{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.8505965498898093,
  "eval_steps": 500,
  "global_step": 11193,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0007599361653621095,
      "grad_norm": 191.63430175303478,
      "learning_rate": 1.1061797752808989e-06,
      "loss": 2.883259582519531,
      "step": 10
    },
    {
      "epoch": 0.001519872330724219,
      "grad_norm": 124.58368761068755,
      "learning_rate": 2.2241573033707866e-06,
      "loss": 1.8781267166137696,
      "step": 20
    },
    {
      "epoch": 0.0022798084960863286,
      "grad_norm": 87.1340034334807,
      "learning_rate": 3.342134831460675e-06,
      "loss": 1.1774589538574218,
      "step": 30
    },
    {
      "epoch": 0.003039744661448438,
      "grad_norm": 844.0347817685458,
      "learning_rate": 4.4601123595505624e-06,
      "loss": 0.9150102615356446,
      "step": 40
    },
    {
      "epoch": 0.0037996808268105477,
      "grad_norm": 26.071347281400854,
      "learning_rate": 5.5780898876404495e-06,
      "loss": 0.7994430065155029,
      "step": 50
    },
    {
      "epoch": 0.004559616992172657,
      "grad_norm": 106.14143580914507,
      "learning_rate": 6.6960674157303374e-06,
      "loss": 0.7530351161956788,
      "step": 60
    },
    {
      "epoch": 0.005319553157534767,
      "grad_norm": 150.7760094384226,
      "learning_rate": 7.814044943820226e-06,
      "loss": 0.7612151622772216,
      "step": 70
    },
    {
      "epoch": 0.006079489322896876,
      "grad_norm": 17.649974068212234,
      "learning_rate": 8.932022471910112e-06,
      "loss": 0.7631851196289062,
      "step": 80
    },
    {
      "epoch": 0.006839425488258986,
      "grad_norm": 74.83280873477604,
      "learning_rate": 1.005e-05,
      "loss": 0.7748248100280761,
      "step": 90
    },
    {
      "epoch": 0.007599361653621095,
      "grad_norm": 923.6270136131341,
      "learning_rate": 1.1167977528089888e-05,
      "loss": 0.7622076988220214,
      "step": 100
    },
    {
      "epoch": 0.008359297818983206,
      "grad_norm": 191.99807923455484,
      "learning_rate": 1.2285955056179776e-05,
      "loss": 0.7419852256774903,
      "step": 110
    },
    {
      "epoch": 0.009119233984345314,
      "grad_norm": 8.180188792042815,
      "learning_rate": 1.3403932584269664e-05,
      "loss": 0.7526190757751465,
      "step": 120
    },
    {
      "epoch": 0.009879170149707425,
      "grad_norm": 6.875791274720428,
      "learning_rate": 1.4521910112359552e-05,
      "loss": 0.7423999309539795,
      "step": 130
    },
    {
      "epoch": 0.010639106315069534,
      "grad_norm": 4.139161233642084,
      "learning_rate": 1.563988764044944e-05,
      "loss": 0.736327075958252,
      "step": 140
    },
    {
      "epoch": 0.011399042480431644,
      "grad_norm": 22.480724459482204,
      "learning_rate": 1.6757865168539328e-05,
      "loss": 0.7192594528198242,
      "step": 150
    },
    {
      "epoch": 0.012158978645793753,
      "grad_norm": 42.20418057032992,
      "learning_rate": 1.7875842696629214e-05,
      "loss": 0.7199824333190918,
      "step": 160
    },
    {
      "epoch": 0.012918914811155863,
      "grad_norm": 22.176716309419565,
      "learning_rate": 1.89938202247191e-05,
      "loss": 0.7676156520843506,
      "step": 170
    },
    {
      "epoch": 0.013678850976517972,
      "grad_norm": 39.06503779613347,
      "learning_rate": 2e-05,
      "loss": 0.723706579208374,
      "step": 180
    },
    {
      "epoch": 0.014438787141880082,
      "grad_norm": 72.77156957569211,
      "learning_rate": 2e-05,
      "loss": 0.7594737052917481,
      "step": 190
    },
    {
      "epoch": 0.01519872330724219,
      "grad_norm": 121.81035558320448,
      "learning_rate": 2e-05,
      "loss": 0.7727550506591797,
      "step": 200
    },
    {
      "epoch": 0.015958659472604303,
      "grad_norm": 17.261218474875754,
      "learning_rate": 2e-05,
      "loss": 0.7319492340087891,
      "step": 210
    },
    {
      "epoch": 0.01671859563796641,
      "grad_norm": 13.6677607141103,
      "learning_rate": 2e-05,
      "loss": 0.7261901378631592,
      "step": 220
    },
    {
      "epoch": 0.01747853180332852,
      "grad_norm": 19.87687501769399,
      "learning_rate": 2e-05,
      "loss": 0.7509442806243897,
      "step": 230
    },
    {
      "epoch": 0.01823846796869063,
      "grad_norm": 71.02797556619042,
      "learning_rate": 2e-05,
      "loss": 0.7291659355163574,
      "step": 240
    },
    {
      "epoch": 0.01899840413405274,
      "grad_norm": 7.361788981118384,
      "learning_rate": 2e-05,
      "loss": 0.7559044837951661,
      "step": 250
    },
    {
      "epoch": 0.01975834029941485,
      "grad_norm": 13.834043992751518,
      "learning_rate": 2e-05,
      "loss": 0.7548636436462403,
      "step": 260
    },
    {
      "epoch": 0.02051827646477696,
      "grad_norm": 18.40212932805268,
      "learning_rate": 2e-05,
      "loss": 0.7729722499847412,
      "step": 270
    },
    {
      "epoch": 0.021278212630139067,
      "grad_norm": 3.7670360731030317,
      "learning_rate": 2e-05,
      "loss": 0.7427117824554443,
      "step": 280
    },
    {
      "epoch": 0.02203814879550118,
      "grad_norm": 3.2551327080904002,
      "learning_rate": 2e-05,
      "loss": 0.696455717086792,
      "step": 290
    },
    {
      "epoch": 0.022798084960863288,
      "grad_norm": 12.681218354218718,
      "learning_rate": 2e-05,
      "loss": 0.7120409965515136,
      "step": 300
    },
    {
      "epoch": 0.023558021126225397,
      "grad_norm": 54.34344654534163,
      "learning_rate": 2e-05,
      "loss": 0.6876229286193848,
      "step": 310
    },
    {
      "epoch": 0.024317957291587505,
      "grad_norm": 5.6674147357259415,
      "learning_rate": 2e-05,
      "loss": 0.7949447631835938,
      "step": 320
    },
    {
      "epoch": 0.025077893456949617,
      "grad_norm": 4.542555847806229,
      "learning_rate": 2e-05,
      "loss": 0.7002223014831543,
      "step": 330
    },
    {
      "epoch": 0.025837829622311726,
      "grad_norm": 6.420887665016644,
      "learning_rate": 2e-05,
      "loss": 0.74368314743042,
      "step": 340
    },
    {
      "epoch": 0.026597765787673835,
      "grad_norm": 66.78918732414809,
      "learning_rate": 2e-05,
      "loss": 0.7450421810150146,
      "step": 350
    },
    {
      "epoch": 0.027357701953035943,
      "grad_norm": 8.341774953333017,
      "learning_rate": 2e-05,
      "loss": 0.7264707565307618,
      "step": 360
    },
    {
      "epoch": 0.028117638118398056,
      "grad_norm": 2.4168868991630457,
      "learning_rate": 2e-05,
      "loss": 0.7024238586425782,
      "step": 370
    },
    {
      "epoch": 0.028877574283760164,
      "grad_norm": 4.431590851345387,
      "learning_rate": 2e-05,
      "loss": 0.7049625396728516,
      "step": 380
    },
    {
      "epoch": 0.029637510449122273,
      "grad_norm": 36.43429145728405,
      "learning_rate": 2e-05,
      "loss": 0.7121419429779052,
      "step": 390
    },
    {
      "epoch": 0.03039744661448438,
      "grad_norm": 8.501774499445727,
      "learning_rate": 2e-05,
      "loss": 0.7381654262542725,
      "step": 400
    },
    {
      "epoch": 0.031157382779846494,
      "grad_norm": 2.1284121866050194,
      "learning_rate": 2e-05,
      "loss": 0.7352887153625488,
      "step": 410
    },
    {
      "epoch": 0.031917318945208606,
      "grad_norm": 1.6959741724878892,
      "learning_rate": 2e-05,
      "loss": 0.709722900390625,
      "step": 420
    },
    {
      "epoch": 0.032677255110570715,
      "grad_norm": 2.777848797318434,
      "learning_rate": 2e-05,
      "loss": 0.6909010887145997,
      "step": 430
    },
    {
      "epoch": 0.03343719127593282,
      "grad_norm": 5.759466391991063,
      "learning_rate": 2e-05,
      "loss": 0.6841095924377442,
      "step": 440
    },
    {
      "epoch": 0.03419712744129493,
      "grad_norm": 1.9894311227745336,
      "learning_rate": 2e-05,
      "loss": 0.6729912757873535,
      "step": 450
    },
    {
      "epoch": 0.03495706360665704,
      "grad_norm": 2.313482340982159,
      "learning_rate": 2e-05,
      "loss": 0.6638118743896484,
      "step": 460
    },
    {
      "epoch": 0.03571699977201915,
      "grad_norm": 7.013232779450187,
      "learning_rate": 2e-05,
      "loss": 0.6692584037780762,
      "step": 470
    },
    {
      "epoch": 0.03647693593738126,
      "grad_norm": 2.574330643991075,
      "learning_rate": 2e-05,
      "loss": 0.70731201171875,
      "step": 480
    },
    {
      "epoch": 0.03723687210274337,
      "grad_norm": 1.5052842559853645,
      "learning_rate": 2e-05,
      "loss": 0.6873155117034913,
      "step": 490
    },
    {
      "epoch": 0.03799680826810548,
      "grad_norm": 1.116178924419754,
      "learning_rate": 2e-05,
      "loss": 0.669003677368164,
      "step": 500
    },
    {
      "epoch": 0.03875674443346759,
      "grad_norm": 1.938914338338345,
      "learning_rate": 2e-05,
      "loss": 0.69815092086792,
      "step": 510
    },
    {
      "epoch": 0.0395166805988297,
      "grad_norm": 1.3811669929651738,
      "learning_rate": 2e-05,
      "loss": 0.712796401977539,
      "step": 520
    },
    {
      "epoch": 0.04027661676419181,
      "grad_norm": 1.091199227854418,
      "learning_rate": 2e-05,
      "loss": 0.6600155830383301,
      "step": 530
    },
    {
      "epoch": 0.04103655292955392,
      "grad_norm": 1.105571901275732,
      "learning_rate": 2e-05,
      "loss": 0.6982758998870849,
      "step": 540
    },
    {
      "epoch": 0.041796489094916026,
      "grad_norm": 0.9941727634113782,
      "learning_rate": 2e-05,
      "loss": 0.6900170803070068,
      "step": 550
    },
    {
      "epoch": 0.042556425260278134,
      "grad_norm": 0.9835023836804377,
      "learning_rate": 2e-05,
      "loss": 0.6558903694152832,
      "step": 560
    },
    {
      "epoch": 0.04331636142564024,
      "grad_norm": 1.1595810180524677,
      "learning_rate": 2e-05,
      "loss": 0.6865039825439453,
      "step": 570
    },
    {
      "epoch": 0.04407629759100236,
      "grad_norm": 2.2186915798904154,
      "learning_rate": 2e-05,
      "loss": 0.6749402999877929,
      "step": 580
    },
    {
      "epoch": 0.04483623375636447,
      "grad_norm": 1.3118057884159302,
      "learning_rate": 2e-05,
      "loss": 0.675851583480835,
      "step": 590
    },
    {
      "epoch": 0.045596169921726576,
      "grad_norm": 1.692496987832561,
      "learning_rate": 2e-05,
      "loss": 0.6745028495788574,
      "step": 600
    },
    {
      "epoch": 0.046356106087088685,
      "grad_norm": 1.9730825283501034,
      "learning_rate": 2e-05,
      "loss": 0.6651962280273438,
      "step": 610
    },
    {
      "epoch": 0.04711604225245079,
      "grad_norm": 1.331976977887419,
      "learning_rate": 2e-05,
      "loss": 0.6501778602600098,
      "step": 620
    },
    {
      "epoch": 0.0478759784178129,
      "grad_norm": 1.5637084886303765,
      "learning_rate": 2e-05,
      "loss": 0.7042662620544433,
      "step": 630
    },
    {
      "epoch": 0.04863591458317501,
      "grad_norm": 1.4726774498399116,
      "learning_rate": 2e-05,
      "loss": 0.6678030967712403,
      "step": 640
    },
    {
      "epoch": 0.049395850748537126,
      "grad_norm": 2.868807910273408,
      "learning_rate": 2e-05,
      "loss": 0.6779534339904785,
      "step": 650
    },
    {
      "epoch": 0.050155786913899235,
      "grad_norm": 1.4317134087883574,
      "learning_rate": 2e-05,
      "loss": 0.6799858093261719,
      "step": 660
    },
    {
      "epoch": 0.050915723079261344,
      "grad_norm": 1.7227333001916365,
      "learning_rate": 2e-05,
      "loss": 0.6573725700378418,
      "step": 670
    },
    {
      "epoch": 0.05167565924462345,
      "grad_norm": 1.3271414274442734,
      "learning_rate": 2e-05,
      "loss": 0.6701514244079589,
      "step": 680
    },
    {
      "epoch": 0.05243559540998556,
      "grad_norm": 1.5421124248057487,
      "learning_rate": 2e-05,
      "loss": 0.6800753593444824,
      "step": 690
    },
    {
      "epoch": 0.05319553157534767,
      "grad_norm": 1.0972092785185035,
      "learning_rate": 2e-05,
      "loss": 0.6628331184387207,
      "step": 700
    },
    {
      "epoch": 0.05395546774070978,
      "grad_norm": 1.8016898257371647,
      "learning_rate": 2e-05,
      "loss": 0.6686877250671387,
      "step": 710
    },
    {
      "epoch": 0.05471540390607189,
      "grad_norm": 1.3817551649658868,
      "learning_rate": 2e-05,
      "loss": 0.691933012008667,
      "step": 720
    },
    {
      "epoch": 0.055475340071434,
      "grad_norm": 4.390986381327533,
      "learning_rate": 2e-05,
      "loss": 0.6459454536437989,
      "step": 730
    },
    {
      "epoch": 0.05623527623679611,
      "grad_norm": 1.049270882156613,
      "learning_rate": 2e-05,
      "loss": 0.7004701137542725,
      "step": 740
    },
    {
      "epoch": 0.05699521240215822,
      "grad_norm": 2.5788017211289134,
      "learning_rate": 2e-05,
      "loss": 0.6631406784057617,
      "step": 750
    },
    {
      "epoch": 0.05775514856752033,
      "grad_norm": 4.003076354285841,
      "learning_rate": 2e-05,
      "loss": 0.6525260925292968,
      "step": 760
    },
    {
      "epoch": 0.05851508473288244,
      "grad_norm": 3.856641478939966,
      "learning_rate": 2e-05,
      "loss": 0.6481318950653077,
      "step": 770
    },
    {
      "epoch": 0.059275020898244546,
      "grad_norm": 2.1050201546733516,
      "learning_rate": 2e-05,
      "loss": 0.6629796504974366,
      "step": 780
    },
    {
      "epoch": 0.060034957063606655,
      "grad_norm": 2.4972482058867684,
      "learning_rate": 2e-05,
      "loss": 0.6793940544128418,
      "step": 790
    },
    {
      "epoch": 0.06079489322896876,
      "grad_norm": 2.9023733973854338,
      "learning_rate": 2e-05,
      "loss": 0.6557552337646484,
      "step": 800
    },
    {
      "epoch": 0.06155482939433088,
      "grad_norm": 1.753191005851473,
      "learning_rate": 2e-05,
      "loss": 0.6550214767456055,
      "step": 810
    },
    {
      "epoch": 0.06231476555969299,
      "grad_norm": 3.6008366995268677,
      "learning_rate": 2e-05,
      "loss": 0.6532149314880371,
      "step": 820
    },
    {
      "epoch": 0.0630747017250551,
      "grad_norm": 3.9574066156632264,
      "learning_rate": 2e-05,
      "loss": 0.664048147201538,
      "step": 830
    },
    {
      "epoch": 0.06383463789041721,
      "grad_norm": 1.4500186208031258,
      "learning_rate": 2e-05,
      "loss": 0.668964147567749,
      "step": 840
    },
    {
      "epoch": 0.06459457405577931,
      "grad_norm": 1.8650680203909065,
      "learning_rate": 2e-05,
      "loss": 0.6597201824188232,
      "step": 850
    },
    {
      "epoch": 0.06535451022114143,
      "grad_norm": 1.286387371803429,
      "learning_rate": 2e-05,
      "loss": 0.6578952312469483,
      "step": 860
    },
    {
      "epoch": 0.06611444638650353,
      "grad_norm": 44.161007706426496,
      "learning_rate": 2e-05,
      "loss": 0.6617650508880615,
      "step": 870
    },
    {
      "epoch": 0.06687438255186565,
      "grad_norm": 8.70858945322814,
      "learning_rate": 2e-05,
      "loss": 0.6729881286621093,
      "step": 880
    },
    {
      "epoch": 0.06763431871722775,
      "grad_norm": 2.0716744352445295,
      "learning_rate": 2e-05,
      "loss": 0.6804790496826172,
      "step": 890
    },
    {
      "epoch": 0.06839425488258986,
      "grad_norm": 2.3995890552247063,
      "learning_rate": 2e-05,
      "loss": 0.6576640605926514,
      "step": 900
    },
    {
      "epoch": 0.06915419104795197,
      "grad_norm": 7.020838251383462,
      "learning_rate": 2e-05,
      "loss": 0.6924066066741943,
      "step": 910
    },
    {
      "epoch": 0.06991412721331408,
      "grad_norm": 1.2111035584957595,
      "learning_rate": 2e-05,
      "loss": 0.6660052299499511,
      "step": 920
    },
    {
      "epoch": 0.0706740633786762,
      "grad_norm": 1.8673454485620073,
      "learning_rate": 2e-05,
      "loss": 0.6890796661376953,
      "step": 930
    },
    {
      "epoch": 0.0714339995440383,
      "grad_norm": 2.462929162354763,
      "learning_rate": 2e-05,
      "loss": 0.6732182979583741,
      "step": 940
    },
    {
      "epoch": 0.07219393570940041,
      "grad_norm": 1.8507788704761199,
      "learning_rate": 2e-05,
      "loss": 0.6905992031097412,
      "step": 950
    },
    {
      "epoch": 0.07295387187476252,
      "grad_norm": 1.5233611761953871,
      "learning_rate": 2e-05,
      "loss": 0.6639741897583008,
      "step": 960
    },
    {
      "epoch": 0.07371380804012463,
      "grad_norm": 1.0039057924990893,
      "learning_rate": 2e-05,
      "loss": 0.6618095397949219,
      "step": 970
    },
    {
      "epoch": 0.07447374420548673,
      "grad_norm": 1.59146144447183,
      "learning_rate": 2e-05,
      "loss": 0.6541851043701172,
      "step": 980
    },
    {
      "epoch": 0.07523368037084885,
      "grad_norm": 1.671042860219738,
      "learning_rate": 2e-05,
      "loss": 0.6654860496520996,
      "step": 990
    },
    {
      "epoch": 0.07599361653621096,
      "grad_norm": 1.198120650365706,
      "learning_rate": 2e-05,
      "loss": 0.6552873134613038,
      "step": 1000
    },
    {
      "epoch": 0.07675355270157307,
      "grad_norm": 8.77583610709108,
      "learning_rate": 2e-05,
      "loss": 0.6433052062988281,
      "step": 1010
    },
    {
      "epoch": 0.07751348886693518,
      "grad_norm": 1.111330096476852,
      "learning_rate": 2e-05,
      "loss": 0.642775297164917,
      "step": 1020
    },
    {
      "epoch": 0.07827342503229728,
      "grad_norm": 1.348192309325266,
      "learning_rate": 2e-05,
      "loss": 0.6692739009857178,
      "step": 1030
    },
    {
      "epoch": 0.0790333611976594,
      "grad_norm": 1.0407807523199857,
      "learning_rate": 2e-05,
      "loss": 0.6457259178161621,
      "step": 1040
    },
    {
      "epoch": 0.0797932973630215,
      "grad_norm": 1.0582735573079416,
      "learning_rate": 2e-05,
      "loss": 0.6610038757324219,
      "step": 1050
    },
    {
      "epoch": 0.08055323352838362,
      "grad_norm": 2.1749317446696894,
      "learning_rate": 2e-05,
      "loss": 0.6602062225341797,
      "step": 1060
    },
    {
      "epoch": 0.08131316969374573,
      "grad_norm": 1.2523925736905532,
      "learning_rate": 2e-05,
      "loss": 0.6483442306518554,
      "step": 1070
    },
    {
      "epoch": 0.08207310585910783,
      "grad_norm": 1.2772297314001628,
      "learning_rate": 2e-05,
      "loss": 0.6683950424194336,
      "step": 1080
    },
    {
      "epoch": 0.08283304202446995,
      "grad_norm": 3.3387060909226665,
      "learning_rate": 2e-05,
      "loss": 0.6565097808837891,
      "step": 1090
    },
    {
      "epoch": 0.08359297818983205,
      "grad_norm": 1.0681126627729676,
      "learning_rate": 2e-05,
      "loss": 0.6453925609588623,
      "step": 1100
    },
    {
      "epoch": 0.08435291435519417,
      "grad_norm": 1.0491354731716318,
      "learning_rate": 2e-05,
      "loss": 0.672363567352295,
      "step": 1110
    },
    {
      "epoch": 0.08511285052055627,
      "grad_norm": 2.5229554034222126,
      "learning_rate": 2e-05,
      "loss": 0.6442315101623535,
      "step": 1120
    },
    {
      "epoch": 0.08587278668591838,
      "grad_norm": 0.9549713568548055,
      "learning_rate": 2e-05,
      "loss": 0.6467645168304443,
      "step": 1130
    },
    {
      "epoch": 0.08663272285128049,
      "grad_norm": 1.0994614474759026,
      "learning_rate": 2e-05,
      "loss": 0.6226651668548584,
      "step": 1140
    },
    {
      "epoch": 0.0873926590166426,
      "grad_norm": 1.5931672722656118,
      "learning_rate": 2e-05,
      "loss": 0.6385224819183349,
      "step": 1150
    },
    {
      "epoch": 0.08815259518200472,
      "grad_norm": 1.3058693072378955,
      "learning_rate": 2e-05,
      "loss": 0.6373834133148193,
      "step": 1160
    },
    {
      "epoch": 0.08891253134736682,
      "grad_norm": 1.498709878738794,
      "learning_rate": 2e-05,
      "loss": 0.6416314601898193,
      "step": 1170
    },
    {
      "epoch": 0.08967246751272893,
      "grad_norm": 1.6779770963133274,
      "learning_rate": 2e-05,
      "loss": 0.6468165874481201,
      "step": 1180
    },
    {
      "epoch": 0.09043240367809104,
      "grad_norm": 1.454215379835705,
      "learning_rate": 2e-05,
      "loss": 0.6477084636688233,
      "step": 1190
    },
    {
      "epoch": 0.09119233984345315,
      "grad_norm": 1.086597371932175,
      "learning_rate": 2e-05,
      "loss": 0.6345207214355468,
      "step": 1200
    },
    {
      "epoch": 0.09195227600881525,
      "grad_norm": 1.1041391627395598,
      "learning_rate": 2e-05,
      "loss": 0.6753363609313965,
      "step": 1210
    },
    {
      "epoch": 0.09271221217417737,
      "grad_norm": 1.0442979911159636,
      "learning_rate": 2e-05,
      "loss": 0.6602447509765625,
      "step": 1220
    },
    {
      "epoch": 0.09347214833953948,
      "grad_norm": 1.1068350401318283,
      "learning_rate": 2e-05,
      "loss": 0.6334839820861816,
      "step": 1230
    },
    {
      "epoch": 0.09423208450490159,
      "grad_norm": 3.9758019839117336,
      "learning_rate": 2e-05,
      "loss": 0.6504351615905761,
      "step": 1240
    },
    {
      "epoch": 0.0949920206702637,
      "grad_norm": 0.9541829838291032,
      "learning_rate": 2e-05,
      "loss": 0.6485028743743897,
      "step": 1250
    },
    {
      "epoch": 0.0957519568356258,
      "grad_norm": 1.164445477460499,
      "learning_rate": 2e-05,
      "loss": 0.6839859962463379,
      "step": 1260
    },
    {
      "epoch": 0.09651189300098792,
      "grad_norm": 1.8779857955368715,
      "learning_rate": 2e-05,
      "loss": 0.6411514282226562,
      "step": 1270
    },
    {
      "epoch": 0.09727182916635002,
      "grad_norm": 1.147603065647347,
      "learning_rate": 2e-05,
      "loss": 0.6255774974822998,
      "step": 1280
    },
    {
      "epoch": 0.09803176533171214,
      "grad_norm": 1.6550437152636595,
      "learning_rate": 2e-05,
      "loss": 0.6291728019714355,
      "step": 1290
    },
    {
      "epoch": 0.09879170149707425,
      "grad_norm": 1.7682464495720873,
      "learning_rate": 2e-05,
      "loss": 0.659593915939331,
      "step": 1300
    },
    {
      "epoch": 0.09955163766243635,
      "grad_norm": 3.7022883686619323,
      "learning_rate": 2e-05,
      "loss": 0.650638484954834,
      "step": 1310
    },
    {
      "epoch": 0.10031157382779847,
      "grad_norm": 1.0466634365071796,
      "learning_rate": 2e-05,
      "loss": 0.6600216865539551,
      "step": 1320
    },
    {
      "epoch": 0.10107150999316057,
      "grad_norm": 6.303642093550111,
      "learning_rate": 2e-05,
      "loss": 0.6597589015960693,
      "step": 1330
    },
    {
      "epoch": 0.10183144615852269,
      "grad_norm": 1.093978067554938,
      "learning_rate": 2e-05,
      "loss": 0.6573177337646484,
      "step": 1340
    },
    {
      "epoch": 0.10259138232388479,
      "grad_norm": 2.0424198171963748,
      "learning_rate": 2e-05,
      "loss": 0.6553116798400879,
      "step": 1350
    },
    {
      "epoch": 0.1033513184892469,
      "grad_norm": 1.5430270988592787,
      "learning_rate": 2e-05,
      "loss": 0.6318343162536622,
      "step": 1360
    },
    {
      "epoch": 0.104111254654609,
      "grad_norm": 2.1316696565809528,
      "learning_rate": 2e-05,
      "loss": 0.6384878158569336,
      "step": 1370
    },
    {
      "epoch": 0.10487119081997112,
      "grad_norm": 2.91741562357673,
      "learning_rate": 2e-05,
      "loss": 0.6429308891296387,
      "step": 1380
    },
    {
      "epoch": 0.10563112698533324,
      "grad_norm": 1.0434912962088092,
      "learning_rate": 2e-05,
      "loss": 0.6408050537109375,
      "step": 1390
    },
    {
      "epoch": 0.10639106315069534,
      "grad_norm": 1.224807798537247,
