{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 2144,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0023320895522388058,
      "grad_norm": 2.1731144958946356,
      "learning_rate": 2.3148148148148148e-06,
      "loss": 0.8497,
      "step": 5
    },
    {
      "epoch": 0.0046641791044776115,
      "grad_norm": 1.4455418381410732,
      "learning_rate": 4.6296296296296296e-06,
      "loss": 0.8271,
      "step": 10
    },
    {
      "epoch": 0.006996268656716418,
      "grad_norm": 1.6554188288129872,
      "learning_rate": 6.944444444444445e-06,
      "loss": 0.816,
      "step": 15
    },
    {
      "epoch": 0.009328358208955223,
      "grad_norm": 1.0605788457441088,
      "learning_rate": 9.259259259259259e-06,
      "loss": 0.7542,
      "step": 20
    },
    {
      "epoch": 0.01166044776119403,
      "grad_norm": 0.720337911055749,
      "learning_rate": 1.1574074074074075e-05,
      "loss": 0.7312,
      "step": 25
    },
    {
      "epoch": 0.013992537313432836,
      "grad_norm": 0.6053048303655763,
      "learning_rate": 1.388888888888889e-05,
      "loss": 0.6962,
      "step": 30
    },
    {
      "epoch": 0.01632462686567164,
      "grad_norm": 0.5365339422370746,
      "learning_rate": 1.6203703703703704e-05,
      "loss": 0.6915,
      "step": 35
    },
    {
      "epoch": 0.018656716417910446,
      "grad_norm": 0.5138446295825095,
      "learning_rate": 1.8518518518518518e-05,
      "loss": 0.6777,
      "step": 40
    },
    {
      "epoch": 0.020988805970149255,
      "grad_norm": 0.5294732057067889,
      "learning_rate": 2.0833333333333336e-05,
      "loss": 0.6738,
      "step": 45
    },
    {
      "epoch": 0.02332089552238806,
      "grad_norm": 0.5216249762051252,
      "learning_rate": 2.314814814814815e-05,
      "loss": 0.649,
      "step": 50
    },
    {
      "epoch": 0.025652985074626867,
      "grad_norm": 0.4635569420456706,
      "learning_rate": 2.5462962962962965e-05,
      "loss": 0.6445,
      "step": 55
    },
    {
      "epoch": 0.027985074626865673,
      "grad_norm": 0.5711576603157688,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.6442,
      "step": 60
    },
    {
      "epoch": 0.03031716417910448,
      "grad_norm": 0.5406532747721863,
      "learning_rate": 3.0092592592592593e-05,
      "loss": 0.6333,
      "step": 65
    },
    {
      "epoch": 0.03264925373134328,
      "grad_norm": 0.4921611180106655,
      "learning_rate": 3.240740740740741e-05,
      "loss": 0.609,
      "step": 70
    },
    {
      "epoch": 0.034981343283582086,
      "grad_norm": 0.5585104564168789,
      "learning_rate": 3.472222222222222e-05,
      "loss": 0.6166,
      "step": 75
    },
    {
      "epoch": 0.03731343283582089,
      "grad_norm": 0.6060789427946596,
      "learning_rate": 3.7037037037037037e-05,
      "loss": 0.6241,
      "step": 80
    },
    {
      "epoch": 0.039645522388059705,
      "grad_norm": 0.5313035000332463,
      "learning_rate": 3.935185185185186e-05,
      "loss": 0.6184,
      "step": 85
    },
    {
      "epoch": 0.04197761194029851,
      "grad_norm": 0.6556913799681943,
      "learning_rate": 4.166666666666667e-05,
      "loss": 0.6211,
      "step": 90
    },
    {
      "epoch": 0.044309701492537316,
      "grad_norm": 0.5778296341723107,
      "learning_rate": 4.3981481481481486e-05,
      "loss": 0.6079,
      "step": 95
    },
    {
      "epoch": 0.04664179104477612,
      "grad_norm": 0.5268598973302592,
      "learning_rate": 4.62962962962963e-05,
      "loss": 0.6038,
      "step": 100
    },
    {
      "epoch": 0.04897388059701493,
      "grad_norm": 0.509339973453908,
      "learning_rate": 4.8611111111111115e-05,
      "loss": 0.6305,
      "step": 105
    },
    {
      "epoch": 0.051305970149253734,
      "grad_norm": 0.5496370287982213,
      "learning_rate": 4.999989285883431e-05,
      "loss": 0.627,
      "step": 110
    },
    {
      "epoch": 0.05363805970149254,
      "grad_norm": 0.5434194242110924,
      "learning_rate": 4.999868753243865e-05,
      "loss": 0.6263,
      "step": 115
    },
    {
      "epoch": 0.055970149253731345,
      "grad_norm": 0.6282164391718019,
      "learning_rate": 4.999614302517356e-05,
      "loss": 0.6202,
      "step": 120
    },
    {
      "epoch": 0.05830223880597015,
      "grad_norm": 0.5814513203649924,
      "learning_rate": 4.9992259488494784e-05,
      "loss": 0.6049,
      "step": 125
    },
    {
      "epoch": 0.06063432835820896,
      "grad_norm": 0.6315852165763514,
      "learning_rate": 4.99870371535606e-05,
      "loss": 0.6043,
      "step": 130
    },
    {
      "epoch": 0.06296641791044776,
      "grad_norm": 0.6282922843004739,
      "learning_rate": 4.9980476331218066e-05,
      "loss": 0.6484,
      "step": 135
    },
    {
      "epoch": 0.06529850746268656,
      "grad_norm": 0.5875693979221333,
      "learning_rate": 4.997257741198456e-05,
      "loss": 0.6189,
      "step": 140
    },
    {
      "epoch": 0.06763059701492537,
      "grad_norm": 0.6058703738937499,
      "learning_rate": 4.996334086602444e-05,
      "loss": 0.6036,
      "step": 145
    },
    {
      "epoch": 0.06996268656716417,
      "grad_norm": 0.5426212351193616,
      "learning_rate": 4.9952767243121146e-05,
      "loss": 0.619,
      "step": 150
    },
    {
      "epoch": 0.07229477611940298,
      "grad_norm": 0.5276707044683704,
      "learning_rate": 4.994085717264443e-05,
      "loss": 0.6042,
      "step": 155
    },
    {
      "epoch": 0.07462686567164178,
      "grad_norm": 0.48342313898689165,
      "learning_rate": 4.992761136351291e-05,
      "loss": 0.6084,
      "step": 160
    },
    {
      "epoch": 0.07695895522388059,
      "grad_norm": 0.43727574748121845,
      "learning_rate": 4.99130306041519e-05,
      "loss": 0.6086,
      "step": 165
