1e-05,
      "loss": 0.2792,
      "step": 215400
    },
    {
      "epoch": 0.002155,
      "grad_norm": 1.3890748023986816,
      "learning_rate": 1e-05,
      "loss": 0.2742,
      "step": 215500
    },
    {
      "epoch": 0.002156,
      "grad_norm": 1.41310453414917,
      "learning_rate": 1e-05,
      "loss": 0.2732,
      "step": 215600
    },
    {
      "epoch": 0.002157,
      "grad_norm": 1.4331954717636108,
      "learning_rate": 1e-05,
      "loss": 0.2775,
      "step": 215700
    },
    {
      "epoch": 0.002158,
      "grad_norm": 1.3733137845993042,
      "learning_rate": 1e-05,
      "loss": 0.2709,
      "step": 215800
    },
    {
      "epoch": 0.002159,
      "grad_norm": 1.2462104558944702,
      "learning_rate": 1e-05,
      "loss": 0.2761,
      "step": 215900
    },
    {
      "epoch": 0.00216,
      "grad_norm": 1.506518006324768,
      "learning_rate": 1e-05,
      "loss": 0.2728,
      "step": 216000
    },
    {
      "epoch": 0.002161,
      "grad_norm": 1.5095326900482178,
      "learning_rate": 1e-05,
      "loss": 0.2656,
      "step": 216100
    },
    {
      "epoch": 0.002162,
      "grad_norm": 1.406726360321045,
      "learning_rate": 1e-05,
      "loss": 0.2774,
      "step": 216200
    },
    {
      "epoch": 0.002163,
      "grad_norm": 1.5820876359939575,
      "learning_rate": 1e-05,
      "loss": 0.2735,
      "step": 216300
    },
    {
      "epoch": 0.002164,
      "grad_norm": 1.4047499895095825,
      "learning_rate": 1e-05,
      "loss": 0.2771,
      "step": 216400
    },
    {
      "epoch": 0.002165,
      "grad_norm": 1.3778250217437744,
      "learning_rate": 1e-05,
      "loss": 0.2708,
      "step": 216500
    },
    {
      "epoch": 0.002166,
      "grad_norm": 1.4947162866592407,
      "learning_rate": 1e-05,
      "loss": 0.2746,
      "step": 216600
    },
    {
      "epoch": 0.002167,
      "grad_norm": 1.6999680995941162,
      "learning_rate": 1e-05,
      "loss": 0.2738,
      "step": 216700
    },
    {
      "epoch": 0.002168,
      "grad_norm": 1.2014107704162598,
      "learning_rate": 1e-05,
      "loss": 0.2782,
      "step": 216800
    },
    {
      "epoch": 0.002169,
      "grad_norm": 1.5182017087936401,
      "learning_rate": 1e-05,
      "loss": 0.2693,
      "step": 216900
    },
    {
      "epoch": 0.00217,
      "grad_norm": 1.5457252264022827,
      "learning_rate": 1e-05,
      "loss": 0.2716,
      "step": 217000
    },
    {
      "epoch": 0.002171,
      "grad_norm": 1.2823336124420166,
      "learning_rate": 1e-05,
      "loss": 0.2699,
      "step": 217100
    },
    {
      "epoch": 0.002172,
      "grad_norm": 1.5415891408920288,
      "learning_rate": 1e-05,
      "loss": 0.2761,
      "step": 217200
    },
    {
      "epoch": 0.002173,
      "grad_norm": 1.4495307207107544,
      "learning_rate": 1e-05,
      "loss": 0.2759,
      "step": 217300
    },
    {
      "epoch": 0.002174,
      "grad_norm": 1.3674854040145874,
      "learning_rate": 1e-05,
      "loss": 0.273,
      "step": 217400
    },
    {
      "epoch": 0.002175,
      "grad_norm": 1.3650346994400024,
      "learning_rate": 1e-05,
      "loss": 0.2745,
      "step": 217500
    },
    {
      "epoch": 0.002176,
      "grad_norm": 1.835408329963684,
      "learning_rate": 1e-05,
      "loss": 0.2683,
      "step": 217600
    },
    {
      "epoch": 0.002177,
      "grad_norm": 1.332398533821106,
      "learning_rate": 1e-05,
      "loss": 0.273,
      "step": 217700
    },
    {
      "epoch": 0.002178,
      "grad_norm": 1.5285613536834717,
      "learning_rate": 1e-05,
      "loss": 0.2775,
      "step": 217800
    },
    {
      "epoch": 0.002179,
      "grad_norm": 1.9105749130249023,
      "learning_rate": 1e-05,
      "loss": 0.2749,
      "step": 217900
    },
    {
      "epoch": 0.00218,
      "grad_norm": 1.4697386026382446,
      "learning_rate": 1e-05,
      "loss": 0.2796,
      "step": 218000
    },
    {
      "epoch": 0.002181,
      "grad_norm": 1.2347923517227173,
      "learning_rate": 1e-05,
      "loss": 0.2715,
      "step": 218100
    },
    {
      "epoch": 0.002182,
      "grad_norm": 1.539542317390442,
      "learning_rate": 1e-05,
      "loss": 0.2731,
      "step": 218200
    },
    {
      "epoch": 0.002183,
      "grad_norm": 1.3916699886322021,
      "learning_rate": 1e-05,
      "loss": 0.2647,
      "step": 218300
    },
    {
      "epoch": 0.002184,
      "grad_norm": 1.3453309535980225,
      "learning_rate": 1e-05,
      "loss": 0.2732,
      "step": 218400
    },
    {
      "epoch": 0.002185,
      "grad_norm": 1.341639518737793,
      "learning_rate": 1e-05,
      "loss": 0.2738,
      "step": 218500
    },