      "learning_rate": 2e-05,
      "loss": 0.6579841613769531,
      "step": 1400
    },
    {
      "epoch": 0.10715099931605745,
      "grad_norm": 2.0847642989918778,
      "learning_rate": 2e-05,
      "loss": 0.6498456954956054,
      "step": 1410
    },
    {
      "epoch": 0.10791093548141956,
      "grad_norm": 1.108872735327033,
      "learning_rate": 2e-05,
      "loss": 0.6492026329040528,
      "step": 1420
    },
    {
      "epoch": 0.10867087164678167,
      "grad_norm": 0.8206739744796532,
      "learning_rate": 2e-05,
      "loss": 0.6415225028991699,
      "step": 1430
    },
    {
      "epoch": 0.10943080781214377,
      "grad_norm": 1.5508982512432574,
      "learning_rate": 2e-05,
      "loss": 0.6333587169647217,
      "step": 1440
    },
    {
      "epoch": 0.11019074397750589,
      "grad_norm": 1.215268801016134,
      "learning_rate": 2e-05,
      "loss": 0.6485206127166748,
      "step": 1450
    },
    {
      "epoch": 0.110950680142868,
      "grad_norm": 1.270810180433496,
      "learning_rate": 2e-05,
      "loss": 0.6261978149414062,
      "step": 1460
    },
    {
      "epoch": 0.1117106163082301,
      "grad_norm": 1.2131825076098512,
      "learning_rate": 2e-05,
      "loss": 0.6661453247070312,
      "step": 1470
    },
    {
      "epoch": 0.11247055247359222,
      "grad_norm": 0.9731960092017137,
      "learning_rate": 2e-05,
      "loss": 0.659299087524414,
      "step": 1480
    },
    {
      "epoch": 0.11323048863895432,
      "grad_norm": 1.433503129247427,
      "learning_rate": 2e-05,
      "loss": 0.6485359191894531,
      "step": 1490
    },
    {
      "epoch": 0.11399042480431644,
      "grad_norm": 1.3500939927983868,
      "learning_rate": 2e-05,
      "loss": 0.6578350067138672,
      "step": 1500
    },
    {
      "epoch": 0.11475036096967854,
      "grad_norm": 1.4918825355601473,
      "learning_rate": 2e-05,
      "loss": 0.6433755397796631,
      "step": 1510
    },
    {
      "epoch": 0.11551029713504066,
      "grad_norm": 1.8872836071366936,
      "learning_rate": 2e-05,
      "loss": 0.6468602657318115,
      "step": 1520
    },
    {
      "epoch": 0.11627023330040277,
      "grad_norm": 2.3590696751049545,
      "learning_rate": 2e-05,
      "loss": 0.6331050395965576,
      "step": 1530
    },
    {
      "epoch": 0.11703016946576487,
      "grad_norm": 1.333851094236431,
      "learning_rate": 2e-05,
      "loss": 0.667790412902832,
      "step": 1540
    },
    {
      "epoch": 0.11779010563112699,
      "grad_norm": 1.259990456534997,
      "learning_rate": 2e-05,
      "loss": 0.6469200134277344,
      "step": 1550
    },
    {
      "epoch": 0.11855004179648909,
      "grad_norm": 1.7108862137007483,
      "learning_rate": 2e-05,
      "loss": 0.6639731884002685,
      "step": 1560
    },
    {
      "epoch": 0.11930997796185121,
      "grad_norm": 1.4148747471318508,
      "learning_rate": 2e-05,
      "loss": 0.6579952239990234,
      "step": 1570
    },
    {
      "epoch": 0.12006991412721331,
      "grad_norm": 2.8707717322914195,
      "learning_rate": 2e-05,
      "loss": 0.6224180221557617,
      "step": 1580
    },
    {
      "epoch": 0.12082985029257542,
      "grad_norm": 2.6283336063490403,
      "learning_rate": 2e-05,
      "loss": 0.6329303741455078,
      "step": 1590
    },
    {
      "epoch": 0.12158978645793753,
      "grad_norm": 3.150067996130026,
      "learning_rate": 2e-05,
      "loss": 0.6741673946380615,
      "step": 1600
    },
    {
      "epoch": 0.12234972262329964,
      "grad_norm": 1.8279905616934546,
      "learning_rate": 2e-05,
      "loss": 0.659599781036377,
      "step": 1610
    },
    {
      "epoch": 0.12310965878866176,
      "grad_norm": 4.748332853059669,
      "learning_rate": 2e-05,
      "loss": 0.6577554702758789,
      "step": 1620
    },
    {
      "epoch": 0.12386959495402386,
      "grad_norm": 1.4163870558356635,
      "learning_rate": 2e-05,
      "loss": 0.6562213897705078,
      "step": 1630
    },
    {
      "epoch": 0.12462953111938598,
      "grad_norm": 6.694979985306562,
      "learning_rate": 2e-05,
      "loss": 0.6411036491394043,
      "step": 1640
    },
    {
      "epoch": 0.1253894672847481,
      "grad_norm": 6.209427745695381,
      "learning_rate": 2e-05,
      "loss": 0.6805436611175537,
      "step": 1650
    },
    {
      "epoch": 0.1261494034501102,
      "grad_norm": 13.756950170865382,
      "learning_rate": 2e-05,
      "loss": 0.6436635017395019,
      "step": 1660
    },
    {
      "epoch": 0.1269093396154723,
      "grad_norm": 3.997563874538566,
      "learning_rate": 2e-05,
      "loss": 0.6643521785736084,
      "step": 1670
    },
    {
      "epoch": 0.12766927578083442,
      "grad_norm": 2.114093971367723,
      "learning_rate": 2e-05,
      "loss": 0.6300763607025146,
      "step": 1680
    },
    {
      "epoch": 0.12842921194619653,
      "grad_norm": 7.777372686568945,
      "learning_rate": 2e-05,
      "loss": 0.6559924125671387,
      "step": 1690
    },
    {
      "epoch": 0.12918914811155863,
      "grad_norm": 5.240897859578015,
      "learning_rate": 2e-05,
      "loss": 0.6357455253601074,
      "step": 1700
    },
    {
      "epoch": 0.12994908427692073,
      "grad_norm": 6.9535155450420545,
      "learning_rate": 2e-05,
      "loss": 0.6478095531463623,
      "step": 1710
    },
    {
      "epoch": 0.13070902044228286,
      "grad_norm": 2.358848628288785,
      "learning_rate": 2e-05,
      "loss": 0.6790260314941406,
      "step": 1720
    },
    {
      "epoch": 0.13146895660764496,
      "grad_norm": 2.0138319749117786,
      "learning_rate": 2e-05,
      "loss": 0.6263014793395996,
      "step": 1730
    },
    {
      "epoch": 0.13222889277300706,
      "grad_norm": 1.4025723929068667,
      "learning_rate": 2e-05,
      "loss": 0.6489272117614746,
      "step": 1740
    },
    {
      "epoch": 0.13298882893836916,
      "grad_norm": 2.723181912456477,
      "learning_rate": 2e-05,
      "loss": 0.6445198059082031,
      "step": 1750
    },
    {
      "epoch": 0.1337487651037313,
      "grad_norm": 2.9660220967238637,
      "learning_rate": 2e-05,
      "loss": 0.6173060417175293,
      "step": 1760
    },
    {
      "epoch": 0.1345087012690934,
      "grad_norm": 1.548833654772482,
      "learning_rate": 2e-05,
      "loss": 0.6361001014709473,
      "step": 1770
    },
    {
      "epoch": 0.1352686374344555,
      "grad_norm": 1.4209226521258322,
      "learning_rate": 2e-05,
      "loss": 0.6101439476013184,
      "step": 1780
    },
    {
      "epoch": 0.13602857359981763,
      "grad_norm": 5.083546421164765,
      "learning_rate": 2e-05,
      "loss": 0.6381206512451172,
      "step": 1790
    },
    {
      "epoch": 0.13678850976517973,
      "grad_norm": 2.496640617510042,
      "learning_rate": 2e-05,
      "loss": 0.6482817649841308,
      "step": 1800
    },
    {
      "epoch": 0.13754844593054183,
      "grad_norm": 1.9610047412853615,
      "learning_rate": 2e-05,
      "loss": 0.6429681301116943,
      "step": 1810
    },
    {
      "epoch": 0.13830838209590393,
      "grad_norm": 1.3428122362755035,
      "learning_rate": 2e-05,
      "loss": 0.6612567901611328,
      "step": 1820
    },
    {
      "epoch": 0.13906831826126606,
      "grad_norm": 1.0480266633178372,
      "learning_rate": 2e-05,
      "loss": 0.634760570526123,
      "step": 1830
    },
    {
      "epoch": 0.13982825442662816,
      "grad_norm": 5.767390275358731,
      "learning_rate": 2e-05,
      "loss": 0.6517911434173584,
      "step": 1840
    },
    {
      "epoch": 0.14058819059199026,
      "grad_norm": 1.0716026596322161,
      "learning_rate": 2e-05,
      "loss": 0.6544319152832031,
      "step": 1850
    },
    {
      "epoch": 0.1413481267573524,
      "grad_norm": 1.846290383372757,
      "learning_rate": 2e-05,
      "loss": 0.6552516460418701,
      "step": 1860
    },
    {
      "epoch": 0.1421080629227145,
      "grad_norm": 1.5831829672964501,
      "learning_rate": 2e-05,
      "loss": 0.6423295497894287,
      "step": 1870
    },
    {
      "epoch": 0.1428679990880766,
      "grad_norm": 5.440399705277345,
      "learning_rate": 2e-05,
      "loss": 0.6394572257995605,
      "step": 1880
    },
    {
      "epoch": 0.1436279352534387,
      "grad_norm": 1.9378312669131104,
      "learning_rate": 2e-05,
      "loss": 0.6410587310791016,
      "step": 1890
    },
    {
      "epoch": 0.14438787141880083,
      "grad_norm": 1.4028689103168956,
      "learning_rate": 2e-05,
      "loss": 0.6463975429534912,
      "step": 1900
    },
    {
      "epoch": 0.14514780758416293,
      "grad_norm": 1.5006186537122213,
      "learning_rate": 2e-05,
      "loss": 0.6331391334533691,
      "step": 1910
    },
    {
      "epoch": 0.14590774374952503,
      "grad_norm": 1.4896410731969767,
      "learning_rate": 2e-05,
      "loss": 0.6415359497070312,
      "step": 1920
    },
    {
      "epoch": 0.14666767991488716,
      "grad_norm": 1.4531220265216627,
      "learning_rate": 2e-05,
      "loss": 0.6483102321624756,
      "step": 1930
    },
    {
      "epoch": 0.14742761608024926,
      "grad_norm": 4.060292836092577,
      "learning_rate": 2e-05,
      "loss": 0.6400858879089355,
      "step": 1940
    },
    {
      "epoch": 0.14818755224561136,
      "grad_norm": 2.01518756849028,
      "learning_rate": 2e-05,
      "loss": 0.6560895442962646,
      "step": 1950
    },
    {
      "epoch": 0.14894748841097347,
      "grad_norm": 1.1943584364475957,
      "learning_rate": 2e-05,
      "loss": 0.6442777156829834,
      "step": 1960
    },
    {
      "epoch": 0.1497074245763356,
      "grad_norm": 1.2365114760287763,
      "learning_rate": 2e-05,
      "loss": 0.615467643737793,
      "step": 1970
    },
    {
      "epoch": 0.1504673607416977,
      "grad_norm": 1.4776860920715695,
      "learning_rate": 2e-05,
      "loss": 0.6682824611663818,
      "step": 1980
    },
    {
      "epoch": 0.1512272969070598,
      "grad_norm": 1.4579971244761316,
      "learning_rate": 2e-05,
      "loss": 0.6257425308227539,
      "step": 1990
    },
    {
      "epoch": 0.15198723307242193,
      "grad_norm": 10.148842969854169,
      "learning_rate": 2e-05,
      "loss": 0.6339943885803223,
      "step": 2000
    },
    {
      "epoch": 0.15274716923778403,
      "grad_norm": 5.442411967582219,
      "learning_rate": 2e-05,
      "loss": 0.6389657497406006,
      "step": 2010
    },
    {
      "epoch": 0.15350710540314613,
      "grad_norm": 16.30352466832802,
      "learning_rate": 2e-05,
      "loss": 0.6474606990814209,
      "step": 2020
    },
    {
      "epoch": 0.15426704156850823,
      "grad_norm": 4.362878921974434,
      "learning_rate": 2e-05,
      "loss": 0.6363828659057618,
      "step": 2030
    },
    {
      "epoch": 0.15502697773387036,
      "grad_norm": 1.2676245703999107,
      "learning_rate": 2e-05,
      "loss": 0.6665027618408204,
      "step": 2040
    },
    {
      "epoch": 0.15578691389923247,
      "grad_norm": 1.0765397406434796,
      "learning_rate": 2e-05,
      "loss": 0.6307790279388428,
      "step": 2050
    },
    {
      "epoch": 0.15654685006459457,
      "grad_norm": 2.6315823864120396,
      "learning_rate": 2e-05,
      "loss": 0.661501693725586,
      "step": 2060
    },
    {
      "epoch": 0.1573067862299567,
      "grad_norm": 0.917152666671035,
      "learning_rate": 2e-05,
      "loss": 0.6306610107421875,
      "step": 2070
    },
    {
      "epoch": 0.1580667223953188,
      "grad_norm": 1.0959801674463898,
      "learning_rate": 2e-05,
      "loss": 0.6308591365814209,
      "step": 2080
    },
    {
      "epoch": 0.1588266585606809,
      "grad_norm": 2.2239738262841513,
      "learning_rate": 2e-05,
      "loss": 0.6311299800872803,
      "step": 2090
    },
    {
      "epoch": 0.159586594726043,
      "grad_norm": 3.4139655597175462,
      "learning_rate": 2e-05,
      "loss": 0.6391436576843261,
      "step": 2100
    },
    {
      "epoch": 0.16034653089140513,
      "grad_norm": 1.5249315901105078,
      "learning_rate": 2e-05,
      "loss": 0.6251719951629638,
      "step": 2110
    },
    {
      "epoch": 0.16110646705676723,
      "grad_norm": 5.887416970765259,
      "learning_rate": 2e-05,
      "loss": 0.6304204940795899,
      "step": 2120
    },
    {
      "epoch": 0.16186640322212933,
      "grad_norm": 1.3922439274444605,
      "learning_rate": 2e-05,
      "loss": 0.6613870143890381,
      "step": 2130
    },
    {
      "epoch": 0.16262633938749146,
      "grad_norm": 1.9337982890171186,
      "learning_rate": 2e-05,
      "loss": 0.6491190910339355,
      "step": 2140
    },
    {
      "epoch": 0.16338627555285357,
      "grad_norm": 1.3724422818318218,
      "learning_rate": 2e-05,
      "loss": 0.6484801769256592,
      "step": 2150
    },
    {
      "epoch": 0.16414621171821567,
      "grad_norm": 2.591228095277901,
      "learning_rate": 2e-05,
      "loss": 0.651193380355835,
      "step": 2160
    },
    {
      "epoch": 0.16490614788357777,
      "grad_norm": 1.6151307202072216,
      "learning_rate": 2e-05,
      "loss": 0.6365288734436035,
      "step": 2170
    },
    {
      "epoch": 0.1656660840489399,
      "grad_norm": 1.546438395722935,
      "learning_rate": 2e-05,
      "loss": 0.6704031944274902,
      "step": 2180
    },
    {
      "epoch": 0.166426020214302,
      "grad_norm": 2.407511950274536,
      "learning_rate": 2e-05,
      "loss": 0.6353227138519287,
      "step": 2190
    },
    {
      "epoch": 0.1671859563796641,
      "grad_norm": 1.6797534787574688,
      "learning_rate": 2e-05,
      "loss": 0.6494597434997559,
      "step": 2200
    },
    {
      "epoch": 0.1679458925450262,
      "grad_norm": 4.8613761266230835,
      "learning_rate": 2e-05,
      "loss": 0.6252596855163575,
      "step": 2210
    },
    {
      "epoch": 0.16870582871038833,
      "grad_norm": 0.99515726828342,
      "learning_rate": 2e-05,
      "loss": 0.6384762763977051,
      "step": 2220
    },
    {
      "epoch": 0.16946576487575044,
      "grad_norm": 1.1520248314117032,
      "learning_rate": 2e-05,
      "loss": 0.6748219966888428,
      "step": 2230
    },
    {
      "epoch": 0.17022570104111254,
      "grad_norm": 1.861088340323805,
      "learning_rate": 2e-05,
      "loss": 0.6419329643249512,
      "step": 2240
    },
    {
      "epoch": 0.17098563720647467,
      "grad_norm": 1.412748370534858,
      "learning_rate": 2e-05,
      "loss": 0.6323102474212646,
      "step": 2250
    },
    {
      "epoch": 0.17174557337183677,
      "grad_norm": 0.8560502316583172,
      "learning_rate": 2e-05,
      "loss": 0.6213559150695801,
      "step": 2260
    },
    {
      "epoch": 0.17250550953719887,
      "grad_norm": 1.2057655077981024,
      "learning_rate": 2e-05,
      "loss": 0.6556430339813233,
      "step": 2270
    },
    {
      "epoch": 0.17326544570256097,
      "grad_norm": 0.9239999124842788,
      "learning_rate": 2e-05,
      "loss": 0.6087024211883545,
      "step": 2280
    },
    {
      "epoch": 0.1740253818679231,
      "grad_norm": 0.9516737623691233,
      "learning_rate": 2e-05,
      "loss": 0.6084947109222412,
      "step": 2290
    },
    {
      "epoch": 0.1747853180332852,
      "grad_norm": 2.815786294259301,
      "learning_rate": 2e-05,
      "loss": 0.6483430385589599,
      "step": 2300
    },
    {
      "epoch": 0.1755452541986473,
      "grad_norm": 0.8652265177148013,
      "learning_rate": 2e-05,
      "loss": 0.6458633422851563,
      "step": 2310
    },
    {
      "epoch": 0.17630519036400943,
      "grad_norm": 1.6069854127598282,
      "learning_rate": 2e-05,
      "loss": 0.639268445968628,
      "step": 2320
    },
    {
      "epoch": 0.17706512652937154,
      "grad_norm": 1.2760797527559364,
      "learning_rate": 2e-05,
      "loss": 0.6464717864990235,
      "step": 2330
    },
    {
      "epoch": 0.17782506269473364,
      "grad_norm": 1.1069755420870164,
      "learning_rate": 2e-05,
      "loss": 0.6435245513916016,
      "step": 2340
    },
    {
      "epoch": 0.17858499886009574,
      "grad_norm": 3.3877427199201224,
      "learning_rate": 2e-05,
      "loss": 0.6325599193572998,
      "step": 2350
    },
    {
      "epoch": 0.17934493502545787,
      "grad_norm": 1.1611639959868179,
      "learning_rate": 2e-05,
      "loss": 0.6352363109588623,
      "step": 2360
    },
    {
      "epoch": 0.18010487119081997,
      "grad_norm": 1.530158351349937,
      "learning_rate": 2e-05,
      "loss": 0.6382848262786865,
      "step": 2370
    },
    {
      "epoch": 0.18086480735618207,
      "grad_norm": 0.828975583603307,
      "learning_rate": 2e-05,
      "loss": 0.6329381465911865,
      "step": 2380
    },
    {
      "epoch": 0.1816247435215442,
      "grad_norm": 1.9691779255291941,
      "learning_rate": 2e-05,
      "loss": 0.6218951225280762,
      "step": 2390
    },
    {
      "epoch": 0.1823846796869063,
      "grad_norm": 8.233905278346574,
      "learning_rate": 2e-05,
      "loss": 0.6378360748291015,
      "step": 2400
    },
    {
      "epoch": 0.1831446158522684,
      "grad_norm": 1.253360788923102,
      "learning_rate": 2e-05,
      "loss": 0.639088773727417,
      "step": 2410
    },
    {
      "epoch": 0.1839045520176305,
      "grad_norm": 1.0639498166629777,
      "learning_rate": 2e-05,
      "loss": 0.648853063583374,
      "step": 2420
    },
    {
      "epoch": 0.18466448818299264,
      "grad_norm": 1.224113932980045,
      "learning_rate": 2e-05,
      "loss": 0.6247108459472657,
      "step": 2430
    },
    {
      "epoch": 0.18542442434835474,
      "grad_norm": 0.862916274591829,
      "learning_rate": 2e-05,
      "loss": 0.6458800315856934,
      "step": 2440
    },
    {
      "epoch": 0.18618436051371684,
      "grad_norm": 1.0838844297963148,
      "learning_rate": 2e-05,
      "loss": 0.6351897239685058,
      "step": 2450
    },
    {
      "epoch": 0.18694429667907897,
      "grad_norm": 0.9652999261625115,
      "learning_rate": 2e-05,
      "loss": 0.6378411293029785,
      "step": 2460
    },
    {
      "epoch": 0.18770423284444107,
      "grad_norm": 0.7979754847702989,
      "learning_rate": 2e-05,
      "loss": 0.6047022819519043,
      "step": 2470
    },
    {
      "epoch": 0.18846416900980317,
      "grad_norm": 2.021558850612358,
      "learning_rate": 2e-05,
      "loss": 0.6230844497680664,
      "step": 2480
    },
    {
      "epoch": 0.18922410517516527,
      "grad_norm": 0.8946816777799294,
      "learning_rate": 2e-05,
      "loss": 0.6341133117675781,
      "step": 2490
    },
    {
      "epoch": 0.1899840413405274,
      "grad_norm": 1.479766364730666,
      "learning_rate": 2e-05,
      "loss": 0.6266453742980957,
      "step": 2500
    },
    {
      "epoch": 0.1907439775058895,
      "grad_norm": 1.9037844174773604,
      "learning_rate": 2e-05,
      "loss": 0.6139904975891113,
      "step": 2510
    },
    {
      "epoch": 0.1915039136712516,
      "grad_norm": 0.8961866794407122,
      "learning_rate": 2e-05,
      "loss": 0.608923864364624,
      "step": 2520
    },
    {
      "epoch": 0.19226384983661374,
      "grad_norm": 1.6522838694053719,
      "learning_rate": 2e-05,
      "loss": 0.6544345378875732,
      "step": 2530
    },
    {
      "epoch": 0.19302378600197584,
      "grad_norm": 1.1182433232512783,
      "learning_rate": 2e-05,
      "loss": 0.6411478042602539,
      "step": 2540
    },
    {
      "epoch": 0.19378372216733794,
      "grad_norm": 1.9063805360822343,
      "learning_rate": 2e-05,
      "loss": 0.6063728332519531,
      "step": 2550
    },
    {
      "epoch": 0.19454365833270004,
      "grad_norm": 1.339954525004715,
      "learning_rate": 2e-05,
      "loss": 0.6316645622253418,
      "step": 2560
    },
    {
      "epoch": 0.19530359449806217,
      "grad_norm": 1.075456792572934,
      "learning_rate": 2e-05,
      "loss": 0.6289247035980224,
      "step": 2570
    },
    {
      "epoch": 0.19606353066342427,
      "grad_norm": 0.8780966177223058,
      "learning_rate": 2e-05,
      "loss": 0.6316461086273193,
      "step": 2580
    },
    {
      "epoch": 0.19682346682878638,
      "grad_norm": 0.8172704754373125,
      "learning_rate": 2e-05,
      "loss": 0.6350976467132569,
      "step": 2590
    },
    {
      "epoch": 0.1975834029941485,
      "grad_norm": 1.314600777113447,
      "learning_rate": 2e-05,
      "loss": 0.6386783599853516,
      "step": 2600
    },
    {
      "epoch": 0.1983433391595106,
      "grad_norm": 1.1204615377337357,
      "learning_rate": 2e-05,
      "loss": 0.6429242134094239,
      "step": 2610
    },
    {
      "epoch": 0.1991032753248727,
      "grad_norm": 0.9207404992707692,
      "learning_rate": 2e-05,
      "loss": 0.6100504875183106,
      "step": 2620
    },
    {
      "epoch": 0.1998632114902348,
      "grad_norm": 1.4479828471747938,
      "learning_rate": 2e-05,
      "loss": 0.6189369201660156,
      "step": 2630
    },
    {
      "epoch": 0.20062314765559694,
      "grad_norm": 1.6538730364392782,
      "learning_rate": 2e-05,
      "loss": 0.6408646583557129,
      "step": 2640
    },
    {
      "epoch": 0.20138308382095904,
      "grad_norm": 1.400833721181578,
      "learning_rate": 2e-05,
      "loss": 0.6331706523895264,
      "step": 2650
    },
    {
      "epoch": 0.20214301998632114,
      "grad_norm": 2.4651840710400217,
      "learning_rate": 2e-05,
      "loss": 0.6359334945678711,
      "step": 2660
    },
    {
      "epoch": 0.20290295615168327,
      "grad_norm": 2.8902390382766168,
      "learning_rate": 2e-05,
      "loss": 0.6367801666259766,
      "step": 2670
    },
    {
      "epoch": 0.20366289231704537,
      "grad_norm": 1.0725705407050916,
      "learning_rate": 2e-05,
      "loss": 0.6252540588378906,
      "step": 2680
    },
    {
      "epoch": 0.20442282848240748,
      "grad_norm": 1.453680714722172,
      "learning_rate": 2e-05,
      "loss": 0.6370233535766602,
      "step": 2690
    },
    {
      "epoch": 0.20518276464776958,
      "grad_norm": 1.1494128351878075,
      "learning_rate": 2e-05,
      "loss": 0.6241844177246094,
      "step": 2700
    },
    {
      "epoch": 0.2059427008131317,
      "grad_norm": 1.2561059487170374,
      "learning_rate": 2e-05,
      "loss": 0.6379550933837891,
      "step": 2710
    },
    {
      "epoch": 0.2067026369784938,
      "grad_norm": 1.1763018989953011,
      "learning_rate": 2e-05,
      "loss": 0.617769193649292,
      "step": 2720
    },
    {
      "epoch": 0.2074625731438559,
      "grad_norm": 0.8301947300077038,
      "learning_rate": 2e-05,
      "loss": 0.6172198295593262,
      "step": 2730
    },
    {
      "epoch": 0.208222509309218,
      "grad_norm": 1.2654029656743424,
      "learning_rate": 2e-05,
      "loss": 0.6206151008605957,
      "step": 2740
    },
    {
      "epoch": 0.20898244547458014,
      "grad_norm": 1.2906344649474475,
      "learning_rate": 2e-05,
      "loss": 0.6183318138122559,
      "step": 2750
    },
    {
      "epoch": 0.20974238163994224,
      "grad_norm": 19.702502388770796,
      "learning_rate": 2e-05,
      "loss": 0.6330374717712403,
      "step": 2760
    },
    {
      "epoch": 0.21050231780530435,
      "grad_norm": 2.896777724084538,
      "learning_rate": 2e-05,
      "loss": 0.6482915878295898,
      "step": 2770
    },
    {
      "epoch": 0.21126225397066647,
      "grad_norm": 1.0591159342855259,
      "learning_rate": 2e-05,
      "loss": 0.6418767452239991,
      "step": 2780
    },
    {
      "epoch": 0.21202219013602858,
      "grad_norm": 0.9814179371810936,
      "learning_rate": 2e-05,
      "loss": 0.6314129829406738,
      "step": 2790
    },
    {
      "epoch": 0.21278212630139068,
      "grad_norm": 0.978536412787398,
      "learning_rate": 2e-05,
      "loss": 0.6173480033874512,
      "step": 2800
    },
    {
      "epoch": 0.21354206246675278,
      "grad_norm": 1.0392932503872843,
      "learning_rate": 2e-05,
      "loss": 0.6070091247558593,
      "step": 2810
    },