    },
    {
      "epoch": 0.07929104477611941,
      "grad_norm": 0.5139967015316651,
      "learning_rate": 4.989711576244639e-05,
      "loss": 0.5885,
      "step": 170
    },
    {
      "epoch": 0.08162313432835822,
      "grad_norm": 0.5169853218760131,
      "learning_rate": 4.9879867785689506e-05,
      "loss": 0.5792,
      "step": 175
    },
    {
      "epoch": 0.08395522388059702,
      "grad_norm": 0.4547424820277734,
      "learning_rate": 4.986128770052603e-05,
      "loss": 0.5979,
      "step": 180
    },
    {
      "epoch": 0.08628731343283583,
      "grad_norm": 0.5278283023520385,
      "learning_rate": 4.9841376612891365e-05,
      "loss": 0.5921,
      "step": 185
    },
    {
      "epoch": 0.08861940298507463,
      "grad_norm": 0.4592914703740021,
      "learning_rate": 4.9820135707945634e-05,
      "loss": 0.5895,
      "step": 190
    },
    {
      "epoch": 0.09095149253731344,
      "grad_norm": 0.5604812331209644,
      "learning_rate": 4.97975662500032e-05,
      "loss": 0.6123,
      "step": 195
    },
    {
      "epoch": 0.09328358208955224,
      "grad_norm": 0.4963710392742468,
      "learning_rate": 4.9773669582457364e-05,
      "loss": 0.5949,
      "step": 200
    },
    {
      "epoch": 0.09561567164179105,
      "grad_norm": 0.5341113967257587,
      "learning_rate": 4.974844712770044e-05,
      "loss": 0.6115,
      "step": 205
    },
    {
      "epoch": 0.09794776119402986,
      "grad_norm": 0.5356900218427062,
      "learning_rate": 4.972190038703905e-05,
      "loss": 0.6062,
      "step": 210
    },
    {
      "epoch": 0.10027985074626866,
      "grad_norm": 0.48271662493768214,
      "learning_rate": 4.96940309406048e-05,
      "loss": 0.5994,
      "step": 215
    },
    {
      "epoch": 0.10261194029850747,
      "grad_norm": 0.4656684615513963,
      "learning_rate": 4.966484044726024e-05,
      "loss": 0.5919,
      "step": 220
    },
    {
      "epoch": 0.10494402985074627,
      "grad_norm": 0.45763728379025154,
      "learning_rate": 4.963433064450002e-05,
      "loss": 0.5897,
      "step": 225
    },
    {
      "epoch": 0.10727611940298508,
      "grad_norm": 0.4790657281481012,
      "learning_rate": 4.9602503348347625e-05,
      "loss": 0.5823,
      "step": 230
    },
    {
      "epoch": 0.10960820895522388,
      "grad_norm": 0.49515386205808626,
      "learning_rate": 4.956936045324716e-05,
      "loss": 0.5709,
      "step": 235
    },
    {
      "epoch": 0.11194029850746269,
      "grad_norm": 0.5030427108835758,
      "learning_rate": 4.953490393195063e-05,
      "loss": 0.6161,
      "step": 240
    },
    {
      "epoch": 0.1142723880597015,
      "grad_norm": 0.45710707251035176,
      "learning_rate": 4.9499135835400526e-05,
      "loss": 0.5983,
      "step": 245
    },
    {
      "epoch": 0.1166044776119403,
      "grad_norm": 0.4321697797530169,
      "learning_rate": 4.9462058292607735e-05,
      "loss": 0.5831,
      "step": 250
    },
    {
      "epoch": 0.11893656716417911,
      "grad_norm": 0.48082763546682483,
      "learning_rate": 4.942367351052481e-05,
      "loss": 0.5788,
      "step": 255
    },
    {
      "epoch": 0.12126865671641791,
      "grad_norm": 0.4447073632938014,
      "learning_rate": 4.938398377391461e-05,
      "loss": 0.5882,
      "step": 260
    },
    {
      "epoch": 0.12360074626865672,
      "grad_norm": 0.42261904193138444,
      "learning_rate": 4.9342991445214334e-05,
      "loss": 0.5844,
      "step": 265
    },
    {
      "epoch": 0.1259328358208955,
      "grad_norm": 0.4455650933452113,
      "learning_rate": 4.930069896439485e-05,
      "loss": 0.5829,
      "step": 270
    },
    {
      "epoch": 0.12826492537313433,
      "grad_norm": 0.4101033046902724,
      "learning_rate": 4.925710884881551e-05,
      "loss": 0.5796,
      "step": 275
    },
    {
      "epoch": 0.13059701492537312,
      "grad_norm": 0.48901968042593585,
      "learning_rate": 4.921222369307427e-05,
      "loss": 0.5914,
      "step": 280
    },
    {
      "epoch": 0.13292910447761194,
      "grad_norm": 0.4900379168491669,
      "learning_rate": 4.916604616885328e-05,
      "loss": 0.5865,
      "step": 285
    },
    {
      "epoch": 0.13526119402985073,
      "grad_norm": 0.4823958325202437,
      "learning_rate": 4.9118579024759854e-05,
      "loss": 0.5901,
      "step": 290
    },
    {
      "epoch": 0.13759328358208955,
      "grad_norm": 0.4431212546352274,
      "learning_rate": 4.9069825086162865e-05,
      "loss": 0.568,
      "step": 295
    },
    {
      "epoch": 0.13992537313432835,
      "grad_norm": 0.43555586427340276,
      "learning_rate": 4.901978725502454e-05,
      "loss": 0.5751,
      "step": 300
    },
    {
      "epoch": 0.14225746268656717,
      "grad_norm": 0.44959729986609015,
      "learning_rate": 4.89684685097278e-05,
      "loss": 0.5648,
      "step": 305
    },
    {
      "epoch": 0.14458955223880596,
      "grad_norm": 0.4775655291361343,
      "learning_rate": 4.891587190489891e-05,
      "loss": 0.5876,
      "step": 310
    },
    {
      "epoch": 0.14692164179104478,
      "grad_norm": 0.44367799195931545,
      "learning_rate": 4.886200057122568e-05,
      "loss": 0.59,
      "step": 315
    },
    {
      "epoch": 0.14925373134328357,
      "grad_norm": 0.46351014448368566,
      "learning_rate": 4.880685771527114e-05,
      "loss": 0.578,
      "step": 320
    },
    {
      "epoch": 0.1515858208955224,
      "grad_norm": 0.4000400178625835,
      "learning_rate": 4.8750446619282646e-05,
      "loss": 0.5782,
      "step": 325
    },
    {
      "epoch": 0.15391791044776118,
      "grad_norm": 0.42361868610958603,
      "learning_rate": 4.869277064099654e-05,
      "loss": 0.5816,
      "step": 330
    },
    {
      "epoch": 0.15625,