    {
      "epoch": 0.002186,
      "grad_norm": 1.5608952045440674,
      "learning_rate": 1e-05,
      "loss": 0.2731,
      "step": 218600
    },
    {
      "epoch": 0.002187,
      "grad_norm": 1.3826178312301636,
      "learning_rate": 1e-05,
      "loss": 0.272,
      "step": 218700
    },
    {
      "epoch": 0.002188,
      "grad_norm": 1.3272230625152588,
      "learning_rate": 1e-05,
      "loss": 0.2663,
      "step": 218800
    },
    {
      "epoch": 0.002189,
      "grad_norm": 2.7572853565216064,
      "learning_rate": 1e-05,
      "loss": 0.2723,
      "step": 218900
    },
    {
      "epoch": 0.00219,
      "grad_norm": 1.3882721662521362,
      "learning_rate": 1e-05,
      "loss": 0.2683,
      "step": 219000
    },
    {
      "epoch": 0.002191,
      "grad_norm": 1.5614526271820068,
      "learning_rate": 1e-05,
      "loss": 0.2685,
      "step": 219100
    },
    {
      "epoch": 0.002192,
      "grad_norm": 1.5847523212432861,
      "learning_rate": 1e-05,
      "loss": 0.2656,
      "step": 219200
    },
    {
      "epoch": 0.002193,
      "grad_norm": 1.4185631275177002,
      "learning_rate": 1e-05,
      "loss": 0.2745,
      "step": 219300
    },
    {
      "epoch": 0.002194,
      "grad_norm": 1.476169228553772,
      "learning_rate": 1e-05,
      "loss": 0.2712,
      "step": 219400
    },
    {
      "epoch": 0.002195,
      "grad_norm": 1.512122631072998,
      "learning_rate": 1e-05,
      "loss": 0.2721,
      "step": 219500
    },
    {
      "epoch": 0.002196,
      "grad_norm": 1.4664489030838013,
      "learning_rate": 1e-05,
      "loss": 0.2706,
      "step": 219600
    },
    {
      "epoch": 0.002197,
      "grad_norm": 1.3304619789123535,
      "learning_rate": 1e-05,
      "loss": 0.272,
      "step": 219700
    },
    {
      "epoch": 0.002198,
      "grad_norm": 1.65354585647583,
      "learning_rate": 1e-05,
      "loss": 0.2716,
      "step": 219800
    },
    {
      "epoch": 0.002199,
      "grad_norm": 1.534138798713684,
      "learning_rate": 1e-05,
      "loss": 0.2689,
      "step": 219900
    },
    {
      "epoch": 0.0022,
      "grad_norm": 1.5334886312484741,
      "learning_rate": 1e-05,
      "loss": 0.2702,
      "step": 220000
    },
    {
      "epoch": 0.0022,
      "eval_loss": 0.2470703125,
      "eval_runtime": 109.4816,
      "eval_samples_per_second": 456.698,
      "eval_steps_per_second": 28.544,
      "step": 220000
    },
    {
      "epoch": 0.002201,
      "grad_norm": 1.6148320436477661,
      "learning_rate": 1e-05,
      "loss": 0.2714,
      "step": 220100
    },
    {
      "epoch": 0.002202,
      "grad_norm": 1.3717103004455566,
      "learning_rate": 1e-05,
      "loss": 0.2714,
      "step": 220200
    },
    {
      "epoch": 0.002203,
      "grad_norm": 1.325392484664917,
      "learning_rate": 1e-05,
      "loss": 0.2685,
      "step": 220300
    },
    {
      "epoch": 0.002204,
      "grad_norm": 1.1573683023452759,
      "learning_rate": 1e-05,
      "loss": 0.2777,
      "step": 220400
    },
    {
      "epoch": 0.002205,
      "grad_norm": 1.4676803350448608,
      "learning_rate": 1e-05,
      "loss": 0.2766,
      "step": 220500
    },
    {
      "epoch": 0.002206,
      "grad_norm": 1.3724902868270874,
      "learning_rate": 1e-05,
      "loss": 0.267,
      "step": 220600
    },
    {
      "epoch": 0.002207,
      "grad_norm": 1.3178821802139282,
      "learning_rate": 1e-05,
      "loss": 0.2677,
      "step": 220700
    },
    {
      "epoch": 0.002208,
      "grad_norm": 1.4839378595352173,
      "learning_rate": 1e-05,
      "loss": 0.2751,
      "step": 220800
    },
    {
      "epoch": 0.002209,
      "grad_norm": 1.4000365734100342,
      "learning_rate": 1e-05,
      "loss": 0.2715,
      "step": 220900
    },
    {
      "epoch": 0.00221,
      "grad_norm": 1.4803142547607422,
      "learning_rate": 1e-05,
      "loss": 0.2633,
      "step": 221000
    },
    {
      "epoch": 0.002211,
      "grad_norm": 1.8429442644119263,
      "learning_rate": 1e-05,
      "loss": 0.2721,
      "step": 221100
    },
    {
      "epoch": 0.002212,
      "grad_norm": 1.3296468257904053,
      "learning_rate": 1e-05,
      "loss": 0.2657,
      "step": 221200
    },
    {
      "epoch": 0.002213,
      "grad_norm": 1.283510446548462,
      "learning_rate": 1e-05,
      "loss": 0.2664,
      "step": 221300
    },
    {
      "epoch": 0.002214,
      "grad_norm": 1.7253391742706299,
      "learning_rate": 1e-05,
      "loss": 0.2719,
      "step": 221400
    },
    {
      "epoch": 0.002215,
      "grad_norm": 1.482222557067871,
      "learning_rate": 1e-05,
      "loss": 0.27,
      "step": 221500
    },