    {
      "epoch": 0.2143019986321149,
      "grad_norm": 0.7630311281085816,
      "learning_rate": 2e-05,
      "loss": 0.603149127960205,
      "step": 2820
    },
    {
      "epoch": 0.215061934797477,
      "grad_norm": 0.7796521616886448,
      "learning_rate": 2e-05,
      "loss": 0.6206292629241943,
      "step": 2830
    },
    {
      "epoch": 0.2158218709628391,
      "grad_norm": 1.135805965040796,
      "learning_rate": 2e-05,
      "loss": 0.6249284267425537,
      "step": 2840
    },
    {
      "epoch": 0.21658180712820124,
      "grad_norm": 1.0034800711429899,
      "learning_rate": 2e-05,
      "loss": 0.6145776271820068,
      "step": 2850
    },
    {
      "epoch": 0.21734174329356334,
      "grad_norm": 1.038072409401356,
      "learning_rate": 2e-05,
      "loss": 0.6203277587890625,
      "step": 2860
    },
    {
      "epoch": 0.21810167945892545,
      "grad_norm": 1.586011294473222,
      "learning_rate": 2e-05,
      "loss": 0.6286195278167724,
      "step": 2870
    },
    {
      "epoch": 0.21886161562428755,
      "grad_norm": 0.8925062852923543,
      "learning_rate": 2e-05,
      "loss": 0.6153251647949218,
      "step": 2880
    },
    {
      "epoch": 0.21962155178964968,
      "grad_norm": 1.9996238470642247,
      "learning_rate": 2e-05,
      "loss": 0.6150319576263428,
      "step": 2890
    },
    {
      "epoch": 0.22038148795501178,
      "grad_norm": 1.851709788656795,
      "learning_rate": 2e-05,
      "loss": 0.6348749160766601,
      "step": 2900
    },
    {
      "epoch": 0.22114142412037388,
      "grad_norm": 1.5503903928626321,
      "learning_rate": 2e-05,
      "loss": 0.6155122280120849,
      "step": 2910
    },
    {
      "epoch": 0.221901360285736,
      "grad_norm": 1.6014915313981002,
      "learning_rate": 2e-05,
      "loss": 0.6487886428833007,
      "step": 2920
    },
    {
      "epoch": 0.2226612964510981,
      "grad_norm": 3.0888999458086306,
      "learning_rate": 2e-05,
      "loss": 0.6402599334716796,
      "step": 2930
    },
    {
      "epoch": 0.2234212326164602,
      "grad_norm": 2.256985072645217,
      "learning_rate": 2e-05,
      "loss": 0.639389181137085,
      "step": 2940
    },
    {
      "epoch": 0.22418116878182232,
      "grad_norm": 1.1953816080056705,
      "learning_rate": 2e-05,
      "loss": 0.6524765014648437,
      "step": 2950
    },
    {
      "epoch": 0.22494110494718444,
      "grad_norm": 1.4837984558035446,
      "learning_rate": 2e-05,
      "loss": 0.6412689208984375,
      "step": 2960
    },
    {
      "epoch": 0.22570104111254655,
      "grad_norm": 1.298620783185863,
      "learning_rate": 2e-05,
      "loss": 0.6097001552581787,
      "step": 2970
    },
    {
      "epoch": 0.22646097727790865,
      "grad_norm": 0.9186623742719742,
      "learning_rate": 2e-05,
      "loss": 0.6124137878417969,
      "step": 2980
    },
    {
      "epoch": 0.22722091344327078,
      "grad_norm": 1.2086327273935487,
      "learning_rate": 2e-05,
      "loss": 0.6018218994140625,
      "step": 2990
    },
    {
      "epoch": 0.22798084960863288,
      "grad_norm": 1.1774896919937214,
      "learning_rate": 2e-05,
      "loss": 0.6359123229980469,
      "step": 3000
    },
    {
      "epoch": 0.22874078577399498,
      "grad_norm": 0.8376140766835578,
      "learning_rate": 2e-05,
      "loss": 0.6134356498718262,
      "step": 3010
    },
    {
      "epoch": 0.22950072193935708,
      "grad_norm": 1.3695480635864736,
      "learning_rate": 2e-05,
      "loss": 0.6094162940979004,
      "step": 3020
    },
    {
      "epoch": 0.2302606581047192,
      "grad_norm": 8.759741635179273,
      "learning_rate": 2e-05,
      "loss": 0.6244755744934082,
      "step": 3030
    },
    {
      "epoch": 0.23102059427008131,
      "grad_norm": 4.401741469652162,
      "learning_rate": 2e-05,
      "loss": 0.6653237819671631,
      "step": 3040
    },
    {
      "epoch": 0.23178053043544342,
      "grad_norm": 0.9651591430969964,
      "learning_rate": 2e-05,
      "loss": 0.6163477897644043,
      "step": 3050
    },
    {
      "epoch": 0.23254046660080555,
      "grad_norm": 1.2540306441172315,
      "learning_rate": 2e-05,
      "loss": 0.6076150417327881,
      "step": 3060
    },
    {
      "epoch": 0.23330040276616765,
      "grad_norm": 1.512197160849794,
      "learning_rate": 2e-05,
      "loss": 0.6530462265014648,
      "step": 3070
    },
    {
      "epoch": 0.23406033893152975,
      "grad_norm": 1.0873268561701026,
      "learning_rate": 2e-05,
      "loss": 0.6455690383911132,
      "step": 3080
    },
    {
      "epoch": 0.23482027509689185,
      "grad_norm": 1.5428987528547071,
      "learning_rate": 2e-05,
      "loss": 0.6147666931152344,
      "step": 3090
    },
    {
      "epoch": 0.23558021126225398,
      "grad_norm": 2.8908752896198657,
      "learning_rate": 2e-05,
      "loss": 0.6317712306976319,
      "step": 3100
    },
    {
      "epoch": 0.23634014742761608,
      "grad_norm": 1.089006444137297,
      "learning_rate": 2e-05,
      "loss": 0.6073556423187256,
      "step": 3110
    },
    {
      "epoch": 0.23710008359297818,
      "grad_norm": 1.062047237532833,
      "learning_rate": 2e-05,
      "loss": 0.629581356048584,
      "step": 3120
    },
    {
      "epoch": 0.2378600197583403,
      "grad_norm": 1.0177972120401673,
      "learning_rate": 2e-05,
      "loss": 0.6323951721191406,
      "step": 3130
    },
    {
      "epoch": 0.23861995592370241,
      "grad_norm": 1.0040279876940803,
      "learning_rate": 2e-05,
      "loss": 0.611222219467163,
      "step": 3140
    },
    {
      "epoch": 0.23937989208906452,
      "grad_norm": 1.3386060294466204,
      "learning_rate": 2e-05,
      "loss": 0.6076837062835694,
      "step": 3150
    },
    {
      "epoch": 0.24013982825442662,
      "grad_norm": 1.003288225197739,
      "learning_rate": 2e-05,
      "loss": 0.6263534545898437,
      "step": 3160
    },
    {
      "epoch": 0.24089976441978875,
      "grad_norm": 0.9295245153210507,
      "learning_rate": 2e-05,
      "loss": 0.629744291305542,
      "step": 3170
    },
    {
      "epoch": 0.24165970058515085,
      "grad_norm": 1.3354273823166982,
      "learning_rate": 2e-05,
      "loss": 0.6224825859069825,
      "step": 3180
    },
    {
      "epoch": 0.24241963675051295,
      "grad_norm": 1.0760003107828335,
      "learning_rate": 2e-05,
      "loss": 0.6299269676208497,
      "step": 3190
    },
    {
      "epoch": 0.24317957291587505,
      "grad_norm": 0.941133634643217,
      "learning_rate": 2e-05,
      "loss": 0.6484769821166992,
      "step": 3200
    },
    {
      "epoch": 0.24393950908123718,
      "grad_norm": 1.3185461630351192,
      "learning_rate": 2e-05,
      "loss": 0.6362860679626465,
      "step": 3210
    },
    {
      "epoch": 0.24469944524659928,
      "grad_norm": 1.183339782363713,
      "learning_rate": 2e-05,
      "loss": 0.6190390586853027,
      "step": 3220
    },
    {
      "epoch": 0.24545938141196139,
      "grad_norm": 0.9387291080116852,
      "learning_rate": 2e-05,
      "loss": 0.6350534439086915,
      "step": 3230
    },
    {
      "epoch": 0.24621931757732352,
      "grad_norm": 2.310301884693289,
      "learning_rate": 2e-05,
      "loss": 0.6099423885345459,
      "step": 3240
    },
    {
      "epoch": 0.24697925374268562,
      "grad_norm": 4.51689150363215,
      "learning_rate": 2e-05,
      "loss": 0.6252852439880371,
      "step": 3250
    },
    {
      "epoch": 0.24773918990804772,
      "grad_norm": 0.9038004061763136,
      "learning_rate": 2e-05,
      "loss": 0.6356831550598144,
      "step": 3260
    },
    {
      "epoch": 0.24849912607340982,
      "grad_norm": 9.22641306881663,
      "learning_rate": 2e-05,
      "loss": 0.639389181137085,
      "step": 3270
    },
    {
      "epoch": 0.24925906223877195,
      "grad_norm": 1.4908438703959253,
      "learning_rate": 2e-05,
      "loss": 0.628369140625,
      "step": 3280
    },
    {
      "epoch": 0.25001899840413405,
      "grad_norm": 7.419626390823035,
      "learning_rate": 2e-05,
      "loss": 0.6123821258544921,
      "step": 3290
    },
    {
      "epoch": 0.2507789345694962,
      "grad_norm": 1.2786580240513115,
      "learning_rate": 2e-05,
      "loss": 0.6278958320617676,
      "step": 3300
    },
    {
      "epoch": 0.25153887073485826,
      "grad_norm": 1.0502472828946412,
      "learning_rate": 2e-05,
      "loss": 0.6296170711517334,
      "step": 3310
    },
    {
      "epoch": 0.2522988069002204,
      "grad_norm": 1.0918861973419551,
      "learning_rate": 2e-05,
      "loss": 0.6308129787445068,
      "step": 3320
    },
    {
      "epoch": 0.2530587430655825,
      "grad_norm": 1.1226136226715475,
      "learning_rate": 2e-05,
      "loss": 0.6414403438568115,
      "step": 3330
    },
    {
      "epoch": 0.2538186792309446,
      "grad_norm": 1.3250141910916213,
      "learning_rate": 2e-05,
      "loss": 0.6283860206604004,
      "step": 3340
    },
    {
      "epoch": 0.2545786153963067,
      "grad_norm": 1.3565252029111057,
      "learning_rate": 2e-05,
      "loss": 0.632968521118164,
      "step": 3350
    },
    {
      "epoch": 0.25533855156166885,
      "grad_norm": 1.128898434457114,
      "learning_rate": 2e-05,
      "loss": 0.6104358673095703,
      "step": 3360
    },
    {
      "epoch": 0.2560984877270309,
      "grad_norm": 1.6612843957044359,
      "learning_rate": 2e-05,
      "loss": 0.6286696434020996,
      "step": 3370
    },
    {
      "epoch": 0.25685842389239305,
      "grad_norm": 1.156000264796932,
      "learning_rate": 2e-05,
      "loss": 0.6091503143310547,
      "step": 3380
    },
    {
      "epoch": 0.2576183600577551,
      "grad_norm": 0.7528740182554841,
      "learning_rate": 2e-05,
      "loss": 0.6218111038208007,
      "step": 3390
    },
    {
      "epoch": 0.25837829622311725,
      "grad_norm": 0.9275869426296971,
      "learning_rate": 2e-05,
      "loss": 0.621043586730957,
      "step": 3400
    },
    {
      "epoch": 0.2591382323884794,
      "grad_norm": 1.0552086309608597,
      "learning_rate": 2e-05,
      "loss": 0.6073514461517334,
      "step": 3410
    },
    {
      "epoch": 0.25989816855384146,
      "grad_norm": 1.1460119713030243,
      "learning_rate": 2e-05,
      "loss": 0.6165019512176514,
      "step": 3420
    },
    {
      "epoch": 0.2606581047192036,
      "grad_norm": 1.0155774831457327,
      "learning_rate": 2e-05,
      "loss": 0.6184118270874024,
      "step": 3430
    },
    {
      "epoch": 0.2614180408845657,
      "grad_norm": 1.3610565099669534,
      "learning_rate": 2e-05,
      "loss": 0.5858828544616699,
      "step": 3440
    },
    {
      "epoch": 0.2621779770499278,
      "grad_norm": 1.6942539291614553,
      "learning_rate": 2e-05,
      "loss": 0.6389949798583985,
      "step": 3450
    },
    {
      "epoch": 0.2629379132152899,
      "grad_norm": 0.9432722353200557,
      "learning_rate": 2e-05,
      "loss": 0.634234619140625,
      "step": 3460
    },
    {
      "epoch": 0.26369784938065205,
      "grad_norm": 9.327485800969509,
      "learning_rate": 2e-05,
      "loss": 0.589467716217041,
      "step": 3470
    },
    {
      "epoch": 0.2644577855460141,
      "grad_norm": 3.071442209769249,
      "learning_rate": 2e-05,
      "loss": 0.6044052124023438,
      "step": 3480
    },
    {
      "epoch": 0.26521772171137625,
      "grad_norm": 1.2925607716528897,
      "learning_rate": 2e-05,
      "loss": 0.6095686912536621,
      "step": 3490
    },
    {
      "epoch": 0.2659776578767383,
      "grad_norm": 16.06253345369598,
      "learning_rate": 2e-05,
      "loss": 0.6121344566345215,
      "step": 3500
    },
    {
      "epoch": 0.26673759404210046,
      "grad_norm": 1.436188332832336,
      "learning_rate": 2e-05,
      "loss": 0.6014012336730957,
      "step": 3510
    },
    {
      "epoch": 0.2674975302074626,
      "grad_norm": 1.7102705577739652,
      "learning_rate": 2e-05,
      "loss": 0.6195626258850098,
      "step": 3520
    },
    {
      "epoch": 0.26825746637282466,
      "grad_norm": 1.4558527764040676,
      "learning_rate": 2e-05,
      "loss": 0.6225949764251709,
      "step": 3530
    },
    {
      "epoch": 0.2690174025381868,
      "grad_norm": 1.3080733031627572,
      "learning_rate": 2e-05,
      "loss": 0.6161701202392578,
      "step": 3540
    },
    {
      "epoch": 0.2697773387035489,
      "grad_norm": 1.1258102284127145,
      "learning_rate": 2e-05,
      "loss": 0.6220259666442871,
      "step": 3550
    },
    {
      "epoch": 0.270537274868911,
      "grad_norm": 1.2422284824681804,
      "learning_rate": 2e-05,
      "loss": 0.645357084274292,
      "step": 3560
    },
    {
      "epoch": 0.2712972110342731,
      "grad_norm": 1.2292013513428475,
      "learning_rate": 2e-05,
      "loss": 0.6096301078796387,
      "step": 3570
    },
    {
      "epoch": 0.27205714719963525,
      "grad_norm": 1.0205303407560335,
      "learning_rate": 2e-05,
      "loss": 0.6461244106292725,
      "step": 3580
    },
    {
      "epoch": 0.2728170833649973,
      "grad_norm": 2.6872881400099273,
      "learning_rate": 2e-05,
      "loss": 0.6358933925628663,
      "step": 3590
    },
    {
      "epoch": 0.27357701953035946,
      "grad_norm": 1.1556669884783608,
      "learning_rate": 2e-05,
      "loss": 0.6323175430297852,
      "step": 3600
    },
    {
      "epoch": 0.2743369556957216,
      "grad_norm": 3.314994024182842,
      "learning_rate": 2e-05,
      "loss": 0.5981500625610352,
      "step": 3610
    },
    {
      "epoch": 0.27509689186108366,
      "grad_norm": 1.5758389088455522,
      "learning_rate": 2e-05,
      "loss": 0.6199132442474365,
      "step": 3620
    },
    {
      "epoch": 0.2758568280264458,
      "grad_norm": 1.1523052929094744,
      "learning_rate": 2e-05,
      "loss": 0.6311004161834717,
      "step": 3630
    },
    {
      "epoch": 0.27661676419180786,
      "grad_norm": 4.209511583399702,
      "learning_rate": 2e-05,
      "loss": 0.6306180477142334,
      "step": 3640
    },
    {
      "epoch": 0.27737670035717,
      "grad_norm": 5.9265607841494905,
      "learning_rate": 2e-05,
      "loss": 0.614080810546875,
      "step": 3650
    },
    {
      "epoch": 0.2781366365225321,
      "grad_norm": 1.339276521039795,
      "learning_rate": 2e-05,
      "loss": 0.6019871711730957,
      "step": 3660
    },
    {
      "epoch": 0.2788965726878942,
      "grad_norm": 1.4009250479535327,
      "learning_rate": 2e-05,
      "loss": 0.6156785011291503,
      "step": 3670
    },
    {
      "epoch": 0.2796565088532563,
      "grad_norm": 1.8701775141946877,
      "learning_rate": 2e-05,
      "loss": 0.6056217193603516,
      "step": 3680
    },
    {
      "epoch": 0.28041644501861845,
      "grad_norm": 1.90893727770188,
      "learning_rate": 2e-05,
      "loss": 0.6284603118896485,
      "step": 3690
    },
    {
      "epoch": 0.28117638118398053,
      "grad_norm": 2.744963869747775,
      "learning_rate": 2e-05,
      "loss": 0.6229351043701172,
      "step": 3700
    },
    {
      "epoch": 0.28193631734934266,
      "grad_norm": 1.18153979601221,
      "learning_rate": 2e-05,
      "loss": 0.6146137237548828,
      "step": 3710
    },
    {
      "epoch": 0.2826962535147048,
      "grad_norm": 1.2996533180738103,
      "learning_rate": 2e-05,
      "loss": 0.6364758491516114,
      "step": 3720
    },
    {
      "epoch": 0.28345618968006686,
      "grad_norm": 1.158844048871961,
      "learning_rate": 2e-05,
      "loss": 0.5962699890136719,
      "step": 3730
    },
    {
      "epoch": 0.284216125845429,
      "grad_norm": 1.2216472402171064,
      "learning_rate": 2e-05,
      "loss": 0.6206924438476562,
      "step": 3740
    },
    {
      "epoch": 0.2849760620107911,
      "grad_norm": 0.890679821509559,
      "learning_rate": 2e-05,
      "loss": 0.6205419540405274,
      "step": 3750
    },
    {
      "epoch": 0.2857359981761532,
      "grad_norm": 1.2912051550907495,
      "learning_rate": 2e-05,
      "loss": 0.624941635131836,
      "step": 3760
    },
    {
      "epoch": 0.2864959343415153,
      "grad_norm": 1.0646251600564485,
      "learning_rate": 2e-05,
      "loss": 0.6251903533935547,
      "step": 3770
    },
    {
      "epoch": 0.2872558705068774,
      "grad_norm": 1.4325260373232669,
      "learning_rate": 2e-05,
      "loss": 0.6093158721923828,
      "step": 3780
    },
    {
      "epoch": 0.2880158066722395,
      "grad_norm": 1.8943205464020347,
      "learning_rate": 2e-05,
      "loss": 0.6362890243530274,
      "step": 3790
    },
    {
      "epoch": 0.28877574283760166,
      "grad_norm": 2.3208592609691543,
      "learning_rate": 2e-05,
      "loss": 0.6275027751922607,
      "step": 3800
    },
    {
      "epoch": 0.28953567900296373,
      "grad_norm": 2.1699449912440745,
      "learning_rate": 2e-05,
      "loss": 0.6083850860595703,
      "step": 3810
    },
    {
      "epoch": 0.29029561516832586,
      "grad_norm": 0.9942819199698443,
      "learning_rate": 2e-05,
      "loss": 0.6335760354995728,
      "step": 3820
    },
    {
      "epoch": 0.291055551333688,
      "grad_norm": 2.1188962468497134,
      "learning_rate": 2e-05,
      "loss": 0.6074498653411865,
      "step": 3830
    },
    {
      "epoch": 0.29181548749905006,
      "grad_norm": 1.1946221873993108,
      "learning_rate": 2e-05,
      "loss": 0.6301320552825928,
      "step": 3840
    },
    {
      "epoch": 0.2925754236644122,
      "grad_norm": 1.4654668135262237,
      "learning_rate": 2e-05,
      "loss": 0.615028190612793,
      "step": 3850
    },
    {
      "epoch": 0.2933353598297743,
      "grad_norm": 0.9704782828686308,
      "learning_rate": 2e-05,
      "loss": 0.6268731594085694,
      "step": 3860
    },
    {
      "epoch": 0.2940952959951364,
      "grad_norm": 0.973881954733979,
      "learning_rate": 2e-05,
      "loss": 0.622544240951538,
      "step": 3870
    },
    {
      "epoch": 0.2948552321604985,
      "grad_norm": 0.9862174972117107,
      "learning_rate": 2e-05,
      "loss": 0.6262887477874756,
      "step": 3880
    },
    {
      "epoch": 0.2956151683258606,
      "grad_norm": 1.8185012915464214,
      "learning_rate": 2e-05,
      "loss": 0.6259152889251709,
      "step": 3890
    },
    {
      "epoch": 0.29637510449122273,
      "grad_norm": 0.9624840612294071,
      "learning_rate": 2e-05,
      "loss": 0.6382728099822998,
      "step": 3900
    },
    {
      "epoch": 0.29713504065658486,
      "grad_norm": 1.27087177885983,
      "learning_rate": 2e-05,
      "loss": 0.6182097911834716,
      "step": 3910
    },
    {
      "epoch": 0.29789497682194693,
      "grad_norm": 2.3890028433012644,
      "learning_rate": 2e-05,
      "loss": 0.6197179317474365,
      "step": 3920
    },
    {
      "epoch": 0.29865491298730906,
      "grad_norm": 1.3648910447241072,
      "learning_rate": 2e-05,
      "loss": 0.6085414409637451,
      "step": 3930
    },
    {
      "epoch": 0.2994148491526712,
      "grad_norm": 1.3264500482112846,
      "learning_rate": 2e-05,
      "loss": 0.6149806022644043,
      "step": 3940
    },
    {
      "epoch": 0.30017478531803327,
      "grad_norm": 1.3200962604276947,
      "learning_rate": 2e-05,
      "loss": 0.6308249473571778,
      "step": 3950
    },
    {
      "epoch": 0.3009347214833954,
      "grad_norm": 0.8595307476045679,
      "learning_rate": 2e-05,
      "loss": 0.6181689262390136,
      "step": 3960
    },
    {
      "epoch": 0.3016946576487575,
      "grad_norm": 14.901845615980008,
      "learning_rate": 2e-05,
      "loss": 0.6143277168273926,
      "step": 3970
    },
    {
      "epoch": 0.3024545938141196,
      "grad_norm": 4.449595695317806,
      "learning_rate": 2e-05,
      "loss": 0.6383512020111084,
      "step": 3980
    },
    {
      "epoch": 0.30321452997948173,
      "grad_norm": 1.127083082452341,
      "learning_rate": 2e-05,
      "loss": 0.6203843116760254,
      "step": 3990
    },
    {
      "epoch": 0.30397446614484386,
      "grad_norm": 0.8860818290392859,
      "learning_rate": 2e-05,
      "loss": 0.6098330974578857,
      "step": 4000
    },
    {
      "epoch": 0.30473440231020593,
      "grad_norm": 1.0437370685884855,
      "learning_rate": 2e-05,
      "loss": 0.6101393699645996,
      "step": 4010
    },
    {
      "epoch": 0.30549433847556806,
      "grad_norm": 1.278469365345407,
      "learning_rate": 2e-05,
      "loss": 0.6249421119689942,
      "step": 4020
    },
    {
      "epoch": 0.30625427464093014,
      "grad_norm": 1.4953459408550605,
      "learning_rate": 2e-05,
      "loss": 0.6222182273864746,
      "step": 4030
    },
    {
      "epoch": 0.30701421080629226,
      "grad_norm": 3.144699685900144,
      "learning_rate": 2e-05,
      "loss": 0.6221820831298828,
      "step": 4040
    },
    {
      "epoch": 0.3077741469716544,
      "grad_norm": 1.6854712024923644,
      "learning_rate": 2e-05,
      "loss": 0.6161329269409179,
      "step": 4050
    },
    {
      "epoch": 0.30853408313701647,
      "grad_norm": 2.3151959150874766,
      "learning_rate": 2e-05,
      "loss": 0.6204587936401367,
      "step": 4060
    },
    {
      "epoch": 0.3092940193023786,
      "grad_norm": 1.4811534982681576,
      "learning_rate": 2e-05,
      "loss": 0.6056510925292968,
      "step": 4070
    },
    {
      "epoch": 0.3100539554677407,
      "grad_norm": 1.0339459517187002,
      "learning_rate": 2e-05,
      "loss": 0.6259673595428467,
      "step": 4080
    },
    {
      "epoch": 0.3108138916331028,
      "grad_norm": 1.2235117780386027,
      "learning_rate": 2e-05,
      "loss": 0.6084205627441406,
      "step": 4090
    },
    {
      "epoch": 0.31157382779846493,
      "grad_norm": 1.7393673646949717,
      "learning_rate": 2e-05,
      "loss": 0.6320530891418457,
      "step": 4100
    },
    {
      "epoch": 0.31233376396382706,
      "grad_norm": 0.9030435760410976,
      "learning_rate": 2e-05,
      "loss": 0.6125435829162598,
      "step": 4110
    },
    {
      "epoch": 0.31309370012918913,
      "grad_norm": 0.9770233863461716,
      "learning_rate": 2e-05,
      "loss": 0.6324932098388671,
      "step": 4120
    },
    {
      "epoch": 0.31385363629455126,
      "grad_norm": 11.164786625545675,
      "learning_rate": 2e-05,
      "loss": 0.6129942893981933,
      "step": 4130
    },
    {
      "epoch": 0.3146135724599134,
      "grad_norm": 1.484858734640339,
      "learning_rate": 2e-05,
      "loss": 0.6236273765563964,
      "step": 4140
    },
    {
      "epoch": 0.31537350862527547,
      "grad_norm": 1.0150188901770478,
      "learning_rate": 2e-05,
      "loss": 0.6191964626312256,
      "step": 4150
    },
    {
      "epoch": 0.3161334447906376,
      "grad_norm": 0.734050746516317,
      "learning_rate": 2e-05,
      "loss": 0.6276535987854004,
      "step": 4160
    },
    {
      "epoch": 0.31689338095599967,
      "grad_norm": 1.4167390319947213,
      "learning_rate": 2e-05,
      "loss": 0.6143552780151367,
      "step": 4170
    },
    {
      "epoch": 0.3176533171213618,
      "grad_norm": 6.988334362027056,
      "learning_rate": 2e-05,
      "loss": 0.595700740814209,
      "step": 4180
    },
    {
      "epoch": 0.31841325328672393,
      "grad_norm": 1.044747728571304,
      "learning_rate": 2e-05,
      "loss": 0.6026046752929688,
      "step": 4190
    },
    {
      "epoch": 0.319173189452086,
      "grad_norm": 1.5761166174685894,
      "learning_rate": 2e-05,
      "loss": 0.6220703125,
      "step": 4200
    },
    {
      "epoch": 0.31993312561744813,
      "grad_norm": 1.368582543235248,
      "learning_rate": 2e-05,
      "loss": 0.6257802009582519,
      "step": 4210
    },
    {
      "epoch": 0.32069306178281026,
      "grad_norm": 0.8593932041129811,
      "learning_rate": 2e-05,
      "loss": 0.6099929809570312,
      "step": 4220
    },
    {
      "epoch": 0.32145299794817234,
      "grad_norm": 2.866418221741655,
      "learning_rate": 2e-05,