      "grad_norm": 0.41929276911015007,
      "learning_rate": 4.8633833213438254e-05,
      "loss": 0.5787,
      "step": 335
    },
    {
      "epoch": 0.15858208955223882,
      "grad_norm": 0.4278081032230215,
      "learning_rate": 4.8573637844718e-05,
      "loss": 0.6021,
      "step": 340
    },
    {
      "epoch": 0.1609141791044776,
      "grad_norm": 0.4435981265670742,
      "learning_rate": 4.851218811782195e-05,
      "loss": 0.5851,
      "step": 345
    },
    {
      "epoch": 0.16324626865671643,
      "grad_norm": 0.4610464853647174,
      "learning_rate": 4.844948769039896e-05,
      "loss": 0.5843,
      "step": 350
    },
    {
      "epoch": 0.16557835820895522,
      "grad_norm": 0.4990418525776248,
      "learning_rate": 4.8385540294542855e-05,
      "loss": 0.5777,
      "step": 355
    },
    {
      "epoch": 0.16791044776119404,
      "grad_norm": 0.41390141315910106,
      "learning_rate": 4.83203497365703e-05,
      "loss": 0.5833,
      "step": 360
    },
    {
      "epoch": 0.17024253731343283,
      "grad_norm": 0.42444589701648866,
      "learning_rate": 4.825391989679422e-05,
      "loss": 0.5779,
      "step": 365
    },
    {
      "epoch": 0.17257462686567165,
      "grad_norm": 0.4048755132488697,
      "learning_rate": 4.818625472929286e-05,
      "loss": 0.5692,
      "step": 370
    },
    {
      "epoch": 0.17490671641791045,
      "grad_norm": 0.44573994111066584,
      "learning_rate": 4.811735826167436e-05,
      "loss": 0.5605,
      "step": 375
    },
    {
      "epoch": 0.17723880597014927,
      "grad_norm": 0.44870138659062336,
      "learning_rate": 4.8047234594837143e-05,
      "loss": 0.572,
      "step": 380
    },
    {
      "epoch": 0.17957089552238806,
      "grad_norm": 0.44214350416809645,
      "learning_rate": 4.7975887902725696e-05,
      "loss": 0.5602,
      "step": 385
    },
    {
      "epoch": 0.18190298507462688,
      "grad_norm": 0.4261029084271384,
      "learning_rate": 4.7903322432082185e-05,
      "loss": 0.5736,
      "step": 390
    },
    {
      "epoch": 0.18423507462686567,
      "grad_norm": 0.36636721217653195,
      "learning_rate": 4.78295425021937e-05,
      "loss": 0.5547,
      "step": 395
    },
    {
      "epoch": 0.1865671641791045,
      "grad_norm": 0.47458322491073895,
      "learning_rate": 4.775455250463507e-05,
      "loss": 0.5788,
      "step": 400
    },
    {
      "epoch": 0.18889925373134328,
      "grad_norm": 0.42726396052859394,
      "learning_rate": 4.767835690300759e-05,
      "loss": 0.577,
      "step": 405
    },
    {
      "epoch": 0.1912313432835821,
      "grad_norm": 0.4153229184095994,
      "learning_rate": 4.760096023267322e-05,
      "loss": 0.5827,
      "step": 410
    },
    {
      "epoch": 0.1935634328358209,
      "grad_norm": 0.39439465691893505,
      "learning_rate": 4.752236710048472e-05,
      "loss": 0.5647,
      "step": 415
    },
    {
      "epoch": 0.1958955223880597,
      "grad_norm": 0.47858737477245633,
      "learning_rate": 4.744258218451135e-05,
      "loss": 0.5787,
      "step": 420
    },
    {
      "epoch": 0.1982276119402985,
      "grad_norm": 0.42698704619750955,
      "learning_rate": 4.736161023376051e-05,
      "loss": 0.5513,
      "step": 425
    },
    {
      "epoch": 0.20055970149253732,
      "grad_norm": 0.43352489747133166,
      "learning_rate": 4.7279456067895e-05,
      "loss": 0.5879,
      "step": 430
    },
    {
      "epoch": 0.20289179104477612,
      "grad_norm": 0.4366534779661185,
      "learning_rate": 4.719612457694616e-05,
      "loss": 0.5726,
      "step": 435
    },
    {
      "epoch": 0.20522388059701493,
      "grad_norm": 0.41581245276436585,
      "learning_rate": 4.71116207210228e-05,
      "loss": 0.5577,
      "step": 440
    },
    {
      "epoch": 0.20755597014925373,
      "grad_norm": 0.39278696980143335,
      "learning_rate": 4.7025949530016e-05,
      "loss": 0.5491,
      "step": 445
    },
    {
      "epoch": 0.20988805970149255,
      "grad_norm": 0.4229385366750606,
      "learning_rate": 4.6939116103299655e-05,
      "loss": 0.5679,
      "step": 450
    },
    {
      "epoch": 0.21222014925373134,
      "grad_norm": 0.43710561997855185,
      "learning_rate": 4.685112560942699e-05,
      "loss": 0.5661,
      "step": 455
    },
    {
      "epoch": 0.21455223880597016,
      "grad_norm": 0.5250968554881381,
      "learning_rate": 4.676198328582288e-05,
      "loss": 0.5692,
      "step": 460
    },
    {
      "epoch": 0.21688432835820895,
      "grad_norm": 0.3953624442644497,
      "learning_rate": 4.6671694438472154e-05,
      "loss": 0.5807,
      "step": 465
    },
    {
      "epoch": 0.21921641791044777,
      "grad_norm": 0.40181633683171075,
      "learning_rate": 4.6580264441603724e-05,
      "loss": 0.5676,
      "step": 470
    },
    {
      "epoch": 0.22154850746268656,
      "grad_norm": 0.44021613584779223,
      "learning_rate": 4.648769873737071e-05,
      "loss": 0.5733,
      "step": 475
    },
    {
      "epoch": 0.22388059701492538,
      "grad_norm": 0.40102131427870424,
      "learning_rate": 4.6394002835526535e-05,
      "loss": 0.5731,
      "step": 480
    },
    {
      "epoch": 0.22621268656716417,
      "grad_norm": 0.4266724170549558,
      "learning_rate": 4.6299182313096916e-05,
      "loss": 0.5909,
      "step": 485
    },
    {
      "epoch": 0.228544776119403,
      "grad_norm": 0.4125686640546716,
      "learning_rate": 4.6203242814047946e-05,
      "loss": 0.5255,
      "step": 490
    },
    {
      "epoch": 0.23087686567164178,
      "grad_norm": 0.4641709249366716,
      "learning_rate": 4.610619004895017e-05,
      "loss": 0.5711,
      "step": 495
    },
    {
      "epoch": 0.2332089552238806,
      "grad_norm": 0.3968357488226857,
      "learning_rate": 4.6008029794638596e-05,