    {
      "epoch": 0.002216,
      "grad_norm": 1.497300148010254,
      "learning_rate": 1e-05,
      "loss": 0.2721,
      "step": 221600
    },
    {
      "epoch": 0.002217,
      "grad_norm": 1.3005281686782837,
      "learning_rate": 1e-05,
      "loss": 0.271,
      "step": 221700
    },
    {
      "epoch": 0.002218,
      "grad_norm": 1.3060221672058105,
      "learning_rate": 1e-05,
      "loss": 0.2728,
      "step": 221800
    },
    {
      "epoch": 0.002219,
      "grad_norm": 1.7013475894927979,
      "learning_rate": 1e-05,
      "loss": 0.2626,
      "step": 221900
    },
    {
      "epoch": 0.00222,
      "grad_norm": 1.3594257831573486,
      "learning_rate": 1e-05,
      "loss": 0.2687,
      "step": 222000
    },
    {
      "epoch": 0.002221,
      "grad_norm": 1.3790507316589355,
      "learning_rate": 1e-05,
      "loss": 0.2747,
      "step": 222100
    },
    {
      "epoch": 0.002222,
      "grad_norm": 1.2899534702301025,
      "learning_rate": 1e-05,
      "loss": 0.2636,
      "step": 222200
    },
    {
      "epoch": 0.002223,
      "grad_norm": 1.3310524225234985,
      "learning_rate": 1e-05,
      "loss": 0.2781,
      "step": 222300
    },
    {
      "epoch": 0.002224,
      "grad_norm": 1.5631316900253296,
      "learning_rate": 1e-05,
      "loss": 0.2668,
      "step": 222400
    },
    {
      "epoch": 0.002225,
      "grad_norm": 1.5219013690948486,
      "learning_rate": 1e-05,
      "loss": 0.2705,
      "step": 222500
    },
    {
      "epoch": 0.002226,
      "grad_norm": 1.2552835941314697,
      "learning_rate": 1e-05,
      "loss": 0.2717,
      "step": 222600
    },
    {
      "epoch": 0.002227,
      "grad_norm": 1.4404419660568237,
      "learning_rate": 1e-05,
      "loss": 0.2696,
      "step": 222700
    },
    {
      "epoch": 0.002228,
      "grad_norm": 1.3716353178024292,
      "learning_rate": 1e-05,
      "loss": 0.2623,
      "step": 222800
    },
    {
      "epoch": 0.002229,
      "grad_norm": 1.5366202592849731,
      "learning_rate": 1e-05,
      "loss": 0.2639,
      "step": 222900
    },
    {
      "epoch": 0.00223,
      "grad_norm": 1.366859793663025,
      "learning_rate": 1e-05,
      "loss": 0.2639,
      "step": 223000
    },
    {
      "epoch": 0.002231,
      "grad_norm": 1.3707561492919922,
      "learning_rate": 1e-05,
      "loss": 0.2705,
      "step": 223100
    },
    {
      "epoch": 0.002232,
      "grad_norm": 1.5594546794891357,
      "learning_rate": 1e-05,
      "loss": 0.2729,
      "step": 223200
    },
    {
      "epoch": 0.002233,
      "grad_norm": 1.404626488685608,
      "learning_rate": 1e-05,
      "loss": 0.2649,
      "step": 223300
    },
    {
      "epoch": 0.002234,
      "grad_norm": 1.6944884061813354,
      "learning_rate": 1e-05,
      "loss": 0.2686,
      "step": 223400
    },
    {
      "epoch": 0.002235,
      "grad_norm": 1.9831916093826294,
      "learning_rate": 1e-05,
      "loss": 0.274,
      "step": 223500
    },
    {
      "epoch": 0.002236,
      "grad_norm": 1.7819033861160278,
      "learning_rate": 1e-05,
      "loss": 0.2734,
      "step": 223600
    },
    {
      "epoch": 0.002237,
      "grad_norm": 1.5690064430236816,
      "learning_rate": 1e-05,
      "loss": 0.2737,
      "step": 223700
    },
    {
      "epoch": 0.002238,
      "grad_norm": 1.2317233085632324,
      "learning_rate": 1e-05,
      "loss": 0.2679,
      "step": 223800
    },
    {
      "epoch": 0.002239,
      "grad_norm": 2.0383543968200684,
      "learning_rate": 1e-05,
      "loss": 0.2679,
      "step": 223900
    },
    {
      "epoch": 0.00224,
      "grad_norm": 1.4997539520263672,
      "learning_rate": 1e-05,
      "loss": 0.2727,
      "step": 224000
    },
    {
      "epoch": 0.002241,
      "grad_norm": 1.3992339372634888,
      "learning_rate": 1e-05,
      "loss": 0.273,
      "step": 224100
    },
    {
      "epoch": 0.002242,
      "grad_norm": 1.5344079732894897,
      "learning_rate": 1e-05,
      "loss": 0.2712,
      "step": 224200
    },
    {
      "epoch": 0.002243,
      "grad_norm": 1.6358779668807983,
      "learning_rate": 1e-05,
      "loss": 0.2728,
      "step": 224300
    },
    {
      "epoch": 0.002244,
      "grad_norm": 1.6222172975540161,
      "learning_rate": 1e-05,
      "loss": 0.2621,
      "step": 224400
    },
    {
      "epoch": 0.002245,
      "grad_norm": 1.5196174383163452,
      "learning_rate": 1e-05,
      "loss": 0.2696,
      "step": 224500
    },
    {
      "epoch": 0.002246,
      "grad_norm": 1.477639079093933,
      "learning_rate": 1e-05,
      "loss": 0.2822,
      "step": 224600
    },
    {
      "epoch": 0.002247,
      "grad_norm": 1.6515791416168213,