      "loss": 0.6190606117248535,
      "step": 4230
    },
    {
      "epoch": 0.32221293411353447,
      "grad_norm": 1.7731877645586673,
      "learning_rate": 2e-05,
      "loss": 0.6337327480316162,
      "step": 4240
    },
    {
      "epoch": 0.3229728702788966,
      "grad_norm": 0.8654444995328492,
      "learning_rate": 2e-05,
      "loss": 0.6130261898040772,
      "step": 4250
    },
    {
      "epoch": 0.32373280644425867,
      "grad_norm": 2.6189925006836585,
      "learning_rate": 2e-05,
      "loss": 0.6288496971130371,
      "step": 4260
    },
    {
      "epoch": 0.3244927426096208,
      "grad_norm": 1.3466638193489775,
      "learning_rate": 2e-05,
      "loss": 0.6257670402526856,
      "step": 4270
    },
    {
      "epoch": 0.32525267877498293,
      "grad_norm": 1.013486446763607,
      "learning_rate": 2e-05,
      "loss": 0.6344550609588623,
      "step": 4280
    },
    {
      "epoch": 0.326012614940345,
      "grad_norm": 1.1856650844101666,
      "learning_rate": 2e-05,
      "loss": 0.6028676509857178,
      "step": 4290
    },
    {
      "epoch": 0.32677255110570713,
      "grad_norm": 1.021941635755477,
      "learning_rate": 2e-05,
      "loss": 0.6205586433410645,
      "step": 4300
    },
    {
      "epoch": 0.3275324872710692,
      "grad_norm": 1.0933995287761475,
      "learning_rate": 2e-05,
      "loss": 0.6085476875305176,
      "step": 4310
    },
    {
      "epoch": 0.32829242343643134,
      "grad_norm": 1.9004377903753171,
      "learning_rate": 2e-05,
      "loss": 0.6211337089538574,
      "step": 4320
    },
    {
      "epoch": 0.32905235960179346,
      "grad_norm": 2.530273155001998,
      "learning_rate": 2e-05,
      "loss": 0.6182102203369141,
      "step": 4330
    },
    {
      "epoch": 0.32981229576715554,
      "grad_norm": 1.7906530704493195,
      "learning_rate": 2e-05,
      "loss": 0.6185105800628662,
      "step": 4340
    },
    {
      "epoch": 0.33057223193251767,
      "grad_norm": 1.670584019028632,
      "learning_rate": 2e-05,
      "loss": 0.6119475364685059,
      "step": 4350
    },
    {
      "epoch": 0.3313321680978798,
      "grad_norm": 1.5455256174452385,
      "learning_rate": 2e-05,
      "loss": 0.6008148193359375,
      "step": 4360
    },
    {
      "epoch": 0.33209210426324187,
      "grad_norm": 3.208457109843326,
      "learning_rate": 2e-05,
      "loss": 0.6103523254394532,
      "step": 4370
    },
    {
      "epoch": 0.332852040428604,
      "grad_norm": 4.734555401179033,
      "learning_rate": 2e-05,
      "loss": 0.6435708045959473,
      "step": 4380
    },
    {
      "epoch": 0.33361197659396613,
      "grad_norm": 0.9834564438968257,
      "learning_rate": 2e-05,
      "loss": 0.613013744354248,
      "step": 4390
    },
    {
      "epoch": 0.3343719127593282,
      "grad_norm": 1.5899940571927773,
      "learning_rate": 2e-05,
      "loss": 0.6370656490325928,
      "step": 4400
    },
    {
      "epoch": 0.33513184892469033,
      "grad_norm": 0.9237244721295171,
      "learning_rate": 2e-05,
      "loss": 0.6245400905609131,
      "step": 4410
    },
    {
      "epoch": 0.3358917850900524,
      "grad_norm": 0.9276064643002765,
      "learning_rate": 2e-05,
      "loss": 0.6174593925476074,
      "step": 4420
    },
    {
      "epoch": 0.33665172125541454,
      "grad_norm": 2.8272846058313927,
      "learning_rate": 2e-05,
      "loss": 0.6339360237121582,
      "step": 4430
    },
    {
      "epoch": 0.33741165742077667,
      "grad_norm": 0.8986064383731139,
      "learning_rate": 2e-05,
      "loss": 0.597989559173584,
      "step": 4440
    },
    {
      "epoch": 0.33817159358613874,
      "grad_norm": 0.8793590754097262,
      "learning_rate": 2e-05,
      "loss": 0.6045150756835938,
      "step": 4450
    },
    {
      "epoch": 0.33893152975150087,
      "grad_norm": 1.1779563215018218,
      "learning_rate": 2e-05,
      "loss": 0.6187587261199952,
      "step": 4460
    },
    {
      "epoch": 0.339691465916863,
      "grad_norm": 1.3468609346078841,
      "learning_rate": 2e-05,
      "loss": 0.6251970291137695,
      "step": 4470
    },
    {
      "epoch": 0.3404514020822251,
      "grad_norm": 0.8988935942814856,
      "learning_rate": 2e-05,
      "loss": 0.597871971130371,
      "step": 4480
    },
    {
      "epoch": 0.3412113382475872,
      "grad_norm": 0.8607023127414716,
      "learning_rate": 2e-05,
      "loss": 0.6100861549377441,
      "step": 4490
    },
    {
      "epoch": 0.34197127441294933,
      "grad_norm": 0.8479817886620996,
      "learning_rate": 2e-05,
      "loss": 0.6026260375976562,
      "step": 4500
    },
    {
      "epoch": 0.3427312105783114,
      "grad_norm": 0.907490203208679,
      "learning_rate": 2e-05,
      "loss": 0.6259716033935547,
      "step": 4510
    },
    {
      "epoch": 0.34349114674367354,
      "grad_norm": 0.9683122114373157,
      "learning_rate": 2e-05,
      "loss": 0.6468429088592529,
      "step": 4520
    },
    {
      "epoch": 0.34425108290903567,
      "grad_norm": 0.9267267803730356,
      "learning_rate": 2e-05,
      "loss": 0.625063419342041,
      "step": 4530
    },
    {
      "epoch": 0.34501101907439774,
      "grad_norm": 1.8581451268995983,
      "learning_rate": 2e-05,
      "loss": 0.6267961502075196,
      "step": 4540
    },
    {
      "epoch": 0.34577095523975987,
      "grad_norm": 1.1135865038363975,
      "learning_rate": 2e-05,
      "loss": 0.6432443618774414,
      "step": 4550
    },
    {
      "epoch": 0.34653089140512194,
      "grad_norm": 1.1244838971476872,
      "learning_rate": 2e-05,
      "loss": 0.6133568763732911,
      "step": 4560
    },
    {
      "epoch": 0.3472908275704841,
      "grad_norm": 3.4229698437830103,
      "learning_rate": 2e-05,
      "loss": 0.602759838104248,
      "step": 4570
    },
    {
      "epoch": 0.3480507637358462,
      "grad_norm": 1.7669657715748632,
      "learning_rate": 2e-05,
      "loss": 0.6133023262023926,
      "step": 4580
    },
    {
      "epoch": 0.3488106999012083,
      "grad_norm": 0.9218128006407286,
      "learning_rate": 2e-05,
      "loss": 0.6523171424865722,
      "step": 4590
    },
    {
      "epoch": 0.3495706360665704,
      "grad_norm": 0.8974549965072942,
      "learning_rate": 2e-05,
      "loss": 0.6212499141693115,
      "step": 4600
    },
    {
      "epoch": 0.35033057223193254,
      "grad_norm": 1.1645952891980642,
      "learning_rate": 2e-05,
      "loss": 0.6346889495849609,
      "step": 4610
    },
    {
      "epoch": 0.3510905083972946,
      "grad_norm": 1.4040413653568118,
      "learning_rate": 2e-05,
      "loss": 0.6068562507629395,
      "step": 4620
    },
    {
      "epoch": 0.35185044456265674,
      "grad_norm": 1.4596732898533262,
      "learning_rate": 2e-05,
      "loss": 0.6284614562988281,
      "step": 4630
    },
    {
      "epoch": 0.35261038072801887,
      "grad_norm": 1.0155555804241654,
      "learning_rate": 2e-05,
      "loss": 0.6058677673339844,
      "step": 4640
    },
    {
      "epoch": 0.35337031689338094,
      "grad_norm": 1.2090689072922334,
      "learning_rate": 2e-05,
      "loss": 0.6285273551940918,
      "step": 4650
    },
    {
      "epoch": 0.35413025305874307,
      "grad_norm": 1.2787108770004478,
      "learning_rate": 2e-05,
      "loss": 0.6266864776611328,
      "step": 4660
    },
    {
      "epoch": 0.3548901892241052,
      "grad_norm": 0.9197107300964413,
      "learning_rate": 2e-05,
      "loss": 0.6152817726135253,
      "step": 4670
    },
    {
      "epoch": 0.3556501253894673,
      "grad_norm": 18.919166842611308,
      "learning_rate": 2e-05,
      "loss": 0.6119517326354981,
      "step": 4680
    },
    {
      "epoch": 0.3564100615548294,
      "grad_norm": 1.4058008566476412,
      "learning_rate": 2e-05,
      "loss": 0.6174506187438965,
      "step": 4690
    },
    {
      "epoch": 0.3571699977201915,
      "grad_norm": 0.9988437823137366,
      "learning_rate": 2e-05,
      "loss": 0.6167577266693115,
      "step": 4700
    },
    {
      "epoch": 0.3579299338855536,
      "grad_norm": 1.1556806181787362,
      "learning_rate": 2e-05,
      "loss": 0.6333663940429688,
      "step": 4710
    },
    {
      "epoch": 0.35868987005091574,
      "grad_norm": 1.1098292654559276,
      "learning_rate": 2e-05,
      "loss": 0.617949104309082,
      "step": 4720
    },
    {
      "epoch": 0.3594498062162778,
      "grad_norm": 1.9299990273309398,
      "learning_rate": 2e-05,
      "loss": 0.6180627346038818,
      "step": 4730
    },
    {
      "epoch": 0.36020974238163994,
      "grad_norm": 1.554199460054541,
      "learning_rate": 2e-05,
      "loss": 0.6386888504028321,
      "step": 4740
    },
    {
      "epoch": 0.36096967854700207,
      "grad_norm": 1.082013002346371,
      "learning_rate": 2e-05,
      "loss": 0.6285783767700195,
      "step": 4750
    },
    {
      "epoch": 0.36172961471236414,
      "grad_norm": 1.0738717257172536,
      "learning_rate": 2e-05,
      "loss": 0.626548957824707,
      "step": 4760
    },
    {
      "epoch": 0.3624895508777263,
      "grad_norm": 1.3833397306073538,
      "learning_rate": 2e-05,
      "loss": 0.6229054450988769,
      "step": 4770
    },
    {
      "epoch": 0.3632494870430884,
      "grad_norm": 0.8969551611024454,
      "learning_rate": 2e-05,
      "loss": 0.6218652248382568,
      "step": 4780
    },
    {
      "epoch": 0.3640094232084505,
      "grad_norm": 0.8497565266191878,
      "learning_rate": 2e-05,
      "loss": 0.6254983425140381,
      "step": 4790
    },
    {
      "epoch": 0.3647693593738126,
      "grad_norm": 0.8030680111207589,
      "learning_rate": 2e-05,
      "loss": 0.5967194557189941,
      "step": 4800
    },
    {
      "epoch": 0.36552929553917474,
      "grad_norm": 2.3725041063513617,
      "learning_rate": 2e-05,
      "loss": 0.6141952514648438,
      "step": 4810
    },
    {
      "epoch": 0.3662892317045368,
      "grad_norm": 1.079090175426383,
      "learning_rate": 2e-05,
      "loss": 0.6035738945007324,
      "step": 4820
    },
    {
      "epoch": 0.36704916786989894,
      "grad_norm": 0.7350181579763702,
      "learning_rate": 2e-05,
      "loss": 0.5995299339294433,
      "step": 4830
    },
    {
      "epoch": 0.367809104035261,
      "grad_norm": 0.6998723708900272,
      "learning_rate": 2e-05,
      "loss": 0.6088802337646484,
      "step": 4840
    },
    {
      "epoch": 0.36856904020062314,
      "grad_norm": 4.656688077966762,
      "learning_rate": 2e-05,
      "loss": 0.6377917766571045,
      "step": 4850
    },
    {
      "epoch": 0.3693289763659853,
      "grad_norm": 0.9362102615883362,
      "learning_rate": 2e-05,
      "loss": 0.5998881816864013,
      "step": 4860
    },
    {
      "epoch": 0.37008891253134735,
      "grad_norm": 0.7777138582558922,
      "learning_rate": 2e-05,
      "loss": 0.6036483764648437,
      "step": 4870
    },
    {
      "epoch": 0.3708488486967095,
      "grad_norm": 0.7566663050177238,
      "learning_rate": 2e-05,
      "loss": 0.5968652725219726,
      "step": 4880
    },
    {
      "epoch": 0.3716087848620716,
      "grad_norm": 0.814818990477887,
      "learning_rate": 2e-05,
      "loss": 0.6251842498779296,
      "step": 4890
    },
    {
      "epoch": 0.3723687210274337,
      "grad_norm": 0.7635616353510719,
      "learning_rate": 2e-05,
      "loss": 0.619797658920288,
      "step": 4900
    },
    {
      "epoch": 0.3731286571927958,
      "grad_norm": 0.7536889302602852,
      "learning_rate": 2e-05,
      "loss": 0.6170392513275147,
      "step": 4910
    },
    {
      "epoch": 0.37388859335815794,
      "grad_norm": 0.9099848828220006,
      "learning_rate": 2e-05,
      "loss": 0.6249567985534668,
      "step": 4920
    },
    {
      "epoch": 0.37464852952352,
      "grad_norm": 15.041580861087684,
      "learning_rate": 2e-05,
      "loss": 0.6162414073944091,
      "step": 4930
    },
    {
      "epoch": 0.37540846568888214,
      "grad_norm": 0.8261042579927552,
      "learning_rate": 2e-05,
      "loss": 0.6113101005554199,
      "step": 4940
    },
    {
      "epoch": 0.3761684018542442,
      "grad_norm": 1.0699305038057592,
      "learning_rate": 2e-05,
      "loss": 0.6028437614440918,
      "step": 4950
    },
    {
      "epoch": 0.37692833801960635,
      "grad_norm": 0.7785165124192742,
      "learning_rate": 2e-05,
      "loss": 0.614793348312378,
      "step": 4960
    },
    {
      "epoch": 0.3776882741849685,
      "grad_norm": 0.8881799045536933,
      "learning_rate": 2e-05,
      "loss": 0.6225779533386231,
      "step": 4970
    },
    {
      "epoch": 0.37844821035033055,
      "grad_norm": 1.4278565114637936,
      "learning_rate": 2e-05,
      "loss": 0.5902219295501709,
      "step": 4980
    },
    {
      "epoch": 0.3792081465156927,
      "grad_norm": 0.8827177845448362,
      "learning_rate": 2e-05,
      "loss": 0.5980781078338623,
      "step": 4990
    },
    {
      "epoch": 0.3799680826810548,
      "grad_norm": 1.0285813124515226,
      "learning_rate": 2e-05,
      "loss": 0.6059083461761474,
      "step": 5000
    },
    {
      "epoch": 0.3807280188464169,
      "grad_norm": 0.820793577407267,
      "learning_rate": 2e-05,
      "loss": 0.6184280395507813,
      "step": 5010
    },
    {
      "epoch": 0.381487955011779,
      "grad_norm": 10.840810148333725,
      "learning_rate": 2e-05,
      "loss": 0.6353204727172852,
      "step": 5020
    },
    {
      "epoch": 0.38224789117714114,
      "grad_norm": 0.8371736556282744,
      "learning_rate": 2e-05,
      "loss": 0.6085888862609863,
      "step": 5030
    },
    {
      "epoch": 0.3830078273425032,
      "grad_norm": 2.557319897319922,
      "learning_rate": 2e-05,
      "loss": 0.6260895729064941,
      "step": 5040
    },
    {
      "epoch": 0.38376776350786534,
      "grad_norm": 0.8642101569233044,
      "learning_rate": 2e-05,
      "loss": 0.6153550148010254,
      "step": 5050
    },
    {
      "epoch": 0.3845276996732275,
      "grad_norm": 1.1032686543382764,
      "learning_rate": 2e-05,
      "loss": 0.6046747207641602,
      "step": 5060
    },
    {
      "epoch": 0.38528763583858955,
      "grad_norm": 0.9365073025939038,
      "learning_rate": 2e-05,
      "loss": 0.5926566123962402,
      "step": 5070
    },
    {
      "epoch": 0.3860475720039517,
      "grad_norm": 1.0246404884418376,
      "learning_rate": 2e-05,
      "loss": 0.6223001480102539,
      "step": 5080
    },
    {
      "epoch": 0.38680750816931375,
      "grad_norm": 0.9300482279143104,
      "learning_rate": 2e-05,
      "loss": 0.5798623561859131,
      "step": 5090
    },
    {
      "epoch": 0.3875674443346759,
      "grad_norm": 1.6053842162544103,
      "learning_rate": 2e-05,
      "loss": 0.6029841423034668,
      "step": 5100
    },
    {
      "epoch": 0.388327380500038,
      "grad_norm": 1.236420346711475,
      "learning_rate": 2e-05,
      "loss": 0.5951147079467773,
      "step": 5110
    },
    {
      "epoch": 0.3890873166654001,
      "grad_norm": 1.4001539643380458,
      "learning_rate": 2e-05,
      "loss": 0.5764058589935303,
      "step": 5120
    },
    {
      "epoch": 0.3898472528307622,
      "grad_norm": 1.8221497895350538,
      "learning_rate": 2e-05,
      "loss": 0.6186111450195313,
      "step": 5130
    },
    {
      "epoch": 0.39060718899612434,
      "grad_norm": 1.4905797767085673,
      "learning_rate": 2e-05,
      "loss": 0.6186910152435303,
      "step": 5140
    },
    {
      "epoch": 0.3913671251614864,
      "grad_norm": 2.3264199588543004,
      "learning_rate": 2e-05,
      "loss": 0.6149563789367676,
      "step": 5150
    },
    {
      "epoch": 0.39212706132684855,
      "grad_norm": 0.7422649442017316,
      "learning_rate": 2e-05,
      "loss": 0.5995307922363281,
      "step": 5160
    },
    {
      "epoch": 0.3928869974922107,
      "grad_norm": 3.5492343837754508,
      "learning_rate": 2e-05,
      "loss": 0.6142280578613282,
      "step": 5170
    },
    {
      "epoch": 0.39364693365757275,
      "grad_norm": 1.3455496627090289,
      "learning_rate": 2e-05,
      "loss": 0.6256613731384277,
      "step": 5180
    },
    {
      "epoch": 0.3944068698229349,
      "grad_norm": 1.5044323702693636,
      "learning_rate": 2e-05,
      "loss": 0.5970637321472168,
      "step": 5190
    },
    {
      "epoch": 0.395166805988297,
      "grad_norm": 1.7044817996926638,
      "learning_rate": 2e-05,
      "loss": 0.604188346862793,
      "step": 5200
    },
    {
      "epoch": 0.3959267421536591,
      "grad_norm": 117.91927987661948,
      "learning_rate": 2e-05,
      "loss": 0.6106534004211426,
      "step": 5210
    },
    {
      "epoch": 0.3966866783190212,
      "grad_norm": 2.0341619973524367,
      "learning_rate": 2e-05,
      "loss": 0.5990291595458984,
      "step": 5220
    },
    {
      "epoch": 0.3974466144843833,
      "grad_norm": 2.0464111527082505,
      "learning_rate": 2e-05,
      "loss": 0.6054695129394532,
      "step": 5230
    },
    {
      "epoch": 0.3982065506497454,
      "grad_norm": 2.572790781277254,
      "learning_rate": 2e-05,
      "loss": 0.5982469081878662,
      "step": 5240
    },
    {
      "epoch": 0.39896648681510755,
      "grad_norm": 1.512942231446957,
      "learning_rate": 2e-05,
      "loss": 0.6332793235778809,
      "step": 5250
    },
    {
      "epoch": 0.3997264229804696,
      "grad_norm": 1.0251175379974675,
      "learning_rate": 2e-05,
      "loss": 0.5920603752136231,
      "step": 5260
    },
    {
      "epoch": 0.40048635914583175,
      "grad_norm": 1.1416793906052707,
      "learning_rate": 2e-05,
      "loss": 0.5849504470825195,
      "step": 5270
    },
    {
      "epoch": 0.4012462953111939,
      "grad_norm": 1.6169523722952144,
      "learning_rate": 2e-05,
      "loss": 0.6397748947143554,
      "step": 5280
    },
    {
      "epoch": 0.40200623147655595,
      "grad_norm": 1.120114737619909,
      "learning_rate": 2e-05,
      "loss": 0.6151798248291016,
      "step": 5290
    },
    {
      "epoch": 0.4027661676419181,
      "grad_norm": 0.8400142838450461,
      "learning_rate": 2e-05,
      "loss": 0.5935912132263184,
      "step": 5300
    },
    {
      "epoch": 0.4035261038072802,
      "grad_norm": 0.7422192477912698,
      "learning_rate": 2e-05,
      "loss": 0.5959334373474121,
      "step": 5310
    },
    {
      "epoch": 0.4042860399726423,
      "grad_norm": 0.7288472129075552,
      "learning_rate": 2e-05,
      "loss": 0.6160897254943848,
      "step": 5320
    },
    {
      "epoch": 0.4050459761380044,
      "grad_norm": 2.2124510762939003,
      "learning_rate": 2e-05,
      "loss": 0.6087497711181641,
      "step": 5330
    },
    {
      "epoch": 0.40580591230336654,
      "grad_norm": 0.989882418515224,
      "learning_rate": 2e-05,
      "loss": 0.6072328090667725,
      "step": 5340
    },
    {
      "epoch": 0.4065658484687286,
      "grad_norm": 0.748453269859091,
      "learning_rate": 2e-05,
      "loss": 0.618111515045166,
      "step": 5350
    },
    {
      "epoch": 0.40732578463409075,
      "grad_norm": 0.8540524008823654,
      "learning_rate": 2e-05,
      "loss": 0.6110595703125,
      "step": 5360
    },
    {
      "epoch": 0.4080857207994528,
      "grad_norm": 0.6736764459896096,
      "learning_rate": 2e-05,
      "loss": 0.6241538047790527,
      "step": 5370
    },
    {
      "epoch": 0.40884565696481495,
      "grad_norm": 0.7466014317361941,
      "learning_rate": 2e-05,
      "loss": 0.6138930320739746,
      "step": 5380
    },
    {
      "epoch": 0.4096055931301771,
      "grad_norm": 0.8819126088804886,
      "learning_rate": 2e-05,
      "loss": 0.6019068717956543,
      "step": 5390
    },
    {
      "epoch": 0.41036552929553916,
      "grad_norm": 0.9004962218204462,
      "learning_rate": 2e-05,
      "loss": 0.6338000297546387,
      "step": 5400
    },
    {
      "epoch": 0.4111254654609013,
      "grad_norm": 0.786302951283075,
      "learning_rate": 2e-05,
      "loss": 0.6065024375915528,
      "step": 5410
    },
    {
      "epoch": 0.4118854016262634,
      "grad_norm": 0.9010468255785674,
      "learning_rate": 2e-05,
      "loss": 0.6121093273162842,
      "step": 5420
    },
    {
      "epoch": 0.4126453377916255,
      "grad_norm": 0.766956867218525,
      "learning_rate": 2e-05,
      "loss": 0.601934814453125,
      "step": 5430
    },
    {
      "epoch": 0.4134052739569876,
      "grad_norm": 0.6962997064172681,
      "learning_rate": 2e-05,
      "loss": 0.6192411422729492,
      "step": 5440
    },
    {
      "epoch": 0.41416521012234975,
      "grad_norm": 0.8127403403699424,
      "learning_rate": 2e-05,
      "loss": 0.6208504676818848,
      "step": 5450
    },
    {
      "epoch": 0.4149251462877118,
      "grad_norm": 0.8622096225488602,
      "learning_rate": 2e-05,
      "loss": 0.6212997436523438,
      "step": 5460
    },
    {
      "epoch": 0.41568508245307395,
      "grad_norm": 1.2958030605371762,
      "learning_rate": 2e-05,
      "loss": 0.6049628734588623,
      "step": 5470
    },
    {
      "epoch": 0.416445018618436,
      "grad_norm": 1.011239117517507,
      "learning_rate": 2e-05,
      "loss": 0.6016694068908691,
      "step": 5480
    },
    {
      "epoch": 0.41720495478379815,
      "grad_norm": 0.8586187339797651,
      "learning_rate": 2e-05,
      "loss": 0.6202512264251709,
      "step": 5490
    },
    {
      "epoch": 0.4179648909491603,
      "grad_norm": 1.2222889163034296,
      "learning_rate": 2e-05,
      "loss": 0.636780071258545,
      "step": 5500
    },
    {
      "epoch": 0.41872482711452236,
      "grad_norm": 1.3899554145284045,
      "learning_rate": 2e-05,
      "loss": 0.5982108116149902,
      "step": 5510
    },
    {
      "epoch": 0.4194847632798845,
      "grad_norm": 0.7257744023936229,
      "learning_rate": 2e-05,
      "loss": 0.6098695278167725,
      "step": 5520
    },
    {
      "epoch": 0.4202446994452466,
      "grad_norm": 1.8678544984743346,
      "learning_rate": 2e-05,
      "loss": 0.6057288169860839,
      "step": 5530
    },
    {
      "epoch": 0.4210046356106087,
      "grad_norm": 3.680962057408083,
      "learning_rate": 2e-05,
      "loss": 0.6138186931610108,
      "step": 5540
    },
    {
      "epoch": 0.4217645717759708,
      "grad_norm": 0.900602413157106,
      "learning_rate": 2e-05,
      "loss": 0.6163470268249511,
      "step": 5550
    },
    {
      "epoch": 0.42252450794133295,
      "grad_norm": 0.8260209688510107,
      "learning_rate": 2e-05,
      "loss": 0.5894516944885254,
      "step": 5560
    },
    {
      "epoch": 0.423284444106695,
      "grad_norm": 0.8736036181737387,
      "learning_rate": 2e-05,
      "loss": 0.6143574714660645,
      "step": 5570
    },
    {
      "epoch": 0.42404438027205715,
      "grad_norm": 1.0061192113106865,
      "learning_rate": 2e-05,
      "loss": 0.6095653533935547,
      "step": 5580
    },
    {
      "epoch": 0.4248043164374193,
      "grad_norm": 1.150762815162264,
      "learning_rate": 2e-05,
      "loss": 0.6121916770935059,
      "step": 5590
    },
    {
      "epoch": 0.42556425260278136,
      "grad_norm": 0.7586185146329245,
      "learning_rate": 2e-05,
      "loss": 0.6273753643035889,
      "step": 5600
    },
    {
      "epoch": 0.4263241887681435,
      "grad_norm": 1.0908434438886199,
      "learning_rate": 2e-05,
      "loss": 0.6004355430603028,
      "step": 5610
    },
    {
      "epoch": 0.42708412493350556,
      "grad_norm": 0.6657273752617863,
      "learning_rate": 2e-05,
      "loss": 0.5922695159912109,
      "step": 5620
    },
    {
      "epoch": 0.4278440610988677,
      "grad_norm": 0.8181277652565763,
      "learning_rate": 2e-05,
      "loss": 0.5982446193695068,
      "step": 5630
    },
    {
      "epoch": 0.4286039972642298,
      "grad_norm": 1.1552411523239745,
      "learning_rate": 2e-05,
      "loss": 0.6214675903320312,
      "step": 5640
    },
    {
      "epoch": 0.4293639334295919,
      "grad_norm": 0.6772524878093737,