      "loss": 0.5452,
      "step": 500
    },
    {
      "epoch": 0.2355410447761194,
      "grad_norm": 0.403217768357308,
      "learning_rate": 4.590876789386893e-05,
      "loss": 0.551,
      "step": 505
    },
    {
      "epoch": 0.23787313432835822,
      "grad_norm": 0.4010193919066992,
      "learning_rate": 4.580841025496974e-05,
      "loss": 0.5595,
      "step": 510
    },
    {
      "epoch": 0.240205223880597,
      "grad_norm": 0.41662390187739595,
      "learning_rate": 4.570696285149084e-05,
      "loss": 0.563,
      "step": 515
    },
    {
      "epoch": 0.24253731343283583,
      "grad_norm": 0.36311719630385086,
      "learning_rate": 4.560443172184763e-05,
      "loss": 0.5652,
      "step": 520
    },
    {
      "epoch": 0.24486940298507462,
      "grad_norm": 0.41407920938729176,
      "learning_rate": 4.55008229689618e-05,
      "loss": 0.5503,
      "step": 525
    },
    {
      "epoch": 0.24720149253731344,
      "grad_norm": 0.4405772717510258,
      "learning_rate": 4.539614275989793e-05,
      "loss": 0.5569,
      "step": 530
    },
    {
      "epoch": 0.24953358208955223,
      "grad_norm": 0.4201353925492403,
      "learning_rate": 4.529039732549653e-05,
      "loss": 0.5584,
      "step": 535
    },
    {
      "epoch": 0.251865671641791,
      "grad_norm": 0.3945439648388656,
      "learning_rate": 4.5183592960003104e-05,
      "loss": 0.5454,
      "step": 540
    },
    {
      "epoch": 0.25419776119402987,
      "grad_norm": 0.3904200336576666,
      "learning_rate": 4.507573602069351e-05,
      "loss": 0.559,
      "step": 545
    },
    {
      "epoch": 0.25652985074626866,
      "grad_norm": 0.46072885500157207,
      "learning_rate": 4.496683292749555e-05,
      "loss": 0.5536,
      "step": 550
    },
    {
      "epoch": 0.25886194029850745,
      "grad_norm": 0.42578053791450554,
      "learning_rate": 4.485689016260686e-05,
      "loss": 0.5469,
      "step": 555
    },
    {
      "epoch": 0.26119402985074625,
      "grad_norm": 0.37965084505723456,
      "learning_rate": 4.4745914270109055e-05,
      "loss": 0.5719,
      "step": 560
    },
    {
      "epoch": 0.2635261194029851,
      "grad_norm": 0.3871506155938371,
      "learning_rate": 4.463391185557822e-05,
      "loss": 0.5621,
      "step": 565
    },
    {
      "epoch": 0.2658582089552239,
      "grad_norm": 0.34520167862600565,
      "learning_rate": 4.4520889585691705e-05,
      "loss": 0.5674,
      "step": 570
    },
    {
      "epoch": 0.2681902985074627,
      "grad_norm": 0.3710748999904097,
      "learning_rate": 4.440685418783135e-05,
      "loss": 0.546,
      "step": 575
    },
    {
      "epoch": 0.27052238805970147,
      "grad_norm": 0.3933601502845124,
      "learning_rate": 4.429181244968301e-05,
      "loss": 0.5513,
      "step": 580
    },
    {
      "epoch": 0.2728544776119403,
      "grad_norm": 0.4185563085937644,
      "learning_rate": 4.417577121883256e-05,
      "loss": 0.5516,
      "step": 585
    },
    {
      "epoch": 0.2751865671641791,
      "grad_norm": 0.4498108663647128,
      "learning_rate": 4.4058737402358295e-05,
      "loss": 0.5352,
      "step": 590
    },
    {
      "epoch": 0.2775186567164179,
      "grad_norm": 0.4227714398109878,
      "learning_rate": 4.394071796641983e-05,
      "loss": 0.555,
      "step": 595
    },
    {
      "epoch": 0.2798507462686567,
      "grad_norm": 0.5187752318511273,
      "learning_rate": 4.38217199358434e-05,
      "loss": 0.5638,
      "step": 600
    },
    {
      "epoch": 0.28218283582089554,
      "grad_norm": 0.4010448762418649,
      "learning_rate": 4.3701750393703786e-05,
      "loss": 0.5313,
      "step": 605
    },
    {
      "epoch": 0.28451492537313433,
      "grad_norm": 0.38989169644191773,
      "learning_rate": 4.3580816480902656e-05,
      "loss": 0.5557,
      "step": 610
    },
    {
      "epoch": 0.2868470149253731,
      "grad_norm": 0.3551587559272834,
      "learning_rate": 4.345892539574359e-05,
      "loss": 0.5629,
      "step": 615
    },
    {
      "epoch": 0.2891791044776119,
      "grad_norm": 0.3752271018747494,
      "learning_rate": 4.3336084393503545e-05,
      "loss": 0.5395,
      "step": 620
    },
    {
      "epoch": 0.29151119402985076,
      "grad_norm": 0.3805402923746079,
      "learning_rate": 4.3212300786001045e-05,
      "loss": 0.5424,
      "step": 625
    },
    {
      "epoch": 0.29384328358208955,
      "grad_norm": 0.3872910865097459,
      "learning_rate": 4.308758194116094e-05,
      "loss": 0.5652,
      "step": 630
    },
    {
      "epoch": 0.29617537313432835,
      "grad_norm": 0.3596225684825542,
      "learning_rate": 4.296193528257586e-05,
      "loss": 0.569,
      "step": 635
    },
    {
      "epoch": 0.29850746268656714,
      "grad_norm": 0.3602254845503326,
      "learning_rate": 4.283536828906436e-05,
      "loss": 0.5612,
      "step": 640
    },
    {
      "epoch": 0.300839552238806,
      "grad_norm": 0.36726034960137754,
      "learning_rate": 4.270788849422572e-05,
      "loss": 0.5456,
      "step": 645
    },
    {
      "epoch": 0.3031716417910448,
      "grad_norm": 0.42290471610215613,
      "learning_rate": 4.2579503485991567e-05,
      "loss": 0.554,
      "step": 650
    },
    {
      "epoch": 0.30550373134328357,
      "grad_norm": 0.3619033441543854,
      "learning_rate": 4.245022090617418e-05,
      "loss": 0.5575,
      "step": 655
    },
    {
      "epoch": 0.30783582089552236,
      "grad_norm": 0.37187893768522473,
      "learning_rate": 4.2320048450011684e-05,
      "loss": 0.5567,
      "step": 660
    },
    {
      "epoch": 0.3101679104477612,
      "grad_norm": 0.3536084758242916,
      "learning_rate": 4.218899386570994e-05,
      "loss": 0.5407,
      "step": 665
    },
    {
      "epoch": 0.3125,