      "learning_rate": 1e-05,
      "loss": 0.2716,
      "step": 224700
    },
    {
      "epoch": 0.002248,
      "grad_norm": 1.387439489364624,
      "learning_rate": 1e-05,
      "loss": 0.2674,
      "step": 224800
    },
    {
      "epoch": 0.002249,
      "grad_norm": 1.264535665512085,
      "learning_rate": 1e-05,
      "loss": 0.2708,
      "step": 224900
    },
    {
      "epoch": 0.00225,
      "grad_norm": 1.4407367706298828,
      "learning_rate": 1e-05,
      "loss": 0.2631,
      "step": 225000
    },
    {
      "epoch": 0.002251,
      "grad_norm": 1.5102108716964722,
      "learning_rate": 1e-05,
      "loss": 0.272,
      "step": 225100
    },
    {
      "epoch": 0.002252,
      "grad_norm": 1.5774545669555664,
      "learning_rate": 1e-05,
      "loss": 0.2675,
      "step": 225200
    },
    {
      "epoch": 0.002253,
      "grad_norm": 1.331253170967102,
      "learning_rate": 1e-05,
      "loss": 0.2638,
      "step": 225300
    },
    {
      "epoch": 0.002254,
      "grad_norm": 1.443367600440979,
      "learning_rate": 1e-05,
      "loss": 0.2756,
      "step": 225400
    },
    {
      "epoch": 0.002255,
      "grad_norm": 1.3950369358062744,
      "learning_rate": 1e-05,
      "loss": 0.2666,
      "step": 225500
    },
    {
      "epoch": 0.002256,
      "grad_norm": 2.172563314437866,
      "learning_rate": 1e-05,
      "loss": 0.2777,
      "step": 225600
    },
    {
      "epoch": 0.002257,
      "grad_norm": 1.1899489164352417,
      "learning_rate": 1e-05,
      "loss": 0.2651,
      "step": 225700
    },
    {
      "epoch": 0.002258,
      "grad_norm": 1.459839105606079,
      "learning_rate": 1e-05,
      "loss": 0.2691,
      "step": 225800
    },
    {
      "epoch": 0.002259,
      "grad_norm": 1.577430248260498,
      "learning_rate": 1e-05,
      "loss": 0.274,
      "step": 225900
    },
    {
      "epoch": 0.00226,
      "grad_norm": 2.454540729522705,
      "learning_rate": 1e-05,
      "loss": 0.2731,
      "step": 226000
    },
    {
      "epoch": 0.002261,
      "grad_norm": 1.4699666500091553,
      "learning_rate": 1e-05,
      "loss": 0.2684,
      "step": 226100
    },
    {
      "epoch": 0.002262,
      "grad_norm": 1.745713472366333,
      "learning_rate": 1e-05,
      "loss": 0.2742,
      "step": 226200
    },
    {
      "epoch": 0.002263,
      "grad_norm": 1.7182310819625854,
      "learning_rate": 1e-05,
      "loss": 0.2695,
      "step": 226300
    },
    {
      "epoch": 0.002264,
      "grad_norm": 1.3382000923156738,
      "learning_rate": 1e-05,
      "loss": 0.2638,
      "step": 226400
    },
    {
      "epoch": 0.002265,
      "grad_norm": 1.5027062892913818,
      "learning_rate": 1e-05,
      "loss": 0.2738,
      "step": 226500
    },
    {
      "epoch": 0.002266,
      "grad_norm": 1.4339383840560913,
      "learning_rate": 1e-05,
      "loss": 0.2651,
      "step": 226600
    },
    {
      "epoch": 0.002267,
      "grad_norm": 1.5018837451934814,
      "learning_rate": 1e-05,
      "loss": 0.2681,
      "step": 226700
    },
    {
      "epoch": 0.002268,
      "grad_norm": 1.4683713912963867,
      "learning_rate": 1e-05,
      "loss": 0.267,
      "step": 226800
    },
    {
      "epoch": 0.002269,
      "grad_norm": 1.5575743913650513,
      "learning_rate": 1e-05,
      "loss": 0.2668,
      "step": 226900
    },
    {
      "epoch": 0.00227,
      "grad_norm": 1.3830864429473877,
      "learning_rate": 1e-05,
      "loss": 0.2701,
      "step": 227000
    },
    {
      "epoch": 0.002271,
      "grad_norm": 1.3803704977035522,
      "learning_rate": 1e-05,
      "loss": 0.266,
      "step": 227100
    },
    {
      "epoch": 0.002272,
      "grad_norm": 1.6529074907302856,
      "learning_rate": 1e-05,
      "loss": 0.2724,
      "step": 227200
    },
    {
      "epoch": 0.002273,
      "grad_norm": 1.5334415435791016,
      "learning_rate": 1e-05,
      "loss": 0.272,
      "step": 227300
    },
    {
      "epoch": 0.002274,
      "grad_norm": 1.4912134408950806,
      "learning_rate": 1e-05,
      "loss": 0.2723,
      "step": 227400
    },
    {
      "epoch": 0.002275,
      "grad_norm": 1.375284194946289,
      "learning_rate": 1e-05,
      "loss": 0.2702,
      "step": 227500
    },
    {
      "epoch": 0.002276,
      "grad_norm": 1.2836390733718872,
      "learning_rate": 1e-05,
      "loss": 0.2738,
      "step": 227600
    },
    {
      "epoch": 0.002277,
      "grad_norm": 1.2948681116104126,
      "learning_rate": 1e-05,
      "loss": 0.275,
      "step": 227700
    },
    {
      "epoch": 0.002278,
      "grad_norm": 1.3349876403808594,
      "learning_rate": 1e-05,
      "loss": 0.2673,