      "learning_rate": 2e-05,
      "loss": 0.6057360649108887,
      "step": 5650
    },
    {
      "epoch": 0.430123869594954,
      "grad_norm": 1.1126776049318463,
      "learning_rate": 2e-05,
      "loss": 0.6137174606323242,
      "step": 5660
    },
    {
      "epoch": 0.43088380576031615,
      "grad_norm": 1.3276737395634266,
      "learning_rate": 2e-05,
      "loss": 0.5930816650390625,
      "step": 5670
    },
    {
      "epoch": 0.4316437419256782,
      "grad_norm": 0.9036475126599194,
      "learning_rate": 2e-05,
      "loss": 0.6211195945739746,
      "step": 5680
    },
    {
      "epoch": 0.43240367809104036,
      "grad_norm": 1.0319657106407263,
      "learning_rate": 2e-05,
      "loss": 0.6194312572479248,
      "step": 5690
    },
    {
      "epoch": 0.4331636142564025,
      "grad_norm": 0.7653032788378785,
      "learning_rate": 2e-05,
      "loss": 0.6090644836425781,
      "step": 5700
    },
    {
      "epoch": 0.43392355042176456,
      "grad_norm": 0.8862122856982452,
      "learning_rate": 2e-05,
      "loss": 0.5964719772338867,
      "step": 5710
    },
    {
      "epoch": 0.4346834865871267,
      "grad_norm": 1.0696121111889672,
      "learning_rate": 2e-05,
      "loss": 0.593308401107788,
      "step": 5720
    },
    {
      "epoch": 0.4354434227524888,
      "grad_norm": 1.0688216224256586,
      "learning_rate": 2e-05,
      "loss": 0.5971372604370118,
      "step": 5730
    },
    {
      "epoch": 0.4362033589178509,
      "grad_norm": 0.8854619257289424,
      "learning_rate": 2e-05,
      "loss": 0.6087642192840577,
      "step": 5740
    },
    {
      "epoch": 0.436963295083213,
      "grad_norm": 1.2861620659986095,
      "learning_rate": 2e-05,
      "loss": 0.599459171295166,
      "step": 5750
    },
    {
      "epoch": 0.4377232312485751,
      "grad_norm": 4.66291213919596,
      "learning_rate": 2e-05,
      "loss": 0.6218099117279052,
      "step": 5760
    },
    {
      "epoch": 0.4384831674139372,
      "grad_norm": 11.756408085990707,
      "learning_rate": 2e-05,
      "loss": 0.6265668869018555,
      "step": 5770
    },
    {
      "epoch": 0.43924310357929935,
      "grad_norm": 1.4438799307215784,
      "learning_rate": 2e-05,
      "loss": 0.6201286315917969,
      "step": 5780
    },
    {
      "epoch": 0.44000303974466143,
      "grad_norm": 0.9274312778281306,
      "learning_rate": 2e-05,
      "loss": 0.6271205902099609,
      "step": 5790
    },
    {
      "epoch": 0.44076297591002356,
      "grad_norm": 2.6076755261511217,
      "learning_rate": 2e-05,
      "loss": 0.6119579315185547,
      "step": 5800
    },
    {
      "epoch": 0.4415229120753857,
      "grad_norm": 0.9729571040565594,
      "learning_rate": 2e-05,
      "loss": 0.614694356918335,
      "step": 5810
    },
    {
      "epoch": 0.44228284824074776,
      "grad_norm": 0.7474643189740268,
      "learning_rate": 2e-05,
      "loss": 0.5935492992401123,
      "step": 5820
    },
    {
      "epoch": 0.4430427844061099,
      "grad_norm": 0.7877624576119056,
      "learning_rate": 2e-05,
      "loss": 0.580467700958252,
      "step": 5830
    },
    {
      "epoch": 0.443802720571472,
      "grad_norm": 0.6964313149835174,
      "learning_rate": 2e-05,
      "loss": 0.5991967678070068,
      "step": 5840
    },
    {
      "epoch": 0.4445626567368341,
      "grad_norm": 2.628006515974151,
      "learning_rate": 2e-05,
      "loss": 0.6020608425140381,
      "step": 5850
    },
    {
      "epoch": 0.4453225929021962,
      "grad_norm": 0.9267585533455466,
      "learning_rate": 2e-05,
      "loss": 0.6050614833831787,
      "step": 5860
    },
    {
      "epoch": 0.4460825290675583,
      "grad_norm": 1.955379220860901,
      "learning_rate": 2e-05,
      "loss": 0.5938424587249755,
      "step": 5870
    },
    {
      "epoch": 0.4468424652329204,
      "grad_norm": 1.9102195749606043,
      "learning_rate": 2e-05,
      "loss": 0.6048601150512696,
      "step": 5880
    },
    {
      "epoch": 0.44760240139828256,
      "grad_norm": 3.8608464837706116,
      "learning_rate": 2e-05,
      "loss": 0.6077241420745849,
      "step": 5890
    },
    {
      "epoch": 0.44836233756364463,
      "grad_norm": 1.1509039433887518,
      "learning_rate": 2e-05,
      "loss": 0.5975329399108886,
      "step": 5900
    },
    {
      "epoch": 0.44912227372900676,
      "grad_norm": 1.0040418622372769,
      "learning_rate": 2e-05,
      "loss": 0.6157717704772949,
      "step": 5910
    },
    {
      "epoch": 0.4498822098943689,
      "grad_norm": 0.8273801383673376,
      "learning_rate": 2e-05,
      "loss": 0.6343053340911865,
      "step": 5920
    },
    {
      "epoch": 0.45064214605973096,
      "grad_norm": 2.289693071330484,
      "learning_rate": 2e-05,
      "loss": 0.5986771583557129,
      "step": 5930
    },
    {
      "epoch": 0.4514020822250931,
      "grad_norm": 1.2357497962072441,
      "learning_rate": 2e-05,
      "loss": 0.6324886322021485,
      "step": 5940
    },
    {
      "epoch": 0.4521620183904552,
      "grad_norm": 1.756822516339203,
      "learning_rate": 2e-05,
      "loss": 0.5928344249725341,
      "step": 5950
    },
    {
      "epoch": 0.4529219545558173,
      "grad_norm": 1.1678551354328846,
      "learning_rate": 2e-05,
      "loss": 0.5949220180511474,
      "step": 5960
    },
    {
      "epoch": 0.4536818907211794,
      "grad_norm": 0.8001068633564044,
      "learning_rate": 2e-05,
      "loss": 0.6108121395111084,
      "step": 5970
    },
    {
      "epoch": 0.45444182688654156,
      "grad_norm": 1.0245759066479119,
      "learning_rate": 2e-05,
      "loss": 0.6035246849060059,
      "step": 5980
    },
    {
      "epoch": 0.45520176305190363,
      "grad_norm": 0.9153609399661357,
      "learning_rate": 2e-05,
      "loss": 0.6026130676269531,
      "step": 5990
    },
    {
      "epoch": 0.45596169921726576,
      "grad_norm": 0.6677710161272575,
      "learning_rate": 2e-05,
      "loss": 0.6253483772277832,
      "step": 6000
    },
    {
      "epoch": 0.45672163538262783,
      "grad_norm": 1.7284012451415256,
      "learning_rate": 2e-05,
      "loss": 0.6016996383666993,
      "step": 6010
    },
    {
      "epoch": 0.45748157154798996,
      "grad_norm": 0.8501319439566086,
      "learning_rate": 2e-05,
      "loss": 0.6069640159606934,
      "step": 6020
    },
    {
      "epoch": 0.4582415077133521,
      "grad_norm": 1.1373268523266649,
      "learning_rate": 2e-05,
      "loss": 0.6038555145263672,
      "step": 6030
    },
    {
      "epoch": 0.45900144387871417,
      "grad_norm": 0.8020796276270284,
      "learning_rate": 2e-05,
      "loss": 0.5952698707580566,
      "step": 6040
    },
    {
      "epoch": 0.4597613800440763,
      "grad_norm": 3.069877940539361,
      "learning_rate": 2e-05,
      "loss": 0.5840969085693359,
      "step": 6050
    },
    {
      "epoch": 0.4605213162094384,
      "grad_norm": 0.908643538928151,
      "learning_rate": 2e-05,
      "loss": 0.5962722778320313,
      "step": 6060
    },
    {
      "epoch": 0.4612812523748005,
      "grad_norm": 0.7878122830594737,
      "learning_rate": 2e-05,
      "loss": 0.6053701400756836,
      "step": 6070
    },
    {
      "epoch": 0.46204118854016263,
      "grad_norm": 7.35125304398459,
      "learning_rate": 2e-05,
      "loss": 0.6195978164672852,
      "step": 6080
    },
    {
      "epoch": 0.46280112470552476,
      "grad_norm": 1.4696893627154255,
      "learning_rate": 2e-05,
      "loss": 0.6204738616943359,
      "step": 6090
    },
    {
      "epoch": 0.46356106087088683,
      "grad_norm": 1.007474548824823,
      "learning_rate": 2e-05,
      "loss": 0.6281867504119873,
      "step": 6100
    },
    {
      "epoch": 0.46432099703624896,
      "grad_norm": 0.8885879511092609,
      "learning_rate": 2e-05,
      "loss": 0.611607551574707,
      "step": 6110
    },
    {
      "epoch": 0.4650809332016111,
      "grad_norm": 1.4059526013753254,
      "learning_rate": 2e-05,
      "loss": 0.6062422752380371,
      "step": 6120
    },
    {
      "epoch": 0.46584086936697316,
      "grad_norm": 0.9553688954931987,
      "learning_rate": 2e-05,
      "loss": 0.59105544090271,
      "step": 6130
    },
    {
      "epoch": 0.4666008055323353,
      "grad_norm": 0.9464690193921229,
      "learning_rate": 2e-05,
      "loss": 0.6160762786865235,
      "step": 6140
    },
    {
      "epoch": 0.46736074169769737,
      "grad_norm": 0.8677615621756016,
      "learning_rate": 2e-05,
      "loss": 0.623110580444336,
      "step": 6150
    },
    {
      "epoch": 0.4681206778630595,
      "grad_norm": 1.386393700399704,
      "learning_rate": 2e-05,
      "loss": 0.6130038261413574,
      "step": 6160
    },
    {
      "epoch": 0.4688806140284216,
      "grad_norm": 0.9063549368448244,
      "learning_rate": 2e-05,
      "loss": 0.6343528747558593,
      "step": 6170
    },
    {
      "epoch": 0.4696405501937837,
      "grad_norm": 0.6786485813086791,
      "learning_rate": 2e-05,
      "loss": 0.5807430267333984,
      "step": 6180
    },
    {
      "epoch": 0.47040048635914583,
      "grad_norm": 2.6030697747007823,
      "learning_rate": 2e-05,
      "loss": 0.6180171489715576,
      "step": 6190
    },
    {
      "epoch": 0.47116042252450796,
      "grad_norm": 0.7902376024996162,
      "learning_rate": 2e-05,
      "loss": 0.6073391914367676,
      "step": 6200
    },
    {
      "epoch": 0.47192035868987003,
      "grad_norm": 0.7697843733094282,
      "learning_rate": 2e-05,
      "loss": 0.5971420288085938,
      "step": 6210
    },
    {
      "epoch": 0.47268029485523216,
      "grad_norm": 0.7895588425934905,
      "learning_rate": 2e-05,
      "loss": 0.6098786354064941,
      "step": 6220
    },
    {
      "epoch": 0.4734402310205943,
      "grad_norm": 0.9569422051454027,
      "learning_rate": 2e-05,
      "loss": 0.5823875427246094,
      "step": 6230
    },
    {
      "epoch": 0.47420016718595637,
      "grad_norm": 0.6783470521094913,
      "learning_rate": 2e-05,
      "loss": 0.6043367385864258,
      "step": 6240
    },
    {
      "epoch": 0.4749601033513185,
      "grad_norm": 0.8135051255853853,
      "learning_rate": 2e-05,
      "loss": 0.6105962753295898,
      "step": 6250
    },
    {
      "epoch": 0.4757200395166806,
      "grad_norm": 0.7999489560795386,
      "learning_rate": 2e-05,
      "loss": 0.6196487426757813,
      "step": 6260
    },
    {
      "epoch": 0.4764799756820427,
      "grad_norm": 0.8837612326198135,
      "learning_rate": 2e-05,
      "loss": 0.6139410972595215,
      "step": 6270
    },
    {
      "epoch": 0.47723991184740483,
      "grad_norm": 0.7577199409776226,
      "learning_rate": 2e-05,
      "loss": 0.5889076232910156,
      "step": 6280
    },
    {
      "epoch": 0.4779998480127669,
      "grad_norm": 0.7152540112695822,
      "learning_rate": 2e-05,
      "loss": 0.5727242469787598,
      "step": 6290
    },
    {
      "epoch": 0.47875978417812903,
      "grad_norm": 0.7298924167380576,
      "learning_rate": 2e-05,
      "loss": 0.5984285354614258,
      "step": 6300
    },
    {
      "epoch": 0.47951972034349116,
      "grad_norm": 1.0713355364304131,
      "learning_rate": 2e-05,
      "loss": 0.6023166179656982,
      "step": 6310
    },
    {
      "epoch": 0.48027965650885324,
      "grad_norm": 0.7172224374416785,
      "learning_rate": 2e-05,
      "loss": 0.6207765579223633,
      "step": 6320
    },
    {
      "epoch": 0.48103959267421537,
      "grad_norm": 1.143952685084706,
      "learning_rate": 2e-05,
      "loss": 0.6132522583007812,
      "step": 6330
    },
    {
      "epoch": 0.4817995288395775,
      "grad_norm": 1.857083421981545,
      "learning_rate": 2e-05,
      "loss": 0.5882185935974121,
      "step": 6340
    },
    {
      "epoch": 0.48255946500493957,
      "grad_norm": 1.1182466204758976,
      "learning_rate": 2e-05,
      "loss": 0.6194880962371826,
      "step": 6350
    },
    {
      "epoch": 0.4833194011703017,
      "grad_norm": 1.0481371027452566,
      "learning_rate": 2e-05,
      "loss": 0.5987126350402832,
      "step": 6360
    },
    {
      "epoch": 0.48407933733566383,
      "grad_norm": 0.8032735711056892,
      "learning_rate": 2e-05,
      "loss": 0.5902304649353027,
      "step": 6370
    },
    {
      "epoch": 0.4848392735010259,
      "grad_norm": 0.9682525117388433,
      "learning_rate": 2e-05,
      "loss": 0.5944116115570068,
      "step": 6380
    },
    {
      "epoch": 0.48559920966638803,
      "grad_norm": 0.9040882511387373,
      "learning_rate": 2e-05,
      "loss": 0.5786711692810058,
      "step": 6390
    },
    {
      "epoch": 0.4863591458317501,
      "grad_norm": 1.919177352975931,
      "learning_rate": 2e-05,
      "loss": 0.6031159400939942,
      "step": 6400
    },
    {
      "epoch": 0.48711908199711224,
      "grad_norm": 0.8434331500674693,
      "learning_rate": 2e-05,
      "loss": 0.586069393157959,
      "step": 6410
    },
    {
      "epoch": 0.48787901816247436,
      "grad_norm": 0.7575535784199139,
      "learning_rate": 2e-05,
      "loss": 0.6031002044677735,
      "step": 6420
    },
    {
      "epoch": 0.48863895432783644,
      "grad_norm": 4.444949244294525,
      "learning_rate": 2e-05,
      "loss": 0.6169734477996827,
      "step": 6430
    },
    {
      "epoch": 0.48939889049319857,
      "grad_norm": 0.9753574848398425,
      "learning_rate": 2e-05,
      "loss": 0.5992822647094727,
      "step": 6440
    },
    {
      "epoch": 0.4901588266585607,
      "grad_norm": 0.7576266796837238,
      "learning_rate": 2e-05,
      "loss": 0.6021771430969238,
      "step": 6450
    },
    {
      "epoch": 0.49091876282392277,
      "grad_norm": 1.1400529355980922,
      "learning_rate": 2e-05,
      "loss": 0.62110595703125,
      "step": 6460
    },
    {
      "epoch": 0.4916786989892849,
      "grad_norm": 2.924263845101203,
      "learning_rate": 2e-05,
      "loss": 0.5857244014739991,
      "step": 6470
    },
    {
      "epoch": 0.49243863515464703,
      "grad_norm": 2.224845636861606,
      "learning_rate": 2e-05,
      "loss": 0.5894128322601319,
      "step": 6480
    },
    {
      "epoch": 0.4931985713200091,
      "grad_norm": 0.9960178429327669,
      "learning_rate": 2e-05,
      "loss": 0.603248119354248,
      "step": 6490
    },
    {
      "epoch": 0.49395850748537123,
      "grad_norm": 1.076137207375585,
      "learning_rate": 2e-05,
      "loss": 0.6163555145263672,
      "step": 6500
    },
    {
      "epoch": 0.49471844365073336,
      "grad_norm": 2.417647377259034,
      "learning_rate": 2e-05,
      "loss": 0.5931193828582764,
      "step": 6510
    },
    {
      "epoch": 0.49547837981609544,
      "grad_norm": 0.9062583280307338,
      "learning_rate": 2e-05,
      "loss": 0.6182891845703125,
      "step": 6520
    },
    {
      "epoch": 0.49623831598145757,
      "grad_norm": 1.0824235407162541,
      "learning_rate": 2e-05,
      "loss": 0.6120864868164062,
      "step": 6530
    },
    {
      "epoch": 0.49699825214681964,
      "grad_norm": 5.042608352443642,
      "learning_rate": 2e-05,
      "loss": 0.6052207946777344,
      "step": 6540
    },
    {
      "epoch": 0.49775818831218177,
      "grad_norm": 1.0696332466528167,
      "learning_rate": 2e-05,
      "loss": 0.6039605617523194,
      "step": 6550
    },
    {
      "epoch": 0.4985181244775439,
      "grad_norm": 1.126452309935091,
      "learning_rate": 2e-05,
      "loss": 0.6063561916351319,
      "step": 6560
    },
    {
      "epoch": 0.499278060642906,
      "grad_norm": 0.8798432195756426,
      "learning_rate": 2e-05,
      "loss": 0.5886040687561035,
      "step": 6570
    },
    {
      "epoch": 0.5000379968082681,
      "grad_norm": 0.9975651136397208,
      "learning_rate": 2e-05,
      "loss": 0.5875444412231445,
      "step": 6580
    },
    {
      "epoch": 0.5007979329736302,
      "grad_norm": 0.8654826564372172,
      "learning_rate": 2e-05,
      "loss": 0.6321134090423584,
      "step": 6590
    },
    {
      "epoch": 0.5015578691389924,
      "grad_norm": 0.8331670902740226,
      "learning_rate": 2e-05,
      "loss": 0.5988465309143066,
      "step": 6600
    },
    {
      "epoch": 0.5023178053043544,
      "grad_norm": 1.8251533682965329,
      "learning_rate": 2e-05,
      "loss": 0.6200665473937989,
      "step": 6610
    },
    {
      "epoch": 0.5030777414697165,
      "grad_norm": 1.1092331302326297,
      "learning_rate": 2e-05,
      "loss": 0.5946533203125,
      "step": 6620
    },
    {
      "epoch": 0.5038376776350787,
      "grad_norm": 0.8752787512484629,
      "learning_rate": 2e-05,
      "loss": 0.6176362037658691,
      "step": 6630
    },
    {
      "epoch": 0.5045976138004408,
      "grad_norm": 1.169661970265,
      "learning_rate": 2e-05,
      "loss": 0.6031059265136719,
      "step": 6640
    },
    {
      "epoch": 0.5053575499658028,
      "grad_norm": 0.7640022921267342,
      "learning_rate": 2e-05,
      "loss": 0.5927863597869873,
      "step": 6650
    },
    {
      "epoch": 0.506117486131165,
      "grad_norm": 0.9635966547623069,
      "learning_rate": 2e-05,
      "loss": 0.5886881828308106,
      "step": 6660
    },
    {
      "epoch": 0.5068774222965271,
      "grad_norm": 1.0566534310675786,
      "learning_rate": 2e-05,
      "loss": 0.5918074607849121,
      "step": 6670
    },
    {
      "epoch": 0.5076373584618892,
      "grad_norm": 1.9493709363580631,
      "learning_rate": 2e-05,
      "loss": 0.597348976135254,
      "step": 6680
    },
    {
      "epoch": 0.5083972946272514,
      "grad_norm": 1.1306409856098496,
      "learning_rate": 2e-05,
      "loss": 0.6138288021087647,
      "step": 6690
    },
    {
      "epoch": 0.5091572307926134,
      "grad_norm": 1.430172654905642,
      "learning_rate": 2e-05,
      "loss": 0.6193533897399902,
      "step": 6700
    },
    {
      "epoch": 0.5099171669579755,
      "grad_norm": 1.8841500895989163,
      "learning_rate": 2e-05,
      "loss": 0.6378265380859375,
      "step": 6710
    },
    {
      "epoch": 0.5106771031233377,
      "grad_norm": 3.466152867278259,
      "learning_rate": 2e-05,
      "loss": 0.606361198425293,
      "step": 6720
    },
    {
      "epoch": 0.5114370392886998,
      "grad_norm": 1.0383469378522734,
      "learning_rate": 2e-05,
      "loss": 0.6033223628997803,
      "step": 6730
    },
    {
      "epoch": 0.5121969754540618,
      "grad_norm": 1.048886587878698,
      "learning_rate": 2e-05,
      "loss": 0.6298946380615235,
      "step": 6740
    },
    {
      "epoch": 0.5129569116194239,
      "grad_norm": 1.146268966415593,
      "learning_rate": 2e-05,
      "loss": 0.6025066375732422,
      "step": 6750
    },
    {
      "epoch": 0.5137168477847861,
      "grad_norm": 0.8048461502814286,
      "learning_rate": 2e-05,
      "loss": 0.5928178787231445,
      "step": 6760
    },
    {
      "epoch": 0.5144767839501482,
      "grad_norm": 0.7553651585281447,
      "learning_rate": 2e-05,
      "loss": 0.6157333374023437,
      "step": 6770
    },
    {
      "epoch": 0.5152367201155102,
      "grad_norm": 0.7504430270766109,
      "learning_rate": 2e-05,
      "loss": 0.6032750606536865,
      "step": 6780
    },
    {
      "epoch": 0.5159966562808724,
      "grad_norm": 0.9176195979640216,
      "learning_rate": 2e-05,
      "loss": 0.6010446071624755,
      "step": 6790
    },
    {
      "epoch": 0.5167565924462345,
      "grad_norm": 1.0108884565814966,
      "learning_rate": 2e-05,
      "loss": 0.6020214080810546,
      "step": 6800
    },
    {
      "epoch": 0.5175165286115966,
      "grad_norm": 1.6235896495626638,
      "learning_rate": 2e-05,
      "loss": 0.6102366924285889,
      "step": 6810
    },
    {
      "epoch": 0.5182764647769588,
      "grad_norm": 2.656370401630795,
      "learning_rate": 2e-05,
      "loss": 0.6363487243652344,
      "step": 6820
    },
    {
      "epoch": 0.5190364009423208,
      "grad_norm": 0.836566805410327,
      "learning_rate": 2e-05,
      "loss": 0.5957118988037109,
      "step": 6830
    },
    {
      "epoch": 0.5197963371076829,
      "grad_norm": 1.3252389884721456,
      "learning_rate": 2e-05,
      "loss": 0.5878758907318116,
      "step": 6840
    },
    {
      "epoch": 0.5205562732730451,
      "grad_norm": 2.1783769577962997,
      "learning_rate": 2e-05,
      "loss": 0.6027635097503662,
      "step": 6850
    },
    {
      "epoch": 0.5213162094384072,
      "grad_norm": 0.716515611974515,
      "learning_rate": 2e-05,
      "loss": 0.5774549007415771,
      "step": 6860
    },
    {
      "epoch": 0.5220761456037692,
      "grad_norm": 0.9959159610492098,
      "learning_rate": 2e-05,
      "loss": 0.5866391658782959,
      "step": 6870
    },
    {
      "epoch": 0.5228360817691314,
      "grad_norm": 0.935436891566206,
      "learning_rate": 2e-05,
      "loss": 0.5805262565612793,
      "step": 6880
    },
    {
      "epoch": 0.5235960179344935,
      "grad_norm": 0.7824655144739179,
      "learning_rate": 2e-05,
      "loss": 0.5957433700561523,
      "step": 6890
    },
    {
      "epoch": 0.5243559540998556,
      "grad_norm": 1.341845328453487,
      "learning_rate": 2e-05,
      "loss": 0.6033261299133301,
      "step": 6900
    },
    {
      "epoch": 0.5251158902652178,
      "grad_norm": 0.9749620272596536,
      "learning_rate": 2e-05,
      "loss": 0.5961309432983398,
      "step": 6910
    },
    {
      "epoch": 0.5258758264305798,
      "grad_norm": 0.9325257334149267,
      "learning_rate": 2e-05,
      "loss": 0.6034560203552246,
      "step": 6920
    },
    {
      "epoch": 0.5266357625959419,
      "grad_norm": 0.7842659843400063,
      "learning_rate": 2e-05,
      "loss": 0.5966909408569336,
      "step": 6930
    },
    {
      "epoch": 0.5273956987613041,
      "grad_norm": 0.8241100942539185,
      "learning_rate": 2e-05,
      "loss": 0.5934206008911133,
      "step": 6940
    },
    {
      "epoch": 0.5281556349266662,
      "grad_norm": 0.9605948927512455,
      "learning_rate": 2e-05,
      "loss": 0.5994479179382324,
      "step": 6950
    },
    {
      "epoch": 0.5289155710920282,
      "grad_norm": 1.0280487863797978,
      "learning_rate": 2e-05,
      "loss": 0.5963088035583496,
      "step": 6960
    },
    {
      "epoch": 0.5296755072573904,
      "grad_norm": 1.2318362797721356,
      "learning_rate": 2e-05,
      "loss": 0.5852559089660645,
      "step": 6970
    },
    {
      "epoch": 0.5304354434227525,
      "grad_norm": 0.8926446024558692,
      "learning_rate": 2e-05,
      "loss": 0.6210917472839356,
      "step": 6980
    },
    {
      "epoch": 0.5311953795881146,
      "grad_norm": 1.3630881638933492,
      "learning_rate": 2e-05,
      "loss": 0.5937554836273193,
      "step": 6990
    },
    {
      "epoch": 0.5319553157534767,
      "grad_norm": 0.8097711246912026,
      "learning_rate": 2e-05,
      "loss": 0.598024559020996,
      "step": 7000
    },
    {
      "epoch": 0.5327152519188388,
      "grad_norm": 0.7311233491152126,
      "learning_rate": 2e-05,
      "loss": 0.6088153839111328,
      "step": 7010
    },
    {
      "epoch": 0.5334751880842009,
      "grad_norm": 2.3751260191110557,
      "learning_rate": 2e-05,
      "loss": 0.576990795135498,
      "step": 7020
    },
    {
      "epoch": 0.534235124249563,
      "grad_norm": 0.904657513176049,
      "learning_rate": 2e-05,
      "loss": 0.5952944755554199,
      "step": 7030
    },
    {
      "epoch": 0.5349950604149252,
      "grad_norm": 0.7172233360333498,
      "learning_rate": 2e-05,
      "loss": 0.6185964107513428,
      "step": 7040
    },
    {
      "epoch": 0.5357549965802872,
      "grad_norm": 0.9854764505267585,
      "learning_rate": 2e-05,
      "loss": 0.5918662548065186,
      "step": 7050
    },
    {
      "epoch": 0.5365149327456493,
      "grad_norm": 1.1474718951402756,
      "learning_rate": 2e-05,
      "loss": 0.6104835510253906,
      "step": 7060
    },
    {
      "epoch": 0.5372748689110115,