      "grad_norm": 0.36137231978307244,
      "learning_rate": 4.205706495398143e-05,
      "loss": 0.5629,
      "step": 670
    },
    {
      "epoch": 0.3148320895522388,
      "grad_norm": 0.3929037572466559,
      "learning_rate": 4.192426956758085e-05,
      "loss": 0.533,
      "step": 675
    },
    {
      "epoch": 0.31716417910447764,
      "grad_norm": 0.3409013299580602,
      "learning_rate": 4.179061561083777e-05,
      "loss": 0.5423,
      "step": 680
    },
    {
      "epoch": 0.31949626865671643,
      "grad_norm": 0.3875302006330962,
      "learning_rate": 4.165611103918612e-05,
      "loss": 0.5624,
      "step": 685
    },
    {
      "epoch": 0.3218283582089552,
      "grad_norm": 0.4117426045707057,
      "learning_rate": 4.1520763858690644e-05,
      "loss": 0.5422,
      "step": 690
    },
    {
      "epoch": 0.324160447761194,
      "grad_norm": 0.3822435014923177,
      "learning_rate": 4.138458212557038e-05,
      "loss": 0.5597,
      "step": 695
    },
    {
      "epoch": 0.32649253731343286,
      "grad_norm": 0.40859751039743775,
      "learning_rate": 4.124757394571914e-05,
      "loss": 0.5375,
      "step": 700
    },
    {
      "epoch": 0.32882462686567165,
      "grad_norm": 0.35376270605093485,
      "learning_rate": 4.110974747422299e-05,
      "loss": 0.5441,
      "step": 705
    },
    {
      "epoch": 0.33115671641791045,
      "grad_norm": 0.40517007949766226,
      "learning_rate": 4.097111091487486e-05,
      "loss": 0.5404,
      "step": 710
    },
    {
      "epoch": 0.33348880597014924,
      "grad_norm": 0.40133799439007284,
      "learning_rate": 4.083167251968625e-05,
      "loss": 0.567,
      "step": 715
    },
    {
      "epoch": 0.3358208955223881,
      "grad_norm": 0.40796007686855523,
      "learning_rate": 4.069144058839605e-05,
      "loss": 0.5536,
      "step": 720
    },
    {
      "epoch": 0.3381529850746269,
      "grad_norm": 0.3586891053467589,
      "learning_rate": 4.055042346797643e-05,
      "loss": 0.5491,
      "step": 725
    },
    {
      "epoch": 0.34048507462686567,
      "grad_norm": 0.38971100356762045,
      "learning_rate": 4.040862955213615e-05,
      "loss": 0.5595,
      "step": 730
    },
    {
      "epoch": 0.34281716417910446,
      "grad_norm": 0.3663608610076594,
      "learning_rate": 4.026606728082082e-05,
      "loss": 0.5483,
      "step": 735
    },
    {
      "epoch": 0.3451492537313433,
      "grad_norm": 0.35456132002369006,
      "learning_rate": 4.012274513971061e-05,
      "loss": 0.5642,
      "step": 740
    },
    {
      "epoch": 0.3474813432835821,
      "grad_norm": 0.3685315896394436,
      "learning_rate": 3.997867165971512e-05,
      "loss": 0.5382,
      "step": 745
    },
    {
      "epoch": 0.3498134328358209,
      "grad_norm": 0.36917897331690136,
      "learning_rate": 3.9833855416465624e-05,
      "loss": 0.5493,
      "step": 750
    },
    {
      "epoch": 0.3521455223880597,
      "grad_norm": 0.40099943476775346,
      "learning_rate": 3.968830502980459e-05,
      "loss": 0.544,
      "step": 755
    },
    {
      "epoch": 0.35447761194029853,
      "grad_norm": 0.38203410425672707,
      "learning_rate": 3.954202916327264e-05,
      "loss": 0.5414,
      "step": 760
    },
    {
      "epoch": 0.3568097014925373,
      "grad_norm": 0.3649022039094651,
      "learning_rate": 3.939503652359287e-05,
      "loss": 0.5428,
      "step": 765
    },
    {
      "epoch": 0.3591417910447761,
      "grad_norm": 0.39137773057483016,
      "learning_rate": 3.924733586015257e-05,
      "loss": 0.5394,
      "step": 770
    },
    {
      "epoch": 0.3614738805970149,
      "grad_norm": 0.35114090355932165,
      "learning_rate": 3.9098935964482476e-05,
      "loss": 0.5411,
      "step": 775
    },
    {
      "epoch": 0.36380597014925375,
      "grad_norm": 0.3473596414558861,
      "learning_rate": 3.894984566973346e-05,
      "loss": 0.549,
      "step": 780
    },
    {
      "epoch": 0.36613805970149255,
      "grad_norm": 0.34411846026118503,
      "learning_rate": 3.880007385015075e-05,
      "loss": 0.5382,
      "step": 785
    },
    {
      "epoch": 0.36847014925373134,
      "grad_norm": 0.3389415203112235,
      "learning_rate": 3.864962942054572e-05,
      "loss": 0.5639,
      "step": 790
    },
    {
      "epoch": 0.37080223880597013,
      "grad_norm": 0.351582240594775,
      "learning_rate": 3.849852133576527e-05,
      "loss": 0.5304,
      "step": 795
    },
    {
      "epoch": 0.373134328358209,
      "grad_norm": 0.36495268124322017,
      "learning_rate": 3.834675859015876e-05,
      "loss": 0.5549,
      "step": 800
    },
    {
      "epoch": 0.37546641791044777,
      "grad_norm": 0.3896220021553757,
      "learning_rate": 3.819435021704274e-05,
      "loss": 0.5446,
      "step": 805
    },
    {
      "epoch": 0.37779850746268656,
      "grad_norm": 0.45143305097848657,
      "learning_rate": 3.804130528816312e-05,
      "loss": 0.5457,
      "step": 810
    },
    {
      "epoch": 0.38013059701492535,
      "grad_norm": 0.3513699985563104,
      "learning_rate": 3.7887632913155355e-05,
      "loss": 0.5344,
      "step": 815
    },
    {
      "epoch": 0.3824626865671642,
      "grad_norm": 0.35958918118902816,
      "learning_rate": 3.77333422390021e-05,
      "loss": 0.5267,
      "step": 820
    },
    {
      "epoch": 0.384794776119403,
      "grad_norm": 0.3871250664530478,
      "learning_rate": 3.75784424494888e-05,
      "loss": 0.5364,
      "step": 825
    },
    {
      "epoch": 0.3871268656716418,
      "grad_norm": 0.3627174602408319,
      "learning_rate": 3.7422942764657054e-05,
      "loss": 0.5472,
      "step": 830
    },
    {
      "epoch": 0.3894589552238806,
      "grad_norm": 0.37549355388736905,
      "learning_rate": 3.726685244025578e-05,