      "step": 227800
    },
    {
      "epoch": 0.002279,
      "grad_norm": 1.422441840171814,
      "learning_rate": 1e-05,
      "loss": 0.267,
      "step": 227900
    },
    {
      "epoch": 0.00228,
      "grad_norm": 1.4071296453475952,
      "learning_rate": 1e-05,
      "loss": 0.2688,
      "step": 228000
    },
    {
      "epoch": 0.002281,
      "grad_norm": 1.3052148818969727,
      "learning_rate": 1e-05,
      "loss": 0.2674,
      "step": 228100
    },
    {
      "epoch": 0.002282,
      "grad_norm": 1.4938229322433472,
      "learning_rate": 1e-05,
      "loss": 0.2693,
      "step": 228200
    },
    {
      "epoch": 0.002283,
      "grad_norm": 1.48699951171875,
      "learning_rate": 1e-05,
      "loss": 0.2681,
      "step": 228300
    },
    {
      "epoch": 0.002284,
      "grad_norm": 1.5458961725234985,
      "learning_rate": 1e-05,
      "loss": 0.2697,
      "step": 228400
    },
    {
      "epoch": 0.002285,
      "grad_norm": 1.394895315170288,
      "learning_rate": 1e-05,
      "loss": 0.2678,
      "step": 228500
    },
    {
      "epoch": 0.002286,
      "grad_norm": 1.3542561531066895,
      "learning_rate": 1e-05,
      "loss": 0.2669,
      "step": 228600
    },
    {
      "epoch": 0.002287,
      "grad_norm": 1.5436086654663086,
      "learning_rate": 1e-05,
      "loss": 0.2681,
      "step": 228700
    },
    {
      "epoch": 0.002288,
      "grad_norm": 1.3657710552215576,
      "learning_rate": 1e-05,
      "loss": 0.2731,
      "step": 228800
    },
    {
      "epoch": 0.002289,
      "grad_norm": 1.2138234376907349,
      "learning_rate": 1e-05,
      "loss": 0.2706,
      "step": 228900
    },
    {
      "epoch": 0.00229,
      "grad_norm": 1.4697974920272827,
      "learning_rate": 1e-05,
      "loss": 0.2695,
      "step": 229000
    },
    {
      "epoch": 0.002291,
      "grad_norm": 1.532043695449829,
      "learning_rate": 1e-05,
      "loss": 0.2691,
      "step": 229100
    },
    {
      "epoch": 0.002292,
      "grad_norm": 1.5469400882720947,
      "learning_rate": 1e-05,
      "loss": 0.2635,
      "step": 229200
    },
    {
      "epoch": 0.002293,
      "grad_norm": 1.3268221616744995,
      "learning_rate": 1e-05,
      "loss": 0.2703,
      "step": 229300
    },
    {
      "epoch": 0.002294,
      "grad_norm": 1.3366742134094238,
      "learning_rate": 1e-05,
      "loss": 0.2658,
      "step": 229400
    },
    {
      "epoch": 0.002295,
      "grad_norm": 1.3616935014724731,
      "learning_rate": 1e-05,
      "loss": 0.2728,
      "step": 229500
    },
    {
      "epoch": 0.002296,
      "grad_norm": 1.3293488025665283,
      "learning_rate": 1e-05,
      "loss": 0.2703,
      "step": 229600
    },
    {
      "epoch": 0.002297,
      "grad_norm": 1.58241605758667,
      "learning_rate": 1e-05,
      "loss": 0.2629,
      "step": 229700
    },
    {
      "epoch": 0.002298,
      "grad_norm": 1.240764856338501,
      "learning_rate": 1e-05,
      "loss": 0.2664,
      "step": 229800
    },
    {
      "epoch": 0.002299,
      "grad_norm": 1.480238676071167,
      "learning_rate": 1e-05,
      "loss": 0.2737,
      "step": 229900
    },
    {
      "epoch": 0.0023,
      "grad_norm": 1.6399883031845093,
      "learning_rate": 1e-05,
      "loss": 0.265,
      "step": 230000
    },
    {
      "epoch": 0.002301,
      "grad_norm": 1.5711274147033691,
      "learning_rate": 1e-05,
      "loss": 0.265,
      "step": 230100
    },
    {
      "epoch": 0.002302,
      "grad_norm": 1.3830393552780151,
      "learning_rate": 1e-05,
      "loss": 0.2708,
      "step": 230200
    },
    {
      "epoch": 0.002303,
      "grad_norm": 1.3761115074157715,
      "learning_rate": 1e-05,
      "loss": 0.2665,
      "step": 230300
    },
    {
      "epoch": 0.002304,
      "grad_norm": 1.4468395709991455,
      "learning_rate": 1e-05,
      "loss": 0.2665,
      "step": 230400
    },
    {
      "epoch": 0.002305,
      "grad_norm": 1.2549031972885132,
      "learning_rate": 1e-05,
      "loss": 0.2704,
      "step": 230500
    },
    {
      "epoch": 0.002306,
      "grad_norm": 1.5276696681976318,
      "learning_rate": 1e-05,
      "loss": 0.2617,
      "step": 230600
    },
    {
      "epoch": 0.002307,
      "grad_norm": 1.222853660583496,
      "learning_rate": 1e-05,
      "loss": 0.2692,
      "step": 230700
    },
    {
      "epoch": 0.002308,
      "grad_norm": 1.504705548286438,
      "learning_rate": 1e-05,
      "loss": 0.2743,
      "step": 230800
    },
    {
      "epoch": 0.002309,
      "grad_norm": 1.380125880241394,
      "learning_rate": 1e-05,
      "loss": 0.273,
      "step": 230900
    },
    {
      "epoch": 0.00231,