      "grad_norm": 1.3101621391829685,
      "learning_rate": 2e-05,
      "loss": 0.5769122123718262,
      "step": 7070
    },
    {
      "epoch": 0.5380348050763736,
      "grad_norm": 0.9909328996155858,
      "learning_rate": 2e-05,
      "loss": 0.614993667602539,
      "step": 7080
    },
    {
      "epoch": 0.5387947412417357,
      "grad_norm": 1.1213203706286214,
      "learning_rate": 2e-05,
      "loss": 0.5877876281738281,
      "step": 7090
    },
    {
      "epoch": 0.5395546774070978,
      "grad_norm": 0.8605264847930613,
      "learning_rate": 2e-05,
      "loss": 0.589046049118042,
      "step": 7100
    },
    {
      "epoch": 0.5403146135724599,
      "grad_norm": 0.815986532420845,
      "learning_rate": 2e-05,
      "loss": 0.5968078136444092,
      "step": 7110
    },
    {
      "epoch": 0.541074549737822,
      "grad_norm": 0.754144758660177,
      "learning_rate": 2e-05,
      "loss": 0.5825037002563477,
      "step": 7120
    },
    {
      "epoch": 0.5418344859031842,
      "grad_norm": 1.1267921458023553,
      "learning_rate": 2e-05,
      "loss": 0.6083510398864747,
      "step": 7130
    },
    {
      "epoch": 0.5425944220685462,
      "grad_norm": 0.9054080555494669,
      "learning_rate": 2e-05,
      "loss": 0.5948697090148926,
      "step": 7140
    },
    {
      "epoch": 0.5433543582339083,
      "grad_norm": 0.9811720368076369,
      "learning_rate": 2e-05,
      "loss": 0.6049449920654297,
      "step": 7150
    },
    {
      "epoch": 0.5441142943992705,
      "grad_norm": 0.9163166178125595,
      "learning_rate": 2e-05,
      "loss": 0.607512092590332,
      "step": 7160
    },
    {
      "epoch": 0.5448742305646326,
      "grad_norm": 0.9075844059602937,
      "learning_rate": 2e-05,
      "loss": 0.6031986236572265,
      "step": 7170
    },
    {
      "epoch": 0.5456341667299947,
      "grad_norm": 0.8659856274681829,
      "learning_rate": 2e-05,
      "loss": 0.6117231369018554,
      "step": 7180
    },
    {
      "epoch": 0.5463941028953568,
      "grad_norm": 2.1424693267968338,
      "learning_rate": 2e-05,
      "loss": 0.5911962509155273,
      "step": 7190
    },
    {
      "epoch": 0.5471540390607189,
      "grad_norm": 0.9505608149350776,
      "learning_rate": 2e-05,
      "loss": 0.579746150970459,
      "step": 7200
    },
    {
      "epoch": 0.547913975226081,
      "grad_norm": 0.9097584581572223,
      "learning_rate": 2e-05,
      "loss": 0.5821444511413574,
      "step": 7210
    },
    {
      "epoch": 0.5486739113914432,
      "grad_norm": 0.7178937170852348,
      "learning_rate": 2e-05,
      "loss": 0.5929744243621826,
      "step": 7220
    },
    {
      "epoch": 0.5494338475568052,
      "grad_norm": 0.8925697719731688,
      "learning_rate": 2e-05,
      "loss": 0.5949054718017578,
      "step": 7230
    },
    {
      "epoch": 0.5501937837221673,
      "grad_norm": 0.7369256733227759,
      "learning_rate": 2e-05,
      "loss": 0.590759563446045,
      "step": 7240
    },
    {
      "epoch": 0.5509537198875295,
      "grad_norm": 0.7072786139276697,
      "learning_rate": 2e-05,
      "loss": 0.6118543148040771,
      "step": 7250
    },
    {
      "epoch": 0.5517136560528916,
      "grad_norm": 0.7569742575317288,
      "learning_rate": 2e-05,
      "loss": 0.6118139266967774,
      "step": 7260
    },
    {
      "epoch": 0.5524735922182537,
      "grad_norm": 0.8971683161947563,
      "learning_rate": 2e-05,
      "loss": 0.5879284858703613,
      "step": 7270
    },
    {
      "epoch": 0.5532335283836157,
      "grad_norm": 1.1085782921998748,
      "learning_rate": 2e-05,
      "loss": 0.6066747665405273,
      "step": 7280
    },
    {
      "epoch": 0.5539934645489779,
      "grad_norm": 1.1308183292067095,
      "learning_rate": 2e-05,
      "loss": 0.5989069938659668,
      "step": 7290
    },
    {
      "epoch": 0.55475340071434,
      "grad_norm": 0.8868019804008953,
      "learning_rate": 2e-05,
      "loss": 0.5804832458496094,
      "step": 7300
    },
    {
      "epoch": 0.5555133368797021,
      "grad_norm": 0.8105974919536051,
      "learning_rate": 2e-05,
      "loss": 0.6020310878753662,
      "step": 7310
    },
    {
      "epoch": 0.5562732730450642,
      "grad_norm": 0.8788641102306634,
      "learning_rate": 2e-05,
      "loss": 0.5996427536010742,
      "step": 7320
    },
    {
      "epoch": 0.5570332092104263,
      "grad_norm": 0.9155287163178579,
      "learning_rate": 2e-05,
      "loss": 0.5854691982269287,
      "step": 7330
    },
    {
      "epoch": 0.5577931453757884,
      "grad_norm": 0.8286240402946982,
      "learning_rate": 2e-05,
      "loss": 0.6054078578948975,
      "step": 7340
    },
    {
      "epoch": 0.5585530815411506,
      "grad_norm": 1.0569146541023444,
      "learning_rate": 2e-05,
      "loss": 0.6081356048583985,
      "step": 7350
    },
    {
      "epoch": 0.5593130177065126,
      "grad_norm": 1.269874043319411,
      "learning_rate": 2e-05,
      "loss": 0.5951518058776856,
      "step": 7360
    },
    {
      "epoch": 0.5600729538718747,
      "grad_norm": 0.9437778309023767,
      "learning_rate": 2e-05,
      "loss": 0.6027524471282959,
      "step": 7370
    },
    {
      "epoch": 0.5608328900372369,
      "grad_norm": 0.6805407839250125,
      "learning_rate": 2e-05,
      "loss": 0.6010254859924317,
      "step": 7380
    },
    {
      "epoch": 0.561592826202599,
      "grad_norm": 0.7699733600662001,
      "learning_rate": 2e-05,
      "loss": 0.6248666763305664,
      "step": 7390
    },
    {
      "epoch": 0.5623527623679611,
      "grad_norm": 1.2416494225997283,
      "learning_rate": 2e-05,
      "loss": 0.6094069480895996,
      "step": 7400
    },
    {
      "epoch": 0.5631126985333232,
      "grad_norm": 0.7627199383206319,
      "learning_rate": 2e-05,
      "loss": 0.6203700065612793,
      "step": 7410
    },
    {
      "epoch": 0.5638726346986853,
      "grad_norm": 0.9283127688591661,
      "learning_rate": 2e-05,
      "loss": 0.5932563781738281,
      "step": 7420
    },
    {
      "epoch": 0.5646325708640474,
      "grad_norm": 0.7501059651696573,
      "learning_rate": 2e-05,
      "loss": 0.6079219341278076,
      "step": 7430
    },
    {
      "epoch": 0.5653925070294096,
      "grad_norm": 0.7781063196609215,
      "learning_rate": 2e-05,
      "loss": 0.6053366661071777,
      "step": 7440
    },
    {
      "epoch": 0.5661524431947716,
      "grad_norm": 1.127942391931181,
      "learning_rate": 2e-05,
      "loss": 0.5858275890350342,
      "step": 7450
    },
    {
      "epoch": 0.5669123793601337,
      "grad_norm": 2.3722349429025034,
      "learning_rate": 2e-05,
      "loss": 0.5845343589782714,
      "step": 7460
    },
    {
      "epoch": 0.5676723155254959,
      "grad_norm": 0.7014848839682034,
      "learning_rate": 2e-05,
      "loss": 0.6059205532073975,
      "step": 7470
    },
    {
      "epoch": 0.568432251690858,
      "grad_norm": 0.8165378759572188,
      "learning_rate": 2e-05,
      "loss": 0.5884305477142334,
      "step": 7480
    },
    {
      "epoch": 0.56919218785622,
      "grad_norm": 1.028122977727255,
      "learning_rate": 2e-05,
      "loss": 0.6078063011169433,
      "step": 7490
    },
    {
      "epoch": 0.5699521240215822,
      "grad_norm": 0.7841349000330422,
      "learning_rate": 2e-05,
      "loss": 0.5974686622619629,
      "step": 7500
    },
    {
      "epoch": 0.5707120601869443,
      "grad_norm": 0.7763543970767546,
      "learning_rate": 2e-05,
      "loss": 0.6002594947814941,
      "step": 7510
    },
    {
      "epoch": 0.5714719963523064,
      "grad_norm": 0.7993320065756627,
      "learning_rate": 2e-05,
      "loss": 0.6091893196105957,
      "step": 7520
    },
    {
      "epoch": 0.5722319325176685,
      "grad_norm": 0.718507824873949,
      "learning_rate": 2e-05,
      "loss": 0.5982720851898193,
      "step": 7530
    },
    {
      "epoch": 0.5729918686830306,
      "grad_norm": 0.6534073756453674,
      "learning_rate": 2e-05,
      "loss": 0.6091331005096435,
      "step": 7540
    },
    {
      "epoch": 0.5737518048483927,
      "grad_norm": 0.674756334135681,
      "learning_rate": 2e-05,
      "loss": 0.5962636470794678,
      "step": 7550
    },
    {
      "epoch": 0.5745117410137548,
      "grad_norm": 1.8635878539150972,
      "learning_rate": 2e-05,
      "loss": 0.6044404983520508,
      "step": 7560
    },
    {
      "epoch": 0.575271677179117,
      "grad_norm": 0.7985935272241238,
      "learning_rate": 2e-05,
      "loss": 0.6055945873260498,
      "step": 7570
    },
    {
      "epoch": 0.576031613344479,
      "grad_norm": 0.9493126849388197,
      "learning_rate": 2e-05,
      "loss": 0.5981311798095703,
      "step": 7580
    },
    {
      "epoch": 0.5767915495098411,
      "grad_norm": 1.6739806873865866,
      "learning_rate": 2e-05,
      "loss": 0.6075414180755615,
      "step": 7590
    },
    {
      "epoch": 0.5775514856752033,
      "grad_norm": 0.8180291367130847,
      "learning_rate": 2e-05,
      "loss": 0.6190924644470215,
      "step": 7600
    },
    {
      "epoch": 0.5783114218405654,
      "grad_norm": 0.9659712644649294,
      "learning_rate": 2e-05,
      "loss": 0.6511466979980469,
      "step": 7610
    },
    {
      "epoch": 0.5790713580059275,
      "grad_norm": 0.8056075271707009,
      "learning_rate": 2e-05,
      "loss": 0.5945711135864258,
      "step": 7620
    },
    {
      "epoch": 0.5798312941712896,
      "grad_norm": 0.6787267687106643,
      "learning_rate": 2e-05,
      "loss": 0.5880901336669921,
      "step": 7630
    },
    {
      "epoch": 0.5805912303366517,
      "grad_norm": 0.7827041929906855,
      "learning_rate": 2e-05,
      "loss": 0.6007994413375854,
      "step": 7640
    },
    {
      "epoch": 0.5813511665020138,
      "grad_norm": 0.8348478774766127,
      "learning_rate": 2e-05,
      "loss": 0.6121800422668457,
      "step": 7650
    },
    {
      "epoch": 0.582111102667376,
      "grad_norm": 0.8796039856746606,
      "learning_rate": 2e-05,
      "loss": 0.6179386138916015,
      "step": 7660
    },
    {
      "epoch": 0.582871038832738,
      "grad_norm": 0.692953688236863,
      "learning_rate": 2e-05,
      "loss": 0.5875237941741943,
      "step": 7670
    },
    {
      "epoch": 0.5836309749981001,
      "grad_norm": 0.9704196955891338,
      "learning_rate": 2e-05,
      "loss": 0.604813528060913,
      "step": 7680
    },
    {
      "epoch": 0.5843909111634623,
      "grad_norm": 0.7375540700314249,
      "learning_rate": 2e-05,
      "loss": 0.5853425979614257,
      "step": 7690
    },
    {
      "epoch": 0.5851508473288244,
      "grad_norm": 1.023606304681373,
      "learning_rate": 2e-05,
      "loss": 0.6159492492675781,
      "step": 7700
    },
    {
      "epoch": 0.5859107834941865,
      "grad_norm": 0.9180897554024244,
      "learning_rate": 2e-05,
      "loss": 0.6059733390808105,
      "step": 7710
    },
    {
      "epoch": 0.5866707196595486,
      "grad_norm": 0.9939867951138021,
      "learning_rate": 2e-05,
      "loss": 0.6010074615478516,
      "step": 7720
    },
    {
      "epoch": 0.5874306558249107,
      "grad_norm": 0.8587070994999046,
      "learning_rate": 2e-05,
      "loss": 0.599238395690918,
      "step": 7730
    },
    {
      "epoch": 0.5881905919902728,
      "grad_norm": 1.349243827874104,
      "learning_rate": 2e-05,
      "loss": 0.6118513107299804,
      "step": 7740
    },
    {
      "epoch": 0.588950528155635,
      "grad_norm": 0.8988416494542097,
      "learning_rate": 2e-05,
      "loss": 0.5836956024169921,
      "step": 7750
    },
    {
      "epoch": 0.589710464320997,
      "grad_norm": 2.170368403570881,
      "learning_rate": 2e-05,
      "loss": 0.5999057292938232,
      "step": 7760
    },
    {
      "epoch": 0.5904704004863591,
      "grad_norm": 0.8268832240077089,
      "learning_rate": 2e-05,
      "loss": 0.5690643310546875,
      "step": 7770
    },
    {
      "epoch": 0.5912303366517212,
      "grad_norm": 0.8927397561720134,
      "learning_rate": 2e-05,
      "loss": 0.6162868499755859,
      "step": 7780
    },
    {
      "epoch": 0.5919902728170834,
      "grad_norm": 0.8393668169900896,
      "learning_rate": 2e-05,
      "loss": 0.608328914642334,
      "step": 7790
    },
    {
      "epoch": 0.5927502089824455,
      "grad_norm": 0.8750607642018079,
      "learning_rate": 2e-05,
      "loss": 0.6036715507507324,
      "step": 7800
    },
    {
      "epoch": 0.5935101451478075,
      "grad_norm": 5.5889702521967095,
      "learning_rate": 2e-05,
      "loss": 0.5919855117797852,
      "step": 7810
    },
    {
      "epoch": 0.5942700813131697,
      "grad_norm": 1.4739197036557226,
      "learning_rate": 2e-05,
      "loss": 0.5963436126708984,
      "step": 7820
    },
    {
      "epoch": 0.5950300174785318,
      "grad_norm": 1.1282425857178822,
      "learning_rate": 2e-05,
      "loss": 0.582796859741211,
      "step": 7830
    },
    {
      "epoch": 0.5957899536438939,
      "grad_norm": 0.8576183612877244,
      "learning_rate": 2e-05,
      "loss": 0.6085587978363037,
      "step": 7840
    },
    {
      "epoch": 0.596549889809256,
      "grad_norm": 1.0848186372620021,
      "learning_rate": 2e-05,
      "loss": 0.6275556087493896,
      "step": 7850
    },
    {
      "epoch": 0.5973098259746181,
      "grad_norm": 1.208933061375334,
      "learning_rate": 2e-05,
      "loss": 0.5996315002441406,
      "step": 7860
    },
    {
      "epoch": 0.5980697621399802,
      "grad_norm": 0.9541738873612453,
      "learning_rate": 2e-05,
      "loss": 0.5928617477416992,
      "step": 7870
    },
    {
      "epoch": 0.5988296983053424,
      "grad_norm": 0.8480269306896233,
      "learning_rate": 2e-05,
      "loss": 0.6152638912200927,
      "step": 7880
    },
    {
      "epoch": 0.5995896344707045,
      "grad_norm": 0.8868547145131801,
      "learning_rate": 2e-05,
      "loss": 0.6036113739013672,
      "step": 7890
    },
    {
      "epoch": 0.6003495706360665,
      "grad_norm": 0.9119470503064178,
      "learning_rate": 2e-05,
      "loss": 0.5794699668884278,
      "step": 7900
    },
    {
      "epoch": 0.6011095068014287,
      "grad_norm": 2.4907456924637414,
      "learning_rate": 2e-05,
      "loss": 0.6054845333099366,
      "step": 7910
    },
    {
      "epoch": 0.6018694429667908,
      "grad_norm": 0.7623228375219487,
      "learning_rate": 2e-05,
      "loss": 0.5926270961761475,
      "step": 7920
    },
    {
      "epoch": 0.6026293791321529,
      "grad_norm": 0.7762188404360846,
      "learning_rate": 2e-05,
      "loss": 0.59921293258667,
      "step": 7930
    },
    {
      "epoch": 0.603389315297515,
      "grad_norm": 0.6557570507062174,
      "learning_rate": 2e-05,
      "loss": 0.5805189132690429,
      "step": 7940
    },
    {
      "epoch": 0.6041492514628771,
      "grad_norm": 0.8307028107734032,
      "learning_rate": 2e-05,
      "loss": 0.5973512172698975,
      "step": 7950
    },
    {
      "epoch": 0.6049091876282392,
      "grad_norm": 1.3096555723991214,
      "learning_rate": 2e-05,
      "loss": 0.6090723037719726,
      "step": 7960
    },
    {
      "epoch": 0.6056691237936014,
      "grad_norm": 1.0704487761614303,
      "learning_rate": 2e-05,
      "loss": 0.585895299911499,
      "step": 7970
    },
    {
      "epoch": 0.6064290599589635,
      "grad_norm": 0.7725260556614094,
      "learning_rate": 2e-05,
      "loss": 0.6134288787841797,
      "step": 7980
    },
    {
      "epoch": 0.6071889961243255,
      "grad_norm": 1.0091690335540113,
      "learning_rate": 2e-05,
      "loss": 0.58992919921875,
      "step": 7990
    },
    {
      "epoch": 0.6079489322896877,
      "grad_norm": 0.7134912856158722,
      "learning_rate": 2e-05,
      "loss": 0.5889213562011719,
      "step": 8000
    },
    {
      "epoch": 0.6087088684550498,
      "grad_norm": 0.6873001864771773,
      "learning_rate": 2e-05,
      "loss": 0.5743563652038575,
      "step": 8010
    },
    {
      "epoch": 0.6094688046204119,
      "grad_norm": 0.8061964723233171,
      "learning_rate": 2e-05,
      "loss": 0.609766674041748,
      "step": 8020
    },
    {
      "epoch": 0.610228740785774,
      "grad_norm": 1.0095505782839571,
      "learning_rate": 2e-05,
      "loss": 0.605078649520874,
      "step": 8030
    },
    {
      "epoch": 0.6109886769511361,
      "grad_norm": 0.8343392296757187,
      "learning_rate": 2e-05,
      "loss": 0.6028301239013671,
      "step": 8040
    },
    {
      "epoch": 0.6117486131164982,
      "grad_norm": 0.7324533883761908,
      "learning_rate": 2e-05,
      "loss": 0.6099300384521484,
      "step": 8050
    },
    {
      "epoch": 0.6125085492818603,
      "grad_norm": 0.6795531575958551,
      "learning_rate": 2e-05,
      "loss": 0.605167293548584,
      "step": 8060
    },
    {
      "epoch": 0.6132684854472225,
      "grad_norm": 0.8300933915269836,
      "learning_rate": 2e-05,
      "loss": 0.5997323036193848,
      "step": 8070
    },
    {
      "epoch": 0.6140284216125845,
      "grad_norm": 0.6300283800275384,
      "learning_rate": 2e-05,
      "loss": 0.5956508636474609,
      "step": 8080
    },
    {
      "epoch": 0.6147883577779466,
      "grad_norm": 0.7728525803100674,
      "learning_rate": 2e-05,
      "loss": 0.6140737056732177,
      "step": 8090
    },
    {
      "epoch": 0.6155482939433088,
      "grad_norm": 0.8343076801442537,
      "learning_rate": 2e-05,
      "loss": 0.5962836265563964,
      "step": 8100
    },
    {
      "epoch": 0.6163082301086709,
      "grad_norm": 0.6587916596670088,
      "learning_rate": 2e-05,
      "loss": 0.6022719383239746,
      "step": 8110
    },
    {
      "epoch": 0.6170681662740329,
      "grad_norm": 0.7595460351027042,
      "learning_rate": 2e-05,
      "loss": 0.5757766723632812,
      "step": 8120
    },
    {
      "epoch": 0.6178281024393951,
      "grad_norm": 0.615393587741323,
      "learning_rate": 2e-05,
      "loss": 0.5781272411346435,
      "step": 8130
    },
    {
      "epoch": 0.6185880386047572,
      "grad_norm": 0.8910211411971501,
      "learning_rate": 2e-05,
      "loss": 0.5962740898132324,
      "step": 8140
    },
    {
      "epoch": 0.6193479747701193,
      "grad_norm": 0.7417916421134424,
      "learning_rate": 2e-05,
      "loss": 0.5986260414123535,
      "step": 8150
    },
    {
      "epoch": 0.6201079109354815,
      "grad_norm": 0.7824981924426299,
      "learning_rate": 2e-05,
      "loss": 0.6056794166564942,
      "step": 8160
    },
    {
      "epoch": 0.6208678471008435,
      "grad_norm": 0.796731934451236,
      "learning_rate": 2e-05,
      "loss": 0.6154984951019287,
      "step": 8170
    },
    {
      "epoch": 0.6216277832662056,
      "grad_norm": 0.7919546673658281,
      "learning_rate": 2e-05,
      "loss": 0.5867465019226075,
      "step": 8180
    },
    {
      "epoch": 0.6223877194315678,
      "grad_norm": 0.636746179459474,
      "learning_rate": 2e-05,
      "loss": 0.6003735542297364,
      "step": 8190
    },
    {
      "epoch": 0.6231476555969299,
      "grad_norm": 0.6682693865561821,
      "learning_rate": 2e-05,
      "loss": 0.5870049953460693,
      "step": 8200
    },
    {
      "epoch": 0.6239075917622919,
      "grad_norm": 0.7196998530945142,
      "learning_rate": 2e-05,
      "loss": 0.6013730525970459,
      "step": 8210
    },
    {
      "epoch": 0.6246675279276541,
      "grad_norm": 0.8153708258335793,
      "learning_rate": 2e-05,
      "loss": 0.5879334926605224,
      "step": 8220
    },
    {
      "epoch": 0.6254274640930162,
      "grad_norm": 0.8919724871891992,
      "learning_rate": 2e-05,
      "loss": 0.5782451629638672,
      "step": 8230
    },
    {
      "epoch": 0.6261874002583783,
      "grad_norm": 0.8726925153005258,
      "learning_rate": 2e-05,
      "loss": 0.5840807914733886,
      "step": 8240
    },
    {
      "epoch": 0.6269473364237405,
      "grad_norm": 0.8928166176839037,
      "learning_rate": 2e-05,
      "loss": 0.5926889419555664,
      "step": 8250
    },
    {
      "epoch": 0.6277072725891025,
      "grad_norm": 0.9806028853312723,
      "learning_rate": 2e-05,
      "loss": 0.5755433082580567,
      "step": 8260
    },
    {
      "epoch": 0.6284672087544646,
      "grad_norm": 0.9616874900937195,
      "learning_rate": 2e-05,
      "loss": 0.6066426753997802,
      "step": 8270
    },
    {
      "epoch": 0.6292271449198268,
      "grad_norm": 0.7345748015972401,
      "learning_rate": 2e-05,
      "loss": 0.5875670433044433,
      "step": 8280
    },
    {
      "epoch": 0.6299870810851889,
      "grad_norm": 0.7242154309079251,
      "learning_rate": 2e-05,
      "loss": 0.6026911735534668,
      "step": 8290
    },
    {
      "epoch": 0.6307470172505509,
      "grad_norm": 0.7435435461169804,
      "learning_rate": 2e-05,
      "loss": 0.6008343696594238,
      "step": 8300
    },
    {
      "epoch": 0.631506953415913,
      "grad_norm": 0.7404656697780453,
      "learning_rate": 2e-05,
      "loss": 0.5818479061126709,
      "step": 8310
    },
    {
      "epoch": 0.6322668895812752,
      "grad_norm": 1.4711790191055332,
      "learning_rate": 2e-05,
      "loss": 0.6014999389648438,
      "step": 8320
    },
    {
      "epoch": 0.6330268257466373,
      "grad_norm": 0.7813251843112322,
      "learning_rate": 2e-05,
      "loss": 0.5982386589050293,
      "step": 8330
    },
    {
      "epoch": 0.6337867619119993,
      "grad_norm": 0.6287106707596353,
      "learning_rate": 2e-05,
      "loss": 0.5961961269378662,
      "step": 8340
    },
    {
      "epoch": 0.6345466980773615,
      "grad_norm": 0.8688451824744893,
      "learning_rate": 2e-05,
      "loss": 0.5741853237152099,
      "step": 8350
    },
    {
      "epoch": 0.6353066342427236,
      "grad_norm": 0.8989689009823774,
      "learning_rate": 2e-05,
      "loss": 0.6156363487243652,
      "step": 8360
    },
    {
      "epoch": 0.6360665704080857,
      "grad_norm": 0.7228278855362366,
      "learning_rate": 2e-05,
      "loss": 0.5770135402679444,
      "step": 8370
    },
    {
      "epoch": 0.6368265065734479,
      "grad_norm": 0.7494388601649471,
      "learning_rate": 2e-05,
      "loss": 0.6089717864990234,
      "step": 8380
    },
    {
      "epoch": 0.6375864427388099,
      "grad_norm": 0.6269421162776351,
      "learning_rate": 2e-05,
      "loss": 0.5973659515380859,
      "step": 8390
    },
    {
      "epoch": 0.638346378904172,
      "grad_norm": 0.6879595679641235,
      "learning_rate": 2e-05,
      "loss": 0.5794967651367188,
      "step": 8400
    },
    {
      "epoch": 0.6391063150695342,
      "grad_norm": 0.7041101587392335,
      "learning_rate": 2e-05,
      "loss": 0.6129167079925537,
      "step": 8410
    },
    {
      "epoch": 0.6398662512348963,
      "grad_norm": 0.7646227191115826,
      "learning_rate": 2e-05,
      "loss": 0.5722673416137696,
      "step": 8420
    },
    {
      "epoch": 0.6406261874002583,
      "grad_norm": 0.7253958963766086,
      "learning_rate": 2e-05,
      "loss": 0.6056432723999023,
      "step": 8430
    },
    {
      "epoch": 0.6413861235656205,
      "grad_norm": 0.8468911760134725,
      "learning_rate": 2e-05,
      "loss": 0.5935896873474121,
      "step": 8440
    },
    {
      "epoch": 0.6421460597309826,
      "grad_norm": 0.8751076908063481,
      "learning_rate": 2e-05,
      "loss": 0.5851859092712403,
      "step": 8450
    },
    {
      "epoch": 0.6429059958963447,
      "grad_norm": 0.7525658481783363,
      "learning_rate": 2e-05,
      "loss": 0.6047962188720704,
      "step": 8460
    },
    {
      "epoch": 0.6436659320617069,
      "grad_norm": 0.6109307206704391,
      "learning_rate": 2e-05,
      "loss": 0.5730659008026123,
      "step": 8470
    },
    {
      "epoch": 0.6444258682270689,
      "grad_norm": 0.6859065710098108,
      "learning_rate": 2e-05,
      "loss": 0.5801695823669434,
      "step": 8480
    },
    {
      "epoch": 0.645185804392431,