      "loss": 0.5503,
      "step": 835
    },
    {
      "epoch": 0.3917910447761194,
      "grad_norm": 0.4381992071243207,
      "learning_rate": 3.711018076719034e-05,
      "loss": 0.5438,
      "step": 840
    },
    {
      "epoch": 0.3941231343283582,
      "grad_norm": 0.36668027314474627,
      "learning_rate": 3.695293707096947e-05,
      "loss": 0.5537,
      "step": 845
    },
    {
      "epoch": 0.396455223880597,
      "grad_norm": 0.3806449160418856,
      "learning_rate": 3.679513071115025e-05,
      "loss": 0.5461,
      "step": 850
    },
    {
      "epoch": 0.3987873134328358,
      "grad_norm": 0.35429888945629073,
      "learning_rate": 3.663677108078094e-05,
      "loss": 0.551,
      "step": 855
    },
    {
      "epoch": 0.40111940298507465,
      "grad_norm": 0.3558912409357651,
      "learning_rate": 3.647786760584194e-05,
      "loss": 0.5434,
      "step": 860
    },
    {
      "epoch": 0.40345149253731344,
      "grad_norm": 0.3527211756527707,
      "learning_rate": 3.6318429744684676e-05,
      "loss": 0.5545,
      "step": 865
    },
    {
      "epoch": 0.40578358208955223,
      "grad_norm": 0.3656847568112223,
      "learning_rate": 3.615846698746869e-05,
      "loss": 0.5468,
      "step": 870
    },
    {
      "epoch": 0.408115671641791,
      "grad_norm": 0.383186699265511,
      "learning_rate": 3.599798885559667e-05,
      "loss": 0.5589,
      "step": 875
    },
    {
      "epoch": 0.41044776119402987,
      "grad_norm": 0.37392311284389557,
      "learning_rate": 3.583700490114776e-05,
      "loss": 0.5273,
      "step": 880
    },
    {
      "epoch": 0.41277985074626866,
      "grad_norm": 0.3635317613741542,
      "learning_rate": 3.5675524706309014e-05,
      "loss": 0.5273,
      "step": 885
    },
    {
      "epoch": 0.41511194029850745,
      "grad_norm": 0.38305068761455124,
      "learning_rate": 3.5513557882805e-05,
      "loss": 0.5559,
      "step": 890
    },
    {
      "epoch": 0.41744402985074625,
      "grad_norm": 0.3870612561938992,
      "learning_rate": 3.5351114071325696e-05,
      "loss": 0.5422,
      "step": 895
    },
    {
      "epoch": 0.4197761194029851,
      "grad_norm": 0.38347742537836593,
      "learning_rate": 3.518820294095267e-05,
      "loss": 0.5446,
      "step": 900
    },
    {
      "epoch": 0.4221082089552239,
      "grad_norm": 0.35092427031052703,
      "learning_rate": 3.50248341885835e-05,
      "loss": 0.5326,
      "step": 905
    },
    {
      "epoch": 0.4244402985074627,
      "grad_norm": 0.3561353515285881,
      "learning_rate": 3.486101753835468e-05,
      "loss": 0.5403,
      "step": 910
    },
    {
      "epoch": 0.42677238805970147,
      "grad_norm": 0.3085374698690792,
      "learning_rate": 3.469676274106271e-05,
      "loss": 0.5315,
      "step": 915
    },
    {
      "epoch": 0.4291044776119403,
      "grad_norm": 0.33075051926007265,
      "learning_rate": 3.453207957358377e-05,
      "loss": 0.5337,
      "step": 920
    },
    {
      "epoch": 0.4314365671641791,
      "grad_norm": 0.3515295495723763,
      "learning_rate": 3.436697783829178e-05,
      "loss": 0.5304,
      "step": 925
    },
    {
      "epoch": 0.4337686567164179,
      "grad_norm": 0.3132788673571813,
      "learning_rate": 3.420146736247487e-05,
      "loss": 0.5161,
      "step": 930
    },
    {
      "epoch": 0.4361007462686567,
      "grad_norm": 0.3739603265493346,
      "learning_rate": 3.4035557997750506e-05,
      "loss": 0.5497,
      "step": 935
    },
    {
      "epoch": 0.43843283582089554,
      "grad_norm": 0.34545120023484394,
      "learning_rate": 3.386925961947906e-05,
      "loss": 0.5439,
      "step": 940
    },
    {
      "epoch": 0.44076492537313433,
      "grad_norm": 0.3757040378506225,
      "learning_rate": 3.370258212617602e-05,
      "loss": 0.5588,
      "step": 945
    },
    {
      "epoch": 0.4430970149253731,
      "grad_norm": 0.3402839804931963,
      "learning_rate": 3.353553543892277e-05,
      "loss": 0.5335,
      "step": 950
    },
    {
      "epoch": 0.4454291044776119,
      "grad_norm": 0.33230952626406335,
      "learning_rate": 3.336812950077611e-05,
      "loss": 0.5388,
      "step": 955
    },
    {
      "epoch": 0.44776119402985076,
      "grad_norm": 0.35261211656761277,
      "learning_rate": 3.320037427617639e-05,
      "loss": 0.5375,
      "step": 960
    },
    {
      "epoch": 0.45009328358208955,
      "grad_norm": 0.3599414963399435,
      "learning_rate": 3.30322797503544e-05,
      "loss": 0.547,
      "step": 965
    },
    {
      "epoch": 0.45242537313432835,
      "grad_norm": 0.35470640352572197,
      "learning_rate": 3.2863855928737026e-05,
      "loss": 0.5199,
      "step": 970
    },
    {
      "epoch": 0.45475746268656714,
      "grad_norm": 0.3452568269058605,
      "learning_rate": 3.2695112836351703e-05,
      "loss": 0.5382,
      "step": 975
    },
    {
      "epoch": 0.457089552238806,
      "grad_norm": 0.3362066514765104,
      "learning_rate": 3.252606051722972e-05,
      "loss": 0.5563,
      "step": 980
    },
    {
      "epoch": 0.4594216417910448,
      "grad_norm": 0.34701565335688916,
      "learning_rate": 3.235670903380832e-05,
      "loss": 0.532,
      "step": 985
    },
    {
      "epoch": 0.46175373134328357,
      "grad_norm": 0.3518883926304785,
      "learning_rate": 3.218706846633183e-05,
      "loss": 0.5305,
      "step": 990
    },
    {
      "epoch": 0.46408582089552236,
      "grad_norm": 0.33370892763193366,
      "learning_rate": 3.201714891225156e-05,
      "loss": 0.5288,
      "step": 995
    },
    {
      "epoch": 0.4664179104477612,
      "grad_norm": 0.3265309643482067,
      "learning_rate": 3.1846960485624886e-05,
      "loss": 0.5261,
      "step": 1000
    },
    {
      "epoch": 0.46875,