      "grad_norm": 1.5362893342971802,
      "learning_rate": 1e-05,
      "loss": 0.2617,
      "step": 231000
    },
    {
      "epoch": 0.002311,
      "grad_norm": 1.308255672454834,
      "learning_rate": 1e-05,
      "loss": 0.2607,
      "step": 231100
    },
    {
      "epoch": 0.002312,
      "grad_norm": 1.46437406539917,
      "learning_rate": 1e-05,
      "loss": 0.2623,
      "step": 231200
    },
    {
      "epoch": 0.002313,
      "grad_norm": 1.7176834344863892,
      "learning_rate": 1e-05,
      "loss": 0.2733,
      "step": 231300
    },
    {
      "epoch": 0.002314,
      "grad_norm": 1.2125487327575684,
      "learning_rate": 1e-05,
      "loss": 0.2661,
      "step": 231400
    },
    {
      "epoch": 0.002315,
      "grad_norm": 1.656404972076416,
      "learning_rate": 1e-05,
      "loss": 0.2719,
      "step": 231500
    },
    {
      "epoch": 0.002316,
      "grad_norm": 1.2319536209106445,
      "learning_rate": 1e-05,
      "loss": 0.2662,
      "step": 231600
    },
    {
      "epoch": 0.002317,
      "grad_norm": 1.4529757499694824,
      "learning_rate": 1e-05,
      "loss": 0.2668,
      "step": 231700
    },
    {
      "epoch": 0.002318,
      "grad_norm": 1.3441050052642822,
      "learning_rate": 1e-05,
      "loss": 0.2664,
      "step": 231800
    },
    {
      "epoch": 0.002319,
      "grad_norm": 1.3161287307739258,
      "learning_rate": 1e-05,
      "loss": 0.2664,
      "step": 231900
    },
    {
      "epoch": 0.00232,
      "grad_norm": 1.5237764120101929,
      "learning_rate": 1e-05,
      "loss": 0.2661,
      "step": 232000
    },
    {
      "epoch": 0.002321,
      "grad_norm": 1.4569116830825806,
      "learning_rate": 1e-05,
      "loss": 0.2683,
      "step": 232100
    },
    {
      "epoch": 0.002322,
      "grad_norm": 1.4047276973724365,
      "learning_rate": 1e-05,
      "loss": 0.2658,
      "step": 232200
    },
    {
      "epoch": 0.002323,
      "grad_norm": 1.6638625860214233,
      "learning_rate": 1e-05,
      "loss": 0.2632,
      "step": 232300
    },
    {
      "epoch": 0.002324,
      "grad_norm": 1.2283929586410522,
      "learning_rate": 1e-05,
      "loss": 0.2639,
      "step": 232400
    },
    {
      "epoch": 0.002325,
      "grad_norm": 1.2033153772354126,
      "learning_rate": 1e-05,
      "loss": 0.2679,
      "step": 232500
    },
    {
      "epoch": 0.002326,
      "grad_norm": 1.37595534324646,
      "learning_rate": 1e-05,
      "loss": 0.2671,
      "step": 232600
    },
    {
      "epoch": 0.002327,
      "grad_norm": 1.4282982349395752,
      "learning_rate": 1e-05,
      "loss": 0.2645,
      "step": 232700
    },
    {
      "epoch": 0.002328,
      "grad_norm": 1.3374396562576294,
      "learning_rate": 1e-05,
      "loss": 0.2702,
      "step": 232800
    },
    {
      "epoch": 0.002329,
      "grad_norm": 1.2464615106582642,
      "learning_rate": 1e-05,
      "loss": 0.2704,
      "step": 232900
    },
    {
      "epoch": 0.00233,
      "grad_norm": 1.4305086135864258,
      "learning_rate": 1e-05,
      "loss": 0.2675,
      "step": 233000
    },
    {
      "epoch": 0.002331,
      "grad_norm": 1.404268503189087,
      "learning_rate": 1e-05,
      "loss": 0.2673,
      "step": 233100
    },
    {
      "epoch": 0.002332,
      "grad_norm": 1.2417124509811401,
      "learning_rate": 1e-05,
      "loss": 0.2652,
      "step": 233200
    },
    {
      "epoch": 0.002333,
      "grad_norm": 1.4305444955825806,
      "learning_rate": 1e-05,
      "loss": 0.264,
      "step": 233300
    },
    {
      "epoch": 0.002334,
      "grad_norm": 1.4114798307418823,
      "learning_rate": 1e-05,
      "loss": 0.2721,
      "step": 233400
    },
    {
      "epoch": 0.002335,
      "grad_norm": 1.5132521390914917,
      "learning_rate": 1e-05,
      "loss": 0.2662,
      "step": 233500
    },
    {
      "epoch": 0.002336,
      "grad_norm": 1.2459546327590942,
      "learning_rate": 1e-05,
      "loss": 0.2608,
      "step": 233600
    },
    {
      "epoch": 0.002337,
      "grad_norm": 1.3642733097076416,
      "learning_rate": 1e-05,
      "loss": 0.2673,
      "step": 233700
    },
    {
      "epoch": 0.002338,
      "grad_norm": 1.2351726293563843,
      "learning_rate": 1e-05,
      "loss": 0.2655,
      "step": 233800
    },
    {
      "epoch": 0.002339,
      "grad_norm": 1.2862097024917603,
      "learning_rate": 1e-05,
      "loss": 0.2726,
      "step": 233900
    },
    {
      "epoch": 0.00234,
      "grad_norm": 1.286604642868042,
      "learning_rate": 1e-05,
      "loss": 0.2626,
      "step": 234000
    },
    {
      "epoch": 0.002341,
      "grad_norm": 1.5336260795593262,