      "grad_norm": 0.691765114917809,
      "learning_rate": 2e-05,
      "loss": 0.5768918037414551,
      "step": 8490
    },
    {
      "epoch": 0.6459457405577932,
      "grad_norm": 0.8412595409989185,
      "learning_rate": 2e-05,
      "loss": 0.5940985679626465,
      "step": 8500
    },
    {
      "epoch": 0.6467056767231553,
      "grad_norm": 0.7114082502432819,
      "learning_rate": 2e-05,
      "loss": 0.5932064056396484,
      "step": 8510
    },
    {
      "epoch": 0.6474656128885173,
      "grad_norm": 1.2174024563872616,
      "learning_rate": 2e-05,
      "loss": 0.5861550807952881,
      "step": 8520
    },
    {
      "epoch": 0.6482255490538795,
      "grad_norm": 0.7581719390063164,
      "learning_rate": 2e-05,
      "loss": 0.5635357856750488,
      "step": 8530
    },
    {
      "epoch": 0.6489854852192416,
      "grad_norm": 0.7525011274426494,
      "learning_rate": 2e-05,
      "loss": 0.6120333671569824,
      "step": 8540
    },
    {
      "epoch": 0.6497454213846037,
      "grad_norm": 1.3391303248867492,
      "learning_rate": 2e-05,
      "loss": 0.581744909286499,
      "step": 8550
    },
    {
      "epoch": 0.6505053575499659,
      "grad_norm": 0.7409995178889197,
      "learning_rate": 2e-05,
      "loss": 0.6137965202331543,
      "step": 8560
    },
    {
      "epoch": 0.6512652937153279,
      "grad_norm": 1.3634632071595136,
      "learning_rate": 2e-05,
      "loss": 0.5984136581420898,
      "step": 8570
    },
    {
      "epoch": 0.65202522988069,
      "grad_norm": 0.9988620881230286,
      "learning_rate": 2e-05,
      "loss": 0.6131124496459961,
      "step": 8580
    },
    {
      "epoch": 0.6527851660460521,
      "grad_norm": 0.881374934543849,
      "learning_rate": 2e-05,
      "loss": 0.6051208019256592,
      "step": 8590
    },
    {
      "epoch": 0.6535451022114143,
      "grad_norm": 0.8342875370449667,
      "learning_rate": 2e-05,
      "loss": 0.5960539817810059,
      "step": 8600
    },
    {
      "epoch": 0.6543050383767763,
      "grad_norm": 1.2474132507378013,
      "learning_rate": 2e-05,
      "loss": 0.5886716842651367,
      "step": 8610
    },
    {
      "epoch": 0.6550649745421384,
      "grad_norm": 1.243214213581915,
      "learning_rate": 2e-05,
      "loss": 0.5813884735107422,
      "step": 8620
    },
    {
      "epoch": 0.6558249107075006,
      "grad_norm": 0.7606529928190813,
      "learning_rate": 2e-05,
      "loss": 0.5905316829681396,
      "step": 8630
    },
    {
      "epoch": 0.6565848468728627,
      "grad_norm": 0.7377859775394447,
      "learning_rate": 2e-05,
      "loss": 0.5877938747406006,
      "step": 8640
    },
    {
      "epoch": 0.6573447830382247,
      "grad_norm": 0.8507253387144404,
      "learning_rate": 2e-05,
      "loss": 0.5901322364807129,
      "step": 8650
    },
    {
      "epoch": 0.6581047192035869,
      "grad_norm": 0.7984819317175492,
      "learning_rate": 2e-05,
      "loss": 0.5729423046112061,
      "step": 8660
    },
    {
      "epoch": 0.658864655368949,
      "grad_norm": 1.0528266142574558,
      "learning_rate": 2e-05,
      "loss": 0.5799797058105469,
      "step": 8670
    },
    {
      "epoch": 0.6596245915343111,
      "grad_norm": 1.0197574834393646,
      "learning_rate": 2e-05,
      "loss": 0.6271066188812255,
      "step": 8680
    },
    {
      "epoch": 0.6603845276996733,
      "grad_norm": 0.7906718916145783,
      "learning_rate": 2e-05,
      "loss": 0.5790286064147949,
      "step": 8690
    },
    {
      "epoch": 0.6611444638650353,
      "grad_norm": 0.841466839107414,
      "learning_rate": 2e-05,
      "loss": 0.5901204586029053,
      "step": 8700
    },
    {
      "epoch": 0.6619044000303974,
      "grad_norm": 0.7028446977123615,
      "learning_rate": 2e-05,
      "loss": 0.5993346214294434,
      "step": 8710
    },
    {
      "epoch": 0.6626643361957596,
      "grad_norm": 0.7078986511968858,
      "learning_rate": 2e-05,
      "loss": 0.600566053390503,
      "step": 8720
    },
    {
      "epoch": 0.6634242723611217,
      "grad_norm": 0.6581292572026805,
      "learning_rate": 2e-05,
      "loss": 0.6093256950378418,
      "step": 8730
    },
    {
      "epoch": 0.6641842085264837,
      "grad_norm": 0.636772972180966,
      "learning_rate": 2e-05,
      "loss": 0.6010907649993896,
      "step": 8740
    },
    {
      "epoch": 0.6649441446918459,
      "grad_norm": 0.6701786172550229,
      "learning_rate": 2e-05,
      "loss": 0.587213659286499,
      "step": 8750
    },
    {
      "epoch": 0.665704080857208,
      "grad_norm": 0.660905024594776,
      "learning_rate": 2e-05,
      "loss": 0.5752870559692382,
      "step": 8760
    },
    {
      "epoch": 0.6664640170225701,
      "grad_norm": 0.6832511663780295,
      "learning_rate": 2e-05,
      "loss": 0.6052382469177247,
      "step": 8770
    },
    {
      "epoch": 0.6672239531879323,
      "grad_norm": 0.7199276738485614,
      "learning_rate": 2e-05,
      "loss": 0.5947525024414062,
      "step": 8780
    },
    {
      "epoch": 0.6679838893532943,
      "grad_norm": 0.821456460120791,
      "learning_rate": 2e-05,
      "loss": 0.6010943412780761,
      "step": 8790
    },
    {
      "epoch": 0.6687438255186564,
      "grad_norm": 0.6624381912712659,
      "learning_rate": 2e-05,
      "loss": 0.5849940299987793,
      "step": 8800
    },
    {
      "epoch": 0.6695037616840186,
      "grad_norm": 0.700521500868997,
      "learning_rate": 2e-05,
      "loss": 0.5700336456298828,
      "step": 8810
    },
    {
      "epoch": 0.6702636978493807,
      "grad_norm": 0.8033229960124904,
      "learning_rate": 2e-05,
      "loss": 0.594280195236206,
      "step": 8820
    },
    {
      "epoch": 0.6710236340147427,
      "grad_norm": 1.3927479596492323,
      "learning_rate": 2e-05,
      "loss": 0.569603157043457,
      "step": 8830
    },
    {
      "epoch": 0.6717835701801048,
      "grad_norm": 0.6861888970493578,
      "learning_rate": 2e-05,
      "loss": 0.5899533271789551,
      "step": 8840
    },
    {
      "epoch": 0.672543506345467,
      "grad_norm": 0.898064627565956,
      "learning_rate": 2e-05,
      "loss": 0.6035385131835938,
      "step": 8850
    },
    {
      "epoch": 0.6733034425108291,
      "grad_norm": 0.8522401190735633,
      "learning_rate": 2e-05,
      "loss": 0.5903943061828614,
      "step": 8860
    },
    {
      "epoch": 0.6740633786761911,
      "grad_norm": 0.7151656851693126,
      "learning_rate": 2e-05,
      "loss": 0.5817798614501953,
      "step": 8870
    },
    {
      "epoch": 0.6748233148415533,
      "grad_norm": 0.8246444312893565,
      "learning_rate": 2e-05,
      "loss": 0.5800168514251709,
      "step": 8880
    },
    {
      "epoch": 0.6755832510069154,
      "grad_norm": 1.1050013443310225,
      "learning_rate": 2e-05,
      "loss": 0.5800858497619629,
      "step": 8890
    },
    {
      "epoch": 0.6763431871722775,
      "grad_norm": 1.0765659124945022,
      "learning_rate": 2e-05,
      "loss": 0.5707725524902344,
      "step": 8900
    },
    {
      "epoch": 0.6771031233376397,
      "grad_norm": 0.729960081050483,
      "learning_rate": 2e-05,
      "loss": 0.6077522277832031,
      "step": 8910
    },
    {
      "epoch": 0.6778630595030017,
      "grad_norm": 0.8294870548068406,
      "learning_rate": 2e-05,
      "loss": 0.5900848388671875,
      "step": 8920
    },
    {
      "epoch": 0.6786229956683638,
      "grad_norm": 1.2258616674150775,
      "learning_rate": 2e-05,
      "loss": 0.5774793148040771,
      "step": 8930
    },
    {
      "epoch": 0.679382931833726,
      "grad_norm": 1.2330799102629177,
      "learning_rate": 2e-05,
      "loss": 0.5875690460205079,
      "step": 8940
    },
    {
      "epoch": 0.6801428679990881,
      "grad_norm": 1.0209345516494845,
      "learning_rate": 2e-05,
      "loss": 0.589243221282959,
      "step": 8950
    },
    {
      "epoch": 0.6809028041644501,
      "grad_norm": 1.3792771671368815,
      "learning_rate": 2e-05,
      "loss": 0.6029260158538818,
      "step": 8960
    },
    {
      "epoch": 0.6816627403298123,
      "grad_norm": 0.8185214829561661,
      "learning_rate": 2e-05,
      "loss": 0.5939120769500732,
      "step": 8970
    },
    {
      "epoch": 0.6824226764951744,
      "grad_norm": 0.7511824436841457,
      "learning_rate": 2e-05,
      "loss": 0.5958861351013184,
      "step": 8980
    },
    {
      "epoch": 0.6831826126605365,
      "grad_norm": 0.9445719613661021,
      "learning_rate": 2e-05,
      "loss": 0.5867552757263184,
      "step": 8990
    },
    {
      "epoch": 0.6839425488258987,
      "grad_norm": 0.8380620489874602,
      "learning_rate": 2e-05,
      "loss": 0.6105091094970703,
      "step": 9000
    },
    {
      "epoch": 0.6847024849912607,
      "grad_norm": 0.8507960935997074,
      "learning_rate": 2e-05,
      "loss": 0.5816993713378906,
      "step": 9010
    },
    {
      "epoch": 0.6854624211566228,
      "grad_norm": 8.28120615893421,
      "learning_rate": 2e-05,
      "loss": 0.5999642372131347,
      "step": 9020
    },
    {
      "epoch": 0.686222357321985,
      "grad_norm": 0.8687326264453524,
      "learning_rate": 2e-05,
      "loss": 0.5581311702728271,
      "step": 9030
    },
    {
      "epoch": 0.6869822934873471,
      "grad_norm": 0.8290703466934447,
      "learning_rate": 2e-05,
      "loss": 0.6090290069580078,
      "step": 9040
    },
    {
      "epoch": 0.6877422296527091,
      "grad_norm": 0.6874383413128794,
      "learning_rate": 2e-05,
      "loss": 0.6011064052581787,
      "step": 9050
    },
    {
      "epoch": 0.6885021658180713,
      "grad_norm": 0.7694894135367721,
      "learning_rate": 2e-05,
      "loss": 0.5797263145446777,
      "step": 9060
    },
    {
      "epoch": 0.6892621019834334,
      "grad_norm": 1.165431221956054,
      "learning_rate": 2e-05,
      "loss": 0.5832767486572266,
      "step": 9070
    },
    {
      "epoch": 0.6900220381487955,
      "grad_norm": 0.8116460566419788,
      "learning_rate": 2e-05,
      "loss": 0.5757497310638428,
      "step": 9080
    },
    {
      "epoch": 0.6907819743141577,
      "grad_norm": 0.948881824695589,
      "learning_rate": 2e-05,
      "loss": 0.5877982139587402,
      "step": 9090
    },
    {
      "epoch": 0.6915419104795197,
      "grad_norm": 0.975732765993571,
      "learning_rate": 2e-05,
      "loss": 0.5739354610443115,
      "step": 9100
    },
    {
      "epoch": 0.6923018466448818,
      "grad_norm": 0.8566536962737435,
      "learning_rate": 2e-05,
      "loss": 0.5980454444885254,
      "step": 9110
    },
    {
      "epoch": 0.6930617828102439,
      "grad_norm": 0.894921576030419,
      "learning_rate": 2e-05,
      "loss": 0.6054080963134766,
      "step": 9120
    },
    {
      "epoch": 0.6938217189756061,
      "grad_norm": 4.578159401595052,
      "learning_rate": 2e-05,
      "loss": 0.61366286277771,
      "step": 9130
    },
    {
      "epoch": 0.6945816551409681,
      "grad_norm": 0.9730201266082712,
      "learning_rate": 2e-05,
      "loss": 0.5760093212127686,
      "step": 9140
    },
    {
      "epoch": 0.6953415913063302,
      "grad_norm": 0.7541566895980375,
      "learning_rate": 2e-05,
      "loss": 0.5806249141693115,
      "step": 9150
    },
    {
      "epoch": 0.6961015274716924,
      "grad_norm": 1.0163688052960889,
      "learning_rate": 2e-05,
      "loss": 0.58531494140625,
      "step": 9160
    },
    {
      "epoch": 0.6968614636370545,
      "grad_norm": 1.5330672949250783,
      "learning_rate": 2e-05,
      "loss": 0.6041053295135498,
      "step": 9170
    },
    {
      "epoch": 0.6976213998024166,
      "grad_norm": 1.0490013342717643,
      "learning_rate": 2e-05,
      "loss": 0.6016427516937256,
      "step": 9180
    },
    {
      "epoch": 0.6983813359677787,
      "grad_norm": 0.723739684181483,
      "learning_rate": 2e-05,
      "loss": 0.5917651176452636,
      "step": 9190
    },
    {
      "epoch": 0.6991412721331408,
      "grad_norm": 2.0897653424178086,
      "learning_rate": 2e-05,
      "loss": 0.5874265193939209,
      "step": 9200
    },
    {
      "epoch": 0.6999012082985029,
      "grad_norm": 0.9265352299084221,
      "learning_rate": 2e-05,
      "loss": 0.5957166671752929,
      "step": 9210
    },
    {
      "epoch": 0.7006611444638651,
      "grad_norm": 1.0925344161127133,
      "learning_rate": 2e-05,
      "loss": 0.5931726455688476,
      "step": 9220
    },
    {
      "epoch": 0.7014210806292271,
      "grad_norm": 1.6025640081360681,
      "learning_rate": 2e-05,
      "loss": 0.5884226799011231,
      "step": 9230
    },
    {
      "epoch": 0.7021810167945892,
      "grad_norm": 0.7243541821125187,
      "learning_rate": 2e-05,
      "loss": 0.6050127029418946,
      "step": 9240
    },
    {
      "epoch": 0.7029409529599514,
      "grad_norm": 1.0632929919824998,
      "learning_rate": 2e-05,
      "loss": 0.5956923484802246,
      "step": 9250
    },
    {
      "epoch": 0.7037008891253135,
      "grad_norm": 1.374850375520713,
      "learning_rate": 2e-05,
      "loss": 0.5854083061218261,
      "step": 9260
    },
    {
      "epoch": 0.7044608252906756,
      "grad_norm": 0.7221313286668255,
      "learning_rate": 2e-05,
      "loss": 0.5897477149963379,
      "step": 9270
    },
    {
      "epoch": 0.7052207614560377,
      "grad_norm": 23.57359269984525,
      "learning_rate": 2e-05,
      "loss": 0.5722711086273193,
      "step": 9280
    },
    {
      "epoch": 0.7059806976213998,
      "grad_norm": 1.0883840350679976,
      "learning_rate": 2e-05,
      "loss": 0.5987275600433349,
      "step": 9290
    },
    {
      "epoch": 0.7067406337867619,
      "grad_norm": 0.7971705016212886,
      "learning_rate": 2e-05,
      "loss": 0.5989093780517578,
      "step": 9300
    },
    {
      "epoch": 0.7075005699521241,
      "grad_norm": 0.9792081620025005,
      "learning_rate": 2e-05,
      "loss": 0.5943959712982178,
      "step": 9310
    },
    {
      "epoch": 0.7082605061174861,
      "grad_norm": 0.7042955607492626,
      "learning_rate": 2e-05,
      "loss": 0.6103343009948731,
      "step": 9320
    },
    {
      "epoch": 0.7090204422828482,
      "grad_norm": 0.8104227442370697,
      "learning_rate": 2e-05,
      "loss": 0.5926801204681397,
      "step": 9330
    },
    {
      "epoch": 0.7097803784482104,
      "grad_norm": 0.6819991974438884,
      "learning_rate": 2e-05,
      "loss": 0.5871978282928467,
      "step": 9340
    },
    {
      "epoch": 0.7105403146135725,
      "grad_norm": 0.7287626317128609,
      "learning_rate": 2e-05,
      "loss": 0.598819637298584,
      "step": 9350
    },
    {
      "epoch": 0.7113002507789346,
      "grad_norm": 0.6788027514100982,
      "learning_rate": 2e-05,
      "loss": 0.6007012367248535,
      "step": 9360
    },
    {
      "epoch": 0.7120601869442966,
      "grad_norm": 0.7633762810412361,
      "learning_rate": 2e-05,
      "loss": 0.5912946224212646,
      "step": 9370
    },
    {
      "epoch": 0.7128201231096588,
      "grad_norm": 0.6755573650956104,
      "learning_rate": 2e-05,
      "loss": 0.6052114009857178,
      "step": 9380
    },
    {
      "epoch": 0.7135800592750209,
      "grad_norm": 0.8011797566022275,
      "learning_rate": 2e-05,
      "loss": 0.5779260635375977,
      "step": 9390
    },
    {
      "epoch": 0.714339995440383,
      "grad_norm": 0.7883132483063823,
      "learning_rate": 2e-05,
      "loss": 0.590411901473999,
      "step": 9400
    },
    {
      "epoch": 0.7150999316057451,
      "grad_norm": 0.8873872098462603,
      "learning_rate": 2e-05,
      "loss": 0.6128024101257324,
      "step": 9410
    },
    {
      "epoch": 0.7158598677711072,
      "grad_norm": 18.59723989829583,
      "learning_rate": 2e-05,
      "loss": 0.5695260047912598,
      "step": 9420
    },
    {
      "epoch": 0.7166198039364693,
      "grad_norm": 0.7300172182285544,
      "learning_rate": 2e-05,
      "loss": 0.6043861389160157,
      "step": 9430
    },
    {
      "epoch": 0.7173797401018315,
      "grad_norm": 0.7894404598714977,
      "learning_rate": 2e-05,
      "loss": 0.584445858001709,
      "step": 9440
    },
    {
      "epoch": 0.7181396762671935,
      "grad_norm": 1.0115309726092288,
      "learning_rate": 2e-05,
      "loss": 0.5953572750091553,
      "step": 9450
    },
    {
      "epoch": 0.7188996124325556,
      "grad_norm": 0.893292337254974,
      "learning_rate": 2e-05,
      "loss": 0.5781230926513672,
      "step": 9460
    },
    {
      "epoch": 0.7196595485979178,
      "grad_norm": 0.8212756194183854,
      "learning_rate": 2e-05,
      "loss": 0.5968575954437256,
      "step": 9470
    },
    {
      "epoch": 0.7204194847632799,
      "grad_norm": 0.7863741840290853,
      "learning_rate": 2e-05,
      "loss": 0.58509840965271,
      "step": 9480
    },
    {
      "epoch": 0.721179420928642,
      "grad_norm": 0.9616354205846387,
      "learning_rate": 2e-05,
      "loss": 0.5720963954925538,
      "step": 9490
    },
    {
      "epoch": 0.7219393570940041,
      "grad_norm": 0.7595195561819638,
      "learning_rate": 2e-05,
      "loss": 0.5851184844970703,
      "step": 9500
    },
    {
      "epoch": 0.7226992932593662,
      "grad_norm": 0.7295491876745545,
      "learning_rate": 2e-05,
      "loss": 0.5861953735351563,
      "step": 9510
    },
    {
      "epoch": 0.7234592294247283,
      "grad_norm": 0.8040338778403924,
      "learning_rate": 2e-05,
      "loss": 0.5695897102355957,
      "step": 9520
    },
    {
      "epoch": 0.7242191655900905,
      "grad_norm": 0.8094172336684529,
      "learning_rate": 2e-05,
      "loss": 0.6031304836273194,
      "step": 9530
    },
    {
      "epoch": 0.7249791017554525,
      "grad_norm": 2.5419745126286073,
      "learning_rate": 2e-05,
      "loss": 0.5888889312744141,
      "step": 9540
    },
    {
      "epoch": 0.7257390379208146,
      "grad_norm": 1.1606242329672367,
      "learning_rate": 2e-05,
      "loss": 0.590380859375,
      "step": 9550
    },
    {
      "epoch": 0.7264989740861768,
      "grad_norm": 18.361824284046193,
      "learning_rate": 2e-05,
      "loss": 0.5910569667816162,
      "step": 9560
    },
    {
      "epoch": 0.7272589102515389,
      "grad_norm": 1.09219304027433,
      "learning_rate": 2e-05,
      "loss": 0.6107835292816162,
      "step": 9570
    },
    {
      "epoch": 0.728018846416901,
      "grad_norm": 0.8576805408647956,
      "learning_rate": 2e-05,
      "loss": 0.6007575035095215,
      "step": 9580
    },
    {
      "epoch": 0.7287787825822631,
      "grad_norm": 1.1167815333724835,
      "learning_rate": 2e-05,
      "loss": 0.6096333503723145,
      "step": 9590
    },
    {
      "epoch": 0.7295387187476252,
      "grad_norm": 4.979712681127174,
      "learning_rate": 2e-05,
      "loss": 0.6010827541351318,
      "step": 9600
    },
    {
      "epoch": 0.7302986549129873,
      "grad_norm": 0.9843331470611454,
      "learning_rate": 2e-05,
      "loss": 0.6048556804656983,
      "step": 9610
    },
    {
      "epoch": 0.7310585910783495,
      "grad_norm": 0.8214268915474056,
      "learning_rate": 2e-05,
      "loss": 0.5980334758758545,
      "step": 9620
    },
    {
      "epoch": 0.7318185272437115,
      "grad_norm": 0.9683571733832363,
      "learning_rate": 2e-05,
      "loss": 0.5767800331115722,
      "step": 9630
    },
    {
      "epoch": 0.7325784634090736,
      "grad_norm": 0.9321615960927353,
      "learning_rate": 2e-05,
      "loss": 0.5745559692382812,
      "step": 9640
    },
    {
      "epoch": 0.7333383995744357,
      "grad_norm": 0.8878222351004416,
      "learning_rate": 2e-05,
      "loss": 0.5796699523925781,
      "step": 9650
    },
    {
      "epoch": 0.7340983357397979,
      "grad_norm": 0.8362763355643507,
      "learning_rate": 2e-05,
      "loss": 0.5799665451049805,
      "step": 9660
    },
    {
      "epoch": 0.73485827190516,
      "grad_norm": 0.8385921599781598,
      "learning_rate": 2e-05,
      "loss": 0.5985394954681397,
      "step": 9670
    },
    {
      "epoch": 0.735618208070522,
      "grad_norm": 1.12480346608128,
      "learning_rate": 2e-05,
      "loss": 0.5816313743591308,
      "step": 9680
    },
    {
      "epoch": 0.7363781442358842,
      "grad_norm": 1.249017212189627,
      "learning_rate": 2e-05,
      "loss": 0.6003907680511474,
      "step": 9690
    },
    {
      "epoch": 0.7371380804012463,
      "grad_norm": 0.7825844939466228,
      "learning_rate": 2e-05,
      "loss": 0.5997039794921875,
      "step": 9700
    },
    {
      "epoch": 0.7378980165666084,
      "grad_norm": 0.8561760239448466,
      "learning_rate": 2e-05,
      "loss": 0.5665740966796875,
      "step": 9710
    },
    {
      "epoch": 0.7386579527319705,
      "grad_norm": 1.0303516869708387,
      "learning_rate": 2e-05,
      "loss": 0.5872193813323975,
      "step": 9720
    },
    {
      "epoch": 0.7394178888973326,
      "grad_norm": 0.926015634002654,
      "learning_rate": 2e-05,
      "loss": 0.6007626533508301,
      "step": 9730
    },
    {
      "epoch": 0.7401778250626947,
      "grad_norm": 2.9914247397055593,
      "learning_rate": 2e-05,
      "loss": 0.5918451309204101,
      "step": 9740
    },
    {
      "epoch": 0.7409377612280569,
      "grad_norm": 2.7368773847562853,
      "learning_rate": 2e-05,
      "loss": 0.5935689449310303,
      "step": 9750
    },
    {
      "epoch": 0.741697697393419,
      "grad_norm": 0.6924396404490831,
      "learning_rate": 2e-05,
      "loss": 0.5899127006530762,
      "step": 9760
    },
    {
      "epoch": 0.742457633558781,
      "grad_norm": 2.6549426500337105,
      "learning_rate": 2e-05,
      "loss": 0.5835705757141113,
      "step": 9770
    },
    {
      "epoch": 0.7432175697241432,
      "grad_norm": 1.7501049522367593,
      "learning_rate": 2e-05,
      "loss": 0.5979074478149414,
      "step": 9780
    },
    {
      "epoch": 0.7439775058895053,
      "grad_norm": 0.8367582976059842,
      "learning_rate": 2e-05,
      "loss": 0.5802057266235352,
      "step": 9790
    },
    {
      "epoch": 0.7447374420548674,
      "grad_norm": 0.8335789237910497,
      "learning_rate": 2e-05,
      "loss": 0.5661105155944824,
      "step": 9800
    },
    {
      "epoch": 0.7454973782202295,
      "grad_norm": 1.6342560787598892,
      "learning_rate": 2e-05,
      "loss": 0.6045132637023926,
      "step": 9810
    },
    {
      "epoch": 0.7462573143855916,
      "grad_norm": 1.3226898558438709,
      "learning_rate": 2e-05,
      "loss": 0.6144696712493897,
      "step": 9820
    },
    {
      "epoch": 0.7470172505509537,
      "grad_norm": 1.8045418573202432,
      "learning_rate": 2e-05,
      "loss": 0.6055825710296631,
      "step": 9830
    },
    {
      "epoch": 0.7477771867163159,
      "grad_norm": 0.7933992888980399,
      "learning_rate": 2e-05,
      "loss": 0.5821048736572265,
      "step": 9840
    },
    {
      "epoch": 0.748537122881678,
      "grad_norm": 0.9337321204867726,
      "learning_rate": 2e-05,
      "loss": 0.5881669521331787,
      "step": 9850
    },
    {
      "epoch": 0.74929705904704,
      "grad_norm": 0.9451427050115224,
      "learning_rate": 2e-05,
      "loss": 0.5911511898040771,
      "step": 9860
    },
    {
      "epoch": 0.7500569952124022,
      "grad_norm": 1.3519910809659763,
      "learning_rate": 2e-05,
      "loss": 0.5648252010345459,
      "step": 9870
    },
    {
      "epoch": 0.7508169313777643,
      "grad_norm": 0.7746135385103889,
      "learning_rate": 2e-05,
      "loss": 0.5962317943572998,
      "step": 9880
    },
    {
      "epoch": 0.7515768675431264,
      "grad_norm": 1.0776755762215189,
      "learning_rate": 2e-05,
      "loss": 0.5946528434753418,
      "step": 9890
    },
    {
      "epoch": 0.7523368037084884,
      "grad_norm": 0.7408841645012999,
      "learning_rate": 2e-05,
      "loss": 0.6039773941040039,
      "step": 9900
    },
    {
      "epoch": 0.7530967398738506,
      "grad_norm": 0.7093832981083452,