      "grad_norm": 0.3428336788771024,
      "learning_rate": 3.1676513316513156e-05,
      "loss": 0.546,
      "step": 1005
    },
    {
      "epoch": 0.4710820895522388,
      "grad_norm": 0.32497034232497185,
      "learning_rate": 3.150581755037877e-05,
      "loss": 0.5327,
      "step": 1010
    },
    {
      "epoch": 0.47341417910447764,
      "grad_norm": 0.32932560895962015,
      "learning_rate": 3.133488334748125e-05,
      "loss": 0.542,
      "step": 1015
    },
    {
      "epoch": 0.47574626865671643,
      "grad_norm": 0.34317416669371287,
      "learning_rate": 3.1163720882272516e-05,
      "loss": 0.5257,
      "step": 1020
    },
    {
      "epoch": 0.4780783582089552,
      "grad_norm": 0.33195906720998813,
      "learning_rate": 3.0992340342791246e-05,
      "loss": 0.5362,
      "step": 1025
    },
    {
      "epoch": 0.480410447761194,
      "grad_norm": 0.3234621303310154,
      "learning_rate": 3.08207519300565e-05,
      "loss": 0.5298,
      "step": 1030
    },
    {
      "epoch": 0.48274253731343286,
      "grad_norm": 0.3439747189931589,
      "learning_rate": 3.064896585746045e-05,
      "loss": 0.5216,
      "step": 1035
    },
    {
      "epoch": 0.48507462686567165,
      "grad_norm": 0.3225618210162447,
      "learning_rate": 3.047699235016056e-05,
      "loss": 0.5271,
      "step": 1040
    },
    {
      "epoch": 0.48740671641791045,
      "grad_norm": 0.4481757291622055,
      "learning_rate": 3.030484164447085e-05,
      "loss": 0.5185,
      "step": 1045
    },
    {
      "epoch": 0.48973880597014924,
      "grad_norm": 0.34946082293471153,
      "learning_rate": 3.0132523987252658e-05,
      "loss": 0.5363,
      "step": 1050
    },
    {
      "epoch": 0.4920708955223881,
      "grad_norm": 0.3337244432348151,
      "learning_rate": 2.9960049635304755e-05,
      "loss": 0.5128,
      "step": 1055
    },
    {
      "epoch": 0.4944029850746269,
      "grad_norm": 0.33744807849843744,
      "learning_rate": 2.9787428854752736e-05,
      "loss": 0.5225,
      "step": 1060
    },
    {
      "epoch": 0.49673507462686567,
      "grad_norm": 0.3589752854354772,
      "learning_rate": 2.961467192043807e-05,
      "loss": 0.5323,
      "step": 1065
    },
    {
      "epoch": 0.49906716417910446,
      "grad_norm": 0.3680177579706306,
      "learning_rate": 2.9441789115306402e-05,
      "loss": 0.5453,
      "step": 1070
    },
    {
      "epoch": 0.5013992537313433,
      "grad_norm": 0.39008642388483844,
      "learning_rate": 2.926879072979558e-05,
      "loss": 0.5363,
      "step": 1075
    },
    {
      "epoch": 0.503731343283582,
      "grad_norm": 0.35182913635444735,
      "learning_rate": 2.9095687061223058e-05,
      "loss": 0.5345,
      "step": 1080
    },
    {
      "epoch": 0.5060634328358209,
      "grad_norm": 0.3687130437242019,
      "learning_rate": 2.8922488413173053e-05,
      "loss": 0.5326,
      "step": 1085
    },
    {
      "epoch": 0.5083955223880597,
      "grad_norm": 0.3896087742442836,
      "learning_rate": 2.874920509488319e-05,
      "loss": 0.5439,
      "step": 1090
    },
    {
      "epoch": 0.5107276119402985,
      "grad_norm": 0.35004354808221283,
      "learning_rate": 2.8575847420630887e-05,
      "loss": 0.5215,
      "step": 1095
    },
    {
      "epoch": 0.5130597014925373,
      "grad_norm": 0.34400548959145894,
      "learning_rate": 2.8402425709119435e-05,
      "loss": 0.5449,
      "step": 1100
    },
    {
      "epoch": 0.5153917910447762,
      "grad_norm": 0.34279814513241086,
      "learning_rate": 2.8228950282863776e-05,
      "loss": 0.5352,
      "step": 1105
    },
    {
      "epoch": 0.5177238805970149,
      "grad_norm": 0.33066562290386337,
      "learning_rate": 2.8055431467576106e-05,
      "loss": 0.5277,
      "step": 1110
    },
    {
      "epoch": 0.5200559701492538,
      "grad_norm": 0.36541005581441127,
      "learning_rate": 2.788187959155124e-05,
      "loss": 0.5208,
      "step": 1115
    },
    {
      "epoch": 0.5223880597014925,
      "grad_norm": 0.31819346746550703,
      "learning_rate": 2.7708304985051868e-05,
      "loss": 0.5477,
      "step": 1120
    },
    {
      "epoch": 0.5247201492537313,
      "grad_norm": 0.3463253117162502,
      "learning_rate": 2.7534717979693647e-05,
      "loss": 0.5243,
      "step": 1125
    },
    {
      "epoch": 0.5270522388059702,
      "grad_norm": 0.32466560863468996,
      "learning_rate": 2.7361128907830253e-05,
      "loss": 0.5266,
      "step": 1130
    },
    {
      "epoch": 0.5293843283582089,
      "grad_norm": 0.32661440076143755,
      "learning_rate": 2.7187548101938353e-05,
      "loss": 0.5166,
      "step": 1135
    },
    {
      "epoch": 0.5317164179104478,
      "grad_norm": 0.35797548941766444,
      "learning_rate": 2.7013985894002623e-05,
      "loss": 0.5153,
      "step": 1140
    },
    {
      "epoch": 0.5340485074626866,
      "grad_norm": 0.3747723691950687,
      "learning_rate": 2.6840452614900726e-05,
      "loss": 0.54,
      "step": 1145
    },
    {
      "epoch": 0.5363805970149254,
      "grad_norm": 0.31227276183327857,
      "learning_rate": 2.6666958593788405e-05,
      "loss": 0.5214,
      "step": 1150
    },
    {
      "epoch": 0.5387126865671642,
      "grad_norm": 0.32028332057260017,
      "learning_rate": 2.649351415748466e-05,
      "loss": 0.5385,
      "step": 1155
    },
    {
      "epoch": 0.5410447761194029,
      "grad_norm": 0.31721311054218543,
      "learning_rate": 2.6320129629857093e-05,
      "loss": 0.5316,
      "step": 1160
    },
    {
      "epoch": 0.5433768656716418,
      "grad_norm": 0.32367205608069305,
      "learning_rate": 2.6146815331207358e-05,
      "loss": 0.5105,
      "step": 1165
    },
    {
      "epoch": 0.5457089552238806,