      "learning_rate": 1e-05,
      "loss": 0.2654,
      "step": 234100
    },
    {
      "epoch": 0.002342,
      "grad_norm": 1.5840083360671997,
      "learning_rate": 1e-05,
      "loss": 0.2664,
      "step": 234200
    },
    {
      "epoch": 0.002343,
      "grad_norm": 1.7138983011245728,
      "learning_rate": 1e-05,
      "loss": 0.2649,
      "step": 234300
    },
    {
      "epoch": 0.002344,
      "grad_norm": 1.624403953552246,
      "learning_rate": 1e-05,
      "loss": 0.2604,
      "step": 234400
    },
    {
      "epoch": 0.002345,
      "grad_norm": 1.43915593624115,
      "learning_rate": 1e-05,
      "loss": 0.2656,
      "step": 234500
    },
    {
      "epoch": 0.002346,
      "grad_norm": 1.4849474430084229,
      "learning_rate": 1e-05,
      "loss": 0.2651,
      "step": 234600
    },
    {
      "epoch": 0.002347,
      "grad_norm": 1.5203824043273926,
      "learning_rate": 1e-05,
      "loss": 0.2628,
      "step": 234700
    },
    {
      "epoch": 0.002348,
      "grad_norm": 1.2643849849700928,
      "learning_rate": 1e-05,
      "loss": 0.2661,
      "step": 234800
    },
    {
      "epoch": 0.002349,
      "grad_norm": 1.2797199487686157,
      "learning_rate": 1e-05,
      "loss": 0.2624,
      "step": 234900
    },
    {
      "epoch": 0.00235,
      "grad_norm": 1.4327815771102905,
      "learning_rate": 1e-05,
      "loss": 0.2635,
      "step": 235000
    },
    {
      "epoch": 0.002351,
      "grad_norm": 1.5249953269958496,
      "learning_rate": 1e-05,
      "loss": 0.265,
      "step": 235100
    },
    {
      "epoch": 0.002352,
      "grad_norm": 1.46501624584198,
      "learning_rate": 1e-05,
      "loss": 0.2676,
      "step": 235200
    },
    {
      "epoch": 0.002353,
      "grad_norm": 1.5722047090530396,
      "learning_rate": 1e-05,
      "loss": 0.2574,
      "step": 235300
    },
    {
      "epoch": 0.002354,
      "grad_norm": 1.2025928497314453,
      "learning_rate": 1e-05,
      "loss": 0.2671,
      "step": 235400
    },
    {
      "epoch": 0.002355,
      "grad_norm": 1.554432988166809,
      "learning_rate": 1e-05,
      "loss": 0.2703,
      "step": 235500
    },
    {
      "epoch": 0.002356,
      "grad_norm": 1.4228944778442383,
      "learning_rate": 1e-05,
      "loss": 0.2728,
      "step": 235600
    },
    {
      "epoch": 0.002357,
      "grad_norm": 1.5761821269989014,
      "learning_rate": 1e-05,
      "loss": 0.2637,
      "step": 235700
    },
    {
      "epoch": 0.002358,
      "grad_norm": 1.1505320072174072,
      "learning_rate": 1e-05,
      "loss": 0.2675,
      "step": 235800
    },
    {
      "epoch": 0.002359,
      "grad_norm": 1.4779144525527954,
      "learning_rate": 1e-05,
      "loss": 0.2716,
      "step": 235900
    },
    {
      "epoch": 0.00236,
      "grad_norm": 1.3939759731292725,
      "learning_rate": 1e-05,
      "loss": 0.2616,
      "step": 236000
    },
    {
      "epoch": 0.002361,
      "grad_norm": 1.5327190160751343,
      "learning_rate": 1e-05,
      "loss": 0.263,
      "step": 236100
    },
    {
      "epoch": 0.002362,
      "grad_norm": 1.338335394859314,
      "learning_rate": 1e-05,
      "loss": 0.2696,
      "step": 236200
    },
    {
      "epoch": 0.002363,
      "grad_norm": 1.4766395092010498,
      "learning_rate": 1e-05,
      "loss": 0.2657,
      "step": 236300
    },
    {
      "epoch": 0.002364,
      "grad_norm": 1.292716145515442,
      "learning_rate": 1e-05,
      "loss": 0.2611,
      "step": 236400
    },
    {
      "epoch": 0.002365,
      "grad_norm": 1.3678349256515503,
      "learning_rate": 1e-05,
      "loss": 0.2663,
      "step": 236500
    },
    {
      "epoch": 0.002366,
      "grad_norm": 1.5033866167068481,
      "learning_rate": 1e-05,
      "loss": 0.2693,
      "step": 236600
    },
    {
      "epoch": 0.002367,
      "grad_norm": 1.1508349180221558,
      "learning_rate": 1e-05,
      "loss": 0.2631,
      "step": 236700
    },
    {
      "epoch": 0.002368,
      "grad_norm": 1.627375602722168,
      "learning_rate": 1e-05,
      "loss": 0.2661,
      "step": 236800
    },
    {
      "epoch": 0.002369,
      "grad_norm": 1.5816211700439453,
      "learning_rate": 1e-05,
      "loss": 0.2655,
      "step": 236900
    },
    {
      "epoch": 0.00237,
      "grad_norm": 1.6481306552886963,
      "learning_rate": 1e-05,
      "loss": 0.2649,
      "step": 237000
    },
    {
      "epoch": 0.002371,
      "grad_norm": 1.6917251348495483,
      "learning_rate": 1e-05,
      "loss": 0.2653,
      "step": 237100
    },
    {
      "epoch": 0.002372,
      "grad_norm": 2.067993640899658,
      "learning_rate": 1e-05,
      "loss": 0.2593,
      "step": 237200