      "learning_rate": 2e-05,
      "loss": 0.6023799896240234,
      "step": 9910
    },
    {
      "epoch": 0.7538566760392127,
      "grad_norm": 0.8213584800796925,
      "learning_rate": 2e-05,
      "loss": 0.572946310043335,
      "step": 9920
    },
    {
      "epoch": 0.7546166122045748,
      "grad_norm": 2.8545977748363334,
      "learning_rate": 2e-05,
      "loss": 0.5971200942993165,
      "step": 9930
    },
    {
      "epoch": 0.755376548369937,
      "grad_norm": 0.7039184804387227,
      "learning_rate": 2e-05,
      "loss": 0.5836564064025879,
      "step": 9940
    },
    {
      "epoch": 0.756136484535299,
      "grad_norm": 0.8076278083268315,
      "learning_rate": 2e-05,
      "loss": 0.6002474784851074,
      "step": 9950
    },
    {
      "epoch": 0.7568964207006611,
      "grad_norm": 0.9561737028305155,
      "learning_rate": 2e-05,
      "loss": 0.5977529048919678,
      "step": 9960
    },
    {
      "epoch": 0.7576563568660233,
      "grad_norm": 0.7731610858845104,
      "learning_rate": 2e-05,
      "loss": 0.5851725578308106,
      "step": 9970
    },
    {
      "epoch": 0.7584162930313854,
      "grad_norm": 2.799435244223758,
      "learning_rate": 2e-05,
      "loss": 0.5961930274963378,
      "step": 9980
    },
    {
      "epoch": 0.7591762291967474,
      "grad_norm": 1.1534924538731863,
      "learning_rate": 2e-05,
      "loss": 0.6015936851501464,
      "step": 9990
    },
    {
      "epoch": 0.7599361653621096,
      "grad_norm": 0.7672423607665784,
      "learning_rate": 2e-05,
      "loss": 0.5977615356445313,
      "step": 10000
    },
    {
      "epoch": 0.7606961015274717,
      "grad_norm": 0.8097279397022222,
      "learning_rate": 2e-05,
      "loss": 0.6081546306610107,
      "step": 10010
    },
    {
      "epoch": 0.7614560376928338,
      "grad_norm": 1.8032912671873316,
      "learning_rate": 2e-05,
      "loss": 0.5819506645202637,
      "step": 10020
    },
    {
      "epoch": 0.762215973858196,
      "grad_norm": 1.0796564656190588,
      "learning_rate": 2e-05,
      "loss": 0.5767606258392334,
      "step": 10030
    },
    {
      "epoch": 0.762975910023558,
      "grad_norm": 1.4701638211182295,
      "learning_rate": 2e-05,
      "loss": 0.574137830734253,
      "step": 10040
    },
    {
      "epoch": 0.7637358461889201,
      "grad_norm": 0.9259367838297374,
      "learning_rate": 2e-05,
      "loss": 0.581744909286499,
      "step": 10050
    },
    {
      "epoch": 0.7644957823542823,
      "grad_norm": 1.5200951960135667,
      "learning_rate": 2e-05,
      "loss": 0.6004110336303711,
      "step": 10060
    },
    {
      "epoch": 0.7652557185196444,
      "grad_norm": 3.077983385077952,
      "learning_rate": 2e-05,
      "loss": 0.5801307201385498,
      "step": 10070
    },
    {
      "epoch": 0.7660156546850064,
      "grad_norm": 3.6443586343163776,
      "learning_rate": 2e-05,
      "loss": 0.5910909652709961,
      "step": 10080
    },
    {
      "epoch": 0.7667755908503686,
      "grad_norm": 0.9808294493116338,
      "learning_rate": 2e-05,
      "loss": 0.5963793754577636,
      "step": 10090
    },
    {
      "epoch": 0.7675355270157307,
      "grad_norm": 0.9427696889577263,
      "learning_rate": 2e-05,
      "loss": 0.6042157649993897,
      "step": 10100
    },
    {
      "epoch": 0.7682954631810928,
      "grad_norm": 0.7493605271884373,
      "learning_rate": 2e-05,
      "loss": 0.5845524787902832,
      "step": 10110
    },
    {
      "epoch": 0.769055399346455,
      "grad_norm": 1.454032829758323,
      "learning_rate": 2e-05,
      "loss": 0.6082051753997803,
      "step": 10120
    },
    {
      "epoch": 0.769815335511817,
      "grad_norm": 0.8897201152433183,
      "learning_rate": 2e-05,
      "loss": 0.5973020076751709,
      "step": 10130
    },
    {
      "epoch": 0.7705752716771791,
      "grad_norm": 1.6162827818072978,
      "learning_rate": 2e-05,
      "loss": 0.593494176864624,
      "step": 10140
    },
    {
      "epoch": 0.7713352078425413,
      "grad_norm": 0.6962476243032999,
      "learning_rate": 2e-05,
      "loss": 0.5677854061126709,
      "step": 10150
    },
    {
      "epoch": 0.7720951440079034,
      "grad_norm": 0.9704449877820779,
      "learning_rate": 2e-05,
      "loss": 0.5956148147583008,
      "step": 10160
    },
    {
      "epoch": 0.7728550801732654,
      "grad_norm": 0.8777720504914048,
      "learning_rate": 2e-05,
      "loss": 0.5787041664123536,
      "step": 10170
    },
    {
      "epoch": 0.7736150163386275,
      "grad_norm": 0.7915488214518618,
      "learning_rate": 1.999957274187931e-05,
      "loss": 0.5795921087265015,
      "step": 10180
    },
    {
      "epoch": 0.7743749525039897,
      "grad_norm": 0.7162201668035736,
      "learning_rate": 1.9991978062682906e-05,
      "loss": 0.5957917213439942,
      "step": 10190
    },
    {
      "epoch": 0.7751348886693518,
      "grad_norm": 1.363720129361685,
      "learning_rate": 1.997489709987378e-05,
      "loss": 0.5646952629089356,
      "step": 10200
    },
    {
      "epoch": 0.7758948248347138,
      "grad_norm": 0.7952810818984362,
      "learning_rate": 1.9948346151502135e-05,
      "loss": 0.6080700874328613,
      "step": 10210
    },
    {
      "epoch": 0.776654761000076,
      "grad_norm": 0.7146224627906625,
      "learning_rate": 1.991235055154325e-05,
      "loss": 0.5732362270355225,
      "step": 10220
    },
    {
      "epoch": 0.7774146971654381,
      "grad_norm": 0.8341146680637818,
      "learning_rate": 1.9866944645724693e-05,
      "loss": 0.596200704574585,
      "step": 10230
    },
    {
      "epoch": 0.7781746333308002,
      "grad_norm": 0.6888923984738532,
      "learning_rate": 1.9812171758754842e-05,
      "loss": 0.584140396118164,
      "step": 10240
    },
    {
      "epoch": 0.7789345694961624,
      "grad_norm": 0.9035795939673269,
      "learning_rate": 1.9748084152983987e-05,
      "loss": 0.5781354904174805,
      "step": 10250
    },
    {
      "epoch": 0.7796945056615244,
      "grad_norm": 1.911555708156575,
      "learning_rate": 1.967474297853744e-05,
      "loss": 0.5824694633483887,
      "step": 10260
    },
    {
      "epoch": 0.7804544418268865,
      "grad_norm": 0.8260823199006383,
      "learning_rate": 1.9592218214968256e-05,
      "loss": 0.5652504920959472,
      "step": 10270
    },
    {
      "epoch": 0.7812143779922487,
      "grad_norm": 0.8387909892118044,
      "learning_rate": 1.9500588604485213e-05,
      "loss": 0.5966443061828614,
      "step": 10280
    },
    {
      "epoch": 0.7819743141576108,
      "grad_norm": 1.107665243376975,
      "learning_rate": 1.939994157681979e-05,
      "loss": 0.591804027557373,
      "step": 10290
    },
    {
      "epoch": 0.7827342503229728,
      "grad_norm": 0.771682500983752,
      "learning_rate": 1.9290373165803803e-05,
      "loss": 0.5888102531433106,
      "step": 10300
    },
    {
      "epoch": 0.783494186488335,
      "grad_norm": 0.8150045752143769,
      "learning_rate": 1.9171987917737328e-05,
      "loss": 0.5642033576965332,
      "step": 10310
    },
    {
      "epoch": 0.7842541226536971,
      "grad_norm": 0.747540459857151,
      "learning_rate": 1.9044898791634322e-05,
      "loss": 0.5836194515228271,
      "step": 10320
    },
    {
      "epoch": 0.7850140588190592,
      "grad_norm": 1.6040698991228148,
      "learning_rate": 1.8909227051441126e-05,
      "loss": 0.5972249507904053,
      "step": 10330
    },
    {
      "epoch": 0.7857739949844214,
      "grad_norm": 1.4606496224079364,
      "learning_rate": 1.8765102150330692e-05,
      "loss": 0.5893157958984375,
      "step": 10340
    },
    {
      "epoch": 0.7865339311497834,
      "grad_norm": 0.7510088383938315,
      "learning_rate": 1.861266160718297e-05,
      "loss": 0.601513671875,
      "step": 10350
    },
    {
      "epoch": 0.7872938673151455,
      "grad_norm": 0.7943003580378605,
      "learning_rate": 1.845205087536923e-05,
      "loss": 0.5728665828704834,
      "step": 10360
    },
    {
      "epoch": 0.7880538034805077,
      "grad_norm": 0.8553633814460313,
      "learning_rate": 1.8283423203965633e-05,
      "loss": 0.5976364135742187,
      "step": 10370
    },
    {
      "epoch": 0.7888137396458698,
      "grad_norm": 0.9952655315986954,
      "learning_rate": 1.810693949152835e-05,
      "loss": 0.5971133232116699,
      "step": 10380
    },
    {
      "epoch": 0.7895736758112318,
      "grad_norm": 0.8897554087765187,
      "learning_rate": 1.7922768132569902e-05,
      "loss": 0.6048148155212403,
      "step": 10390
    },
    {
      "epoch": 0.790333611976594,
      "grad_norm": 0.7386441402610299,
      "learning_rate": 1.773108485688302e-05,
      "loss": 0.5739955902099609,
      "step": 10400
    },
    {
      "epoch": 0.7910935481419561,
      "grad_norm": 1.0080694735576334,
      "learning_rate": 1.7532072561865537e-05,
      "loss": 0.5828299522399902,
      "step": 10410
    },
    {
      "epoch": 0.7918534843073182,
      "grad_norm": 1.417611075481486,
      "learning_rate": 1.732592113800616e-05,
      "loss": 0.5581233501434326,
      "step": 10420
    },
    {
      "epoch": 0.7926134204726802,
      "grad_norm": 1.5542045839467526,
      "learning_rate": 1.7112827287697668e-05,
      "loss": 0.5806106090545654,
      "step": 10430
    },
    {
      "epoch": 0.7933733566380424,
      "grad_norm": 0.7417799081469686,
      "learning_rate": 1.6892994337550452e-05,
      "loss": 0.5809941768646241,
      "step": 10440
    },
    {
      "epoch": 0.7941332928034045,
      "grad_norm": 0.7349939013000607,
      "learning_rate": 1.6666632044385487e-05,
      "loss": 0.5874391555786133,
      "step": 10450
    },
    {
      "epoch": 0.7948932289687666,
      "grad_norm": 0.7796458395942164,
      "learning_rate": 1.6433956395091757e-05,
      "loss": 0.5821702480316162,
      "step": 10460
    },
    {
      "epoch": 0.7956531651341288,
      "grad_norm": 2.6341748058720977,
      "learning_rate": 1.6195189400539238e-05,
      "loss": 0.5847431182861328,
      "step": 10470
    },
    {
      "epoch": 0.7964131012994908,
      "grad_norm": 0.7403690625748686,
      "learning_rate": 1.595055888374398e-05,
      "loss": 0.611750316619873,
      "step": 10480
    },
    {
      "epoch": 0.7971730374648529,
      "grad_norm": 0.7267586943845273,
      "learning_rate": 1.5700298262487434e-05,
      "loss": 0.5807172298431397,
      "step": 10490
    },
    {
      "epoch": 0.7979329736302151,
      "grad_norm": 0.6323243303708149,
      "learning_rate": 1.5444646326597467e-05,
      "loss": 0.5631939888000488,
      "step": 10500
    },
    {
      "epoch": 0.7986929097955772,
      "grad_norm": 1.6718561175584437,
      "learning_rate": 1.5183847010103586e-05,
      "loss": 0.5870999336242676,
      "step": 10510
    },
    {
      "epoch": 0.7994528459609392,
      "grad_norm": 0.6967536963863697,
      "learning_rate": 1.4918149158483686e-05,
      "loss": 0.5667330741882324,
      "step": 10520
    },
    {
      "epoch": 0.8002127821263014,
      "grad_norm": 0.6762890616145982,
      "learning_rate": 1.464780629122452e-05,
      "loss": 0.5975906372070312,
      "step": 10530
    },
    {
      "epoch": 0.8009727182916635,
      "grad_norm": 0.838437467833338,
      "learning_rate": 1.4373076359922374e-05,
      "loss": 0.5782775402069091,
      "step": 10540
    },
    {
      "epoch": 0.8017326544570256,
      "grad_norm": 1.430787622476739,
      "learning_rate": 1.4094221502154774e-05,
      "loss": 0.5730947017669678,
      "step": 10550
    },
    {
      "epoch": 0.8024925906223878,
      "grad_norm": 0.6366097228939659,
      "learning_rate": 1.3811507791358113e-05,
      "loss": 0.5719327449798584,
      "step": 10560
    },
    {
      "epoch": 0.8032525267877498,
      "grad_norm": 0.7347087172168436,
      "learning_rate": 1.3525204982949753e-05,
      "loss": 0.5822739601135254,
      "step": 10570
    },
    {
      "epoch": 0.8040124629531119,
      "grad_norm": 0.7647439272344558,
      "learning_rate": 1.3235586256936996e-05,
      "loss": 0.5633986473083497,
      "step": 10580
    },
    {
      "epoch": 0.8047723991184741,
      "grad_norm": 0.886324919431922,
      "learning_rate": 1.294292795725836e-05,
      "loss": 0.5658357620239258,
      "step": 10590
    },
    {
      "epoch": 0.8055323352838362,
      "grad_norm": 0.6411029650242228,
      "learning_rate": 1.2647509328105978e-05,
      "loss": 0.5738885879516602,
      "step": 10600
    },
    {
      "epoch": 0.8062922714491982,
      "grad_norm": 2.5062846548006106,
      "learning_rate": 1.2349612247480686e-05,
      "loss": 0.5832395553588867,
      "step": 10610
    },
    {
      "epoch": 0.8070522076145604,
      "grad_norm": 1.2826174258506133,
      "learning_rate": 1.2049520958234e-05,
      "loss": 0.5865778923034668,
      "step": 10620
    },
    {
      "epoch": 0.8078121437799225,
      "grad_norm": 6.231127518799065,
      "learning_rate": 1.1747521796853638e-05,
      "loss": 0.5781538963317872,
      "step": 10630
    },
    {
      "epoch": 0.8085720799452846,
      "grad_norm": 1.0664460594578196,
      "learning_rate": 1.1443902920251395e-05,
      "loss": 0.5603087425231934,
      "step": 10640
    },
    {
      "epoch": 0.8093320161106468,
      "grad_norm": 0.8265217577159412,
      "learning_rate": 1.1138954030814009e-05,
      "loss": 0.5624680519104004,
      "step": 10650
    },
    {
      "epoch": 0.8100919522760088,
      "grad_norm": 0.7642221601871532,
      "learning_rate": 1.0832966099979388e-05,
      "loss": 0.59715576171875,
      "step": 10660
    },
    {
      "epoch": 0.8108518884413709,
      "grad_norm": 0.5888184298497028,
      "learning_rate": 1.0526231090601969e-05,
      "loss": 0.5589974880218506,
      "step": 10670
    },
    {
      "epoch": 0.8116118246067331,
      "grad_norm": 3.96401233327766,
      "learning_rate": 1.0219041678372106e-05,
      "loss": 0.5676442623138428,
      "step": 10680
    },
    {
      "epoch": 0.8123717607720952,
      "grad_norm": 0.8158259967590461,
      "learning_rate": 9.911690972555268e-06,
      "loss": 0.570868444442749,
      "step": 10690
    },
    {
      "epoch": 0.8131316969374572,
      "grad_norm": 23.999709710074026,
      "learning_rate": 9.604472236317541e-06,
      "loss": 0.57431058883667,
      "step": 10700
    },
    {
      "epoch": 0.8138916331028193,
      "grad_norm": 1.3364035735150577,
      "learning_rate": 9.297678606904335e-06,
      "loss": 0.5814786434173584,
      "step": 10710
    },
    {
      "epoch": 0.8146515692681815,
      "grad_norm": 0.7234190820046779,
      "learning_rate": 8.991602815939155e-06,
      "loss": 0.5634496212005615,
      "step": 10720
    },
    {
      "epoch": 0.8154115054335436,
      "grad_norm": 0.62833004089418,
      "learning_rate": 8.686536910109464e-06,
      "loss": 0.5674325466156006,
      "step": 10730
    },
    {
      "epoch": 0.8161714415989056,
      "grad_norm": 1.1569815963489893,
      "learning_rate": 8.382771972506084e-06,
      "loss": 0.590165090560913,
      "step": 10740
    },
    {
      "epoch": 0.8169313777642678,
      "grad_norm": 0.7133790563695029,
      "learning_rate": 8.08059784488198e-06,
      "loss": 0.5609392166137696,
      "step": 10750
    },
    {
      "epoch": 0.8176913139296299,
      "grad_norm": 0.6372574532600473,
      "learning_rate": 7.780302851095544e-06,
      "loss": 0.5638291358947753,
      "step": 10760
    },
    {
      "epoch": 0.818451250094992,
      "grad_norm": 0.8482001833586504,
      "learning_rate": 7.482173522002166e-06,
      "loss": 0.5853194236755371,
      "step": 10770
    },
    {
      "epoch": 0.8192111862603542,
      "grad_norm": 0.7772223771772502,
      "learning_rate": 7.186494322056624e-06,
      "loss": 0.5519120693206787,
      "step": 10780
    },
    {
      "epoch": 0.8199711224257162,
      "grad_norm": 0.831097881781595,
      "learning_rate": 6.893547377887183e-06,
      "loss": 0.5676751136779785,
      "step": 10790
    },
    {
      "epoch": 0.8207310585910783,
      "grad_norm": 0.70779953712275,
      "learning_rate": 6.6036122091003365e-06,
      "loss": 0.5723977088928223,
      "step": 10800
    },
    {
      "epoch": 0.8214909947564405,
      "grad_norm": 0.6815858575419211,
      "learning_rate": 6.316965461573097e-06,
      "loss": 0.5617259979248047,
      "step": 10810
    },
    {
      "epoch": 0.8222509309218026,
      "grad_norm": 0.7495880859155738,
      "learning_rate": 6.0338806434872965e-06,
      "loss": 0.5690215587615967,
      "step": 10820
    },
    {
      "epoch": 0.8230108670871646,
      "grad_norm": 0.6924096824789082,
      "learning_rate": 5.754627864357749e-06,
      "loss": 0.5699169158935546,
      "step": 10830
    },
    {
      "epoch": 0.8237708032525268,
      "grad_norm": 0.6757236463868339,
      "learning_rate": 5.479473577303319e-06,
      "loss": 0.5584075927734375,
      "step": 10840
    },
    {
      "epoch": 0.8245307394178889,
      "grad_norm": 1.9546263670160724,
      "learning_rate": 5.208680324806778e-06,
      "loss": 0.5769439697265625,
      "step": 10850
    },
    {
      "epoch": 0.825290675583251,
      "grad_norm": 0.7229267226648866,
      "learning_rate": 4.94250648820606e-06,
      "loss": 0.5496849060058594,
      "step": 10860
    },
    {
      "epoch": 0.8260506117486132,
      "grad_norm": 0.7253892662772315,
      "learning_rate": 4.681206041155951e-06,
      "loss": 0.5594316005706788,
      "step": 10870
    },
    {
      "epoch": 0.8268105479139752,
      "grad_norm": 1.0413558278549726,
      "learning_rate": 4.42502830729538e-06,
      "loss": 0.567435359954834,
      "step": 10880
    },
    {
      "epoch": 0.8275704840793373,
      "grad_norm": 0.6711320461186329,
      "learning_rate": 4.174217722351669e-06,
      "loss": 0.5450252532958985,
      "step": 10890
    },
    {
      "epoch": 0.8283304202446995,
      "grad_norm": 0.8848291439769558,
      "learning_rate": 3.929013600908575e-06,
      "loss": 0.5831691741943359,
      "step": 10900
    },
    {
      "epoch": 0.8290903564100616,
      "grad_norm": 1.1812413266265367,
      "learning_rate": 3.689649908060807e-06,
      "loss": 0.5639975547790528,
      "step": 10910
    },
    {
      "epoch": 0.8298502925754236,
      "grad_norm": 0.6178555691536994,
      "learning_rate": 3.4563550361728028e-06,
      "loss": 0.5978845119476318,
      "step": 10920
    },
    {
      "epoch": 0.8306102287407858,
      "grad_norm": 0.6501575632031658,
      "learning_rate": 3.2293515869548216e-06,
      "loss": 0.5772077083587647,
      "step": 10930
    },
    {
      "epoch": 0.8313701649061479,
      "grad_norm": 0.7221853133984486,
      "learning_rate": 3.0088561590642746e-06,
      "loss": 0.5637149333953857,
      "step": 10940
    },
    {
      "epoch": 0.83213010107151,
      "grad_norm": 0.6163544375199015,
      "learning_rate": 2.795079141434941e-06,
      "loss": 0.5388998031616211,
      "step": 10950
    },
    {
      "epoch": 0.832890037236872,
      "grad_norm": 0.726274027708327,
      "learning_rate": 2.588224512531299e-06,
      "loss": 0.5641888618469239,
      "step": 10960
    },
    {
      "epoch": 0.8336499734022342,
      "grad_norm": 1.0672493885924672,
      "learning_rate": 2.388489645719495e-06,
      "loss": 0.5616416931152344,
      "step": 10970
    },
    {
      "epoch": 0.8344099095675963,
      "grad_norm": 0.9518962216886122,
      "learning_rate": 2.196065120940663e-06,
      "loss": 0.5582421779632568,
      "step": 10980
    },
    {
      "epoch": 0.8351698457329584,
      "grad_norm": 0.7339816436712041,
      "learning_rate": 2.0111345428662808e-06,
      "loss": 0.5738423347473145,
      "step": 10990
    },
    {
      "epoch": 0.8359297818983206,
      "grad_norm": 0.6585927039011988,
      "learning_rate": 1.8338743657091102e-06,
      "loss": 0.5529128074645996,
      "step": 11000
    },
    {
      "epoch": 0.8366897180636826,
      "grad_norm": 0.7452079924281754,
      "learning_rate": 1.6644537248568113e-06,
      "loss": 0.5621377944946289,
      "step": 11010
    },
    {
      "epoch": 0.8374496542290447,
      "grad_norm": 1.0972462249963206,
      "learning_rate": 1.503034275488951e-06,
      "loss": 0.560430908203125,
      "step": 11020
    },
    {
      "epoch": 0.8382095903944069,
      "grad_norm": 0.6799769698720262,
      "learning_rate": 1.3497700383313495e-06,
      "loss": 0.5330728054046631,
      "step": 11030
    },
    {
      "epoch": 0.838969526559769,
      "grad_norm": 0.6197300003459382,
      "learning_rate": 1.2048072526949562e-06,
      "loss": 0.5601110458374023,
      "step": 11040
    },
    {
      "epoch": 0.839729462725131,
      "grad_norm": 0.8359662320993213,
      "learning_rate": 1.0682842369394756e-06,
      "loss": 0.5462487697601318,
      "step": 11050
    },
    {
      "epoch": 0.8404893988904932,
      "grad_norm": 1.6368192169683355,
      "learning_rate": 9.403312564948792e-07,
      "loss": 0.5276519775390625,
      "step": 11060
    },
    {
      "epoch": 0.8412493350558553,
      "grad_norm": 0.7634105146392935,
      "learning_rate": 8.210703995667471e-07,
      "loss": 0.5467713832855224,
      "step": 11070
    },
    {
      "epoch": 0.8420092712212174,
      "grad_norm": 0.8474274898277121,
      "learning_rate": 7.106154606440245e-07,
      "loss": 0.5679131984710694,
      "step": 11080
    },
    {
      "epoch": 0.8427692073865796,
      "grad_norm": 0.7946111381793197,
      "learning_rate": 6.090718319203393e-07,
      "loss": 0.5504266738891601,
      "step": 11090
    },
    {
      "epoch": 0.8435291435519416,
      "grad_norm": 1.2140907047518703,
      "learning_rate": 5.165364027325089e-07,
      "loss": 0.5715602397918701,
      "step": 11100
    },
    {
      "epoch": 0.8442890797173037,
      "grad_norm": 0.7251009663252196,
      "learning_rate": 4.330974671121585e-07,
      "loss": 0.5597578048706054,
      "step": 11110
    },
    {
      "epoch": 0.8450490158826659,
      "grad_norm": 0.7592576547583885,
      "learning_rate": 3.588346395386846e-07,
      "loss": 0.5609750270843505,
      "step": 11120
    },
    {
      "epoch": 0.845808952048028,
      "grad_norm": 0.7884820727041193,
      "learning_rate": 2.938187789739395e-07,
      "loss": 0.5577229976654052,
      "step": 11130
    },
    {
      "epoch": 0.84656888821339,
      "grad_norm": 1.1667086958727189,
      "learning_rate": 2.3811192125110724e-07,
      "loss": 0.5567448616027832,
      "step": 11140
    },
    {
      "epoch": 0.8473288243787522,
      "grad_norm": 0.8900532225497632,
      "learning_rate": 1.917672198823177e-07,
      "loss": 0.5545243263244629,
      "step": 11150
    },
    {
      "epoch": 0.8480887605441143,
      "grad_norm": 0.6926253372820196,
      "learning_rate": 1.5482889534143795e-07,
      "loss": 0.5580403327941894,
      "step": 11160
    },
    {
      "epoch": 0.8488486967094764,
      "grad_norm": 0.8310531287401584,
      "learning_rate": 1.273321928704694e-07,
      "loss": 0.5566168785095215,
      "step": 11170
    },
    {
      "epoch": 0.8496086328748386,
      "grad_norm": 0.6965507468341606,
      "learning_rate": 1.0930334884977866e-07,
      "loss": 0.5642939567565918,
      "step": 11180
    },
    {
      "epoch": 0.8503685690402006,
      "grad_norm": 0.8573097119615448,
      "learning_rate": 1.0075956576427487e-07,
      "loss": 0.5571513175964355,
      "step": 11190
    },
    {
      "epoch": 0.8505965498898093,
      "step": 11193,
      "total_flos": 5300110976090112.0,
      "train_loss": 0.05198442630278759,
      "train_runtime": 14772.7657,
      "train_samples_per_second": 9.092,
      "train_steps_per_second": 0.758
    }
  ],
  "logging_steps": 10,
  "max_steps": 11193,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 636,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5300110976090112.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}