    },
    {
      "epoch": 0.002373,
      "grad_norm": 1.3310753107070923,
      "learning_rate": 1e-05,
      "loss": 0.2667,
      "step": 237300
    },
    {
      "epoch": 0.002374,
      "grad_norm": 1.6005228757858276,
      "learning_rate": 1e-05,
      "loss": 0.2658,
      "step": 237400
    },
    {
      "epoch": 0.002375,
      "grad_norm": 1.285658836364746,
      "learning_rate": 1e-05,
      "loss": 0.2605,
      "step": 237500
    },
    {
      "epoch": 0.002376,
      "grad_norm": 1.37336003780365,
      "learning_rate": 1e-05,
      "loss": 0.2691,
      "step": 237600
    },
    {
      "epoch": 0.002377,
      "grad_norm": 1.3081128597259521,
      "learning_rate": 1e-05,
      "loss": 0.2597,
      "step": 237700
    },
    {
      "epoch": 0.002378,
      "grad_norm": 1.4721745252609253,
      "learning_rate": 1e-05,
      "loss": 0.2679,
      "step": 237800
    },
    {
      "epoch": 0.002379,
      "grad_norm": 1.4306268692016602,
      "learning_rate": 1e-05,
      "loss": 0.267,
      "step": 237900
    },
    {
      "epoch": 0.00238,
      "grad_norm": 1.323388934135437,
      "learning_rate": 1e-05,
      "loss": 0.2656,
      "step": 238000
    },
    {
      "epoch": 0.002381,
      "grad_norm": 1.5063282251358032,
      "learning_rate": 1e-05,
      "loss": 0.2647,
      "step": 238100
    },
    {
      "epoch": 0.002382,
      "grad_norm": 1.50105619430542,
      "learning_rate": 1e-05,
      "loss": 0.2627,
      "step": 238200
    },
    {
      "epoch": 0.002383,
      "grad_norm": 1.325971245765686,
      "learning_rate": 1e-05,
      "loss": 0.2691,
      "step": 238300
    },
    {
      "epoch": 0.002384,
      "grad_norm": 1.3668450117111206,
      "learning_rate": 1e-05,
      "loss": 0.2607,
      "step": 238400
    },
    {
      "epoch": 0.002385,
      "grad_norm": 1.299739122390747,
      "learning_rate": 1e-05,
      "loss": 0.2658,
      "step": 238500
    },
    {
      "epoch": 0.002386,
      "grad_norm": 1.3577795028686523,
      "learning_rate": 1e-05,
      "loss": 0.267,
      "step": 238600
    },
    {
      "epoch": 0.002387,
      "grad_norm": 2.8680946826934814,
      "learning_rate": 1e-05,
      "loss": 0.2659,
      "step": 238700
    },
    {
      "epoch": 0.002388,
      "grad_norm": 1.344599962234497,
      "learning_rate": 1e-05,
      "loss": 0.2677,
      "step": 238800
    },
    {
      "epoch": 0.002389,
      "grad_norm": 1.3766945600509644,
      "learning_rate": 1e-05,
      "loss": 0.269,
      "step": 238900
    },
    {
      "epoch": 0.00239,
      "grad_norm": 1.4144660234451294,
      "learning_rate": 1e-05,
      "loss": 0.2602,
      "step": 239000
    },
    {
      "epoch": 0.002391,
      "grad_norm": 1.5478261709213257,
      "learning_rate": 1e-05,
      "loss": 0.2665,
      "step": 239100
    },
    {
      "epoch": 0.002392,
      "grad_norm": 2.6023824214935303,
      "learning_rate": 1e-05,
      "loss": 0.267,
      "step": 239200
    },
    {
      "epoch": 0.002393,
      "grad_norm": 1.681246042251587,
      "learning_rate": 1e-05,
      "loss": 0.2617,
      "step": 239300
    },
    {
      "epoch": 0.002394,
      "grad_norm": 1.3988147974014282,
      "learning_rate": 1e-05,
      "loss": 0.2642,
      "step": 239400
    },
    {
      "epoch": 0.002395,
      "grad_norm": 1.481540560722351,
      "learning_rate": 1e-05,
      "loss": 0.2671,
      "step": 239500
    },
    {
      "epoch": 0.002396,
      "grad_norm": 1.4419687986373901,
      "learning_rate": 1e-05,
      "loss": 0.2587,
      "step": 239600
    },
    {
      "epoch": 0.002397,
      "grad_norm": 1.3948259353637695,
      "learning_rate": 1e-05,
      "loss": 0.2675,
      "step": 239700
    },
    {
      "epoch": 0.002398,
      "grad_norm": 1.5281838178634644,
      "learning_rate": 1e-05,
      "loss": 0.2645,
      "step": 239800
    },
    {
      "epoch": 0.002399,
      "grad_norm": 1.4255881309509277,
      "learning_rate": 1e-05,
      "loss": 0.2655,
      "step": 239900
    },
    {
      "epoch": 0.0024,
      "grad_norm": 1.332002878189087,
      "learning_rate": 1e-05,
      "loss": 0.2671,
      "step": 240000
    },
    {
      "epoch": 0.0024,
      "eval_loss": 0.239013671875,
      "eval_runtime": 114.9087,
      "eval_samples_per_second": 435.128,
      "eval_steps_per_second": 27.195,
      "step": 240000
    }
  ],
  "logging_steps": 100,
  "max_steps": 100000000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 20000,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 200,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 5.0168070144e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}