      "grad_norm": 0.32785600244724733,
      "learning_rate": 2.597358157765692e-05,
      "loss": 0.5349,
      "step": 1170
    },
    {
      "epoch": 0.5480410447761194,
      "grad_norm": 0.3418244281371759,
      "learning_rate": 2.5800438680532974e-05,
      "loss": 0.5193,
      "step": 1175
    },
    {
      "epoch": 0.5503731343283582,
      "grad_norm": 0.31552567017064354,
      "learning_rate": 2.56273969457547e-05,
      "loss": 0.5206,
      "step": 1180
    },
    {
      "epoch": 0.5527052238805971,
      "grad_norm": 0.3397428895495846,
      "learning_rate": 2.545446667321984e-05,
      "loss": 0.5379,
      "step": 1185
    },
    {
      "epoch": 0.5550373134328358,
      "grad_norm": 0.3286049050973758,
      "learning_rate": 2.528165815619162e-05,
      "loss": 0.5222,
      "step": 1190
    },
    {
      "epoch": 0.5573694029850746,
      "grad_norm": 0.3331787192546376,
      "learning_rate": 2.5108981680686035e-05,
      "loss": 0.5255,
      "step": 1195
    },
    {
      "epoch": 0.5597014925373134,
      "grad_norm": 0.3311603874036119,
      "learning_rate": 2.4936447524859625e-05,
      "loss": 0.5168,
      "step": 1200
    },
    {
      "epoch": 0.5620335820895522,
      "grad_norm": 0.3303918399172089,
      "learning_rate": 2.4764065958397715e-05,
      "loss": 0.5174,
      "step": 1205
    },
    {
      "epoch": 0.5643656716417911,
      "grad_norm": 0.33363993303401823,
      "learning_rate": 2.459184724190308e-05,
      "loss": 0.5097,
      "step": 1210
    },
    {
      "epoch": 0.5666977611940298,
      "grad_norm": 0.3338211217947914,
      "learning_rate": 2.441980162628527e-05,
      "loss": 0.5226,
      "step": 1215
    },
    {
      "epoch": 0.5690298507462687,
      "grad_norm": 0.3505424652287924,
      "learning_rate": 2.4247939352150386e-05,
      "loss": 0.5325,
      "step": 1220
    },
    {
      "epoch": 0.5713619402985075,
      "grad_norm": 0.34944994698597065,
      "learning_rate": 2.4076270649191573e-05,
      "loss": 0.5409,
      "step": 1225
    },
    {
      "epoch": 0.5736940298507462,
      "grad_norm": 0.32340035702891917,
      "learning_rate": 2.390480573558012e-05,
      "loss": 0.5273,
      "step": 1230
    },
    {
      "epoch": 0.5760261194029851,
      "grad_norm": 0.31938992678180467,
      "learning_rate": 2.3733554817357246e-05,
      "loss": 0.5244,
      "step": 1235
    },
    {
      "epoch": 0.5783582089552238,
      "grad_norm": 0.3166139465080639,
      "learning_rate": 2.3562528087826573e-05,
      "loss": 0.5131,
      "step": 1240
    },
    {
      "epoch": 0.5806902985074627,
      "grad_norm": 0.3487335569590694,
      "learning_rate": 2.339173572694746e-05,
      "loss": 0.5192,
      "step": 1245
    },
    {
      "epoch": 0.5830223880597015,
      "grad_norm": 0.30683594258607133,
      "learning_rate": 2.3221187900729003e-05,
      "loss": 0.5262,
      "step": 1250
    },
    {
      "epoch": 0.5853544776119403,
      "grad_norm": 0.3575092291210576,
      "learning_rate": 2.3050894760624982e-05,
      "loss": 0.5305,
      "step": 1255
    },
    {
      "epoch": 0.5876865671641791,
      "grad_norm": 0.34137275146006896,
      "learning_rate": 2.2880866442929544e-05,
      "loss": 0.5171,
      "step": 1260
    },
    {
      "epoch": 0.590018656716418,
      "grad_norm": 0.3314783873014681,
      "learning_rate": 2.271111306817396e-05,
      "loss": 0.5362,
      "step": 1265
    },
    {
      "epoch": 0.5923507462686567,
      "grad_norm": 0.3160634415556086,
      "learning_rate": 2.254164474052416e-05,
      "loss": 0.5177,
      "step": 1270
    },
    {
      "epoch": 0.5946828358208955,
      "grad_norm": 0.3014157107337593,
      "learning_rate": 2.237247154717932e-05,
      "loss": 0.5112,
      "step": 1275
    },
    {
      "epoch": 0.5970149253731343,
      "grad_norm": 0.3291310606063492,
      "learning_rate": 2.2203603557771447e-05,
      "loss": 0.5116,
      "step": 1280
    },
    {
      "epoch": 0.5993470149253731,
      "grad_norm": 0.32560419347412506,
      "learning_rate": 2.2035050823766008e-05,
      "loss": 0.5212,
      "step": 1285
    },
    {
      "epoch": 0.601679104477612,
      "grad_norm": 0.3472024703269534,
      "learning_rate": 2.186682337786365e-05,
      "loss": 0.526,
      "step": 1290
    },
    {
      "epoch": 0.6040111940298507,
      "grad_norm": 0.30442706575237044,
      "learning_rate": 2.1698931233403013e-05,
      "loss": 0.5089,
      "step": 1295
    },
    {
      "epoch": 0.6063432835820896,
      "grad_norm": 0.33335137810186144,
      "learning_rate": 2.153138438376473e-05,
      "loss": 0.5131,
      "step": 1300
    },
    {
      "epoch": 0.6086753731343284,
      "grad_norm": 0.3210060843108937,
      "learning_rate": 2.136419280177655e-05,
      "loss": 0.5089,
      "step": 1305
    },
    {
      "epoch": 0.6110074626865671,
      "grad_norm": 0.3177567286650887,
      "learning_rate": 2.119736643911979e-05,
      "loss": 0.5301,
      "step": 1310
    },
    {
      "epoch": 0.613339552238806,
      "grad_norm": 0.2990196515604672,
      "learning_rate": 2.1030915225736947e-05,
      "loss": 0.52,
      "step": 1315
    },
    {
      "epoch": 0.6156716417910447,
      "grad_norm": 0.3276526936140588,
      "learning_rate": 2.0864849069240645e-05,
      "loss": 0.5268,
      "step": 1320
    },
    {
      "epoch": 0.6180037313432836,
      "grad_norm": 0.2971098930268874,
      "learning_rate": 2.0699177854323902e-05,
      "loss": 0.5203,
      "step": 1325
    },
    {
      "epoch": 0.6203358208955224,
      "grad_norm": 0.30828873420690545,
      "learning_rate": 2.0533911442171805e-05,
      "loss": 0.5181,
      "step": 1330
    },
    {
      "epoch": 0.6226679104477612,
      "grad_norm": 0.32910014680345584,
      "learning_rate": 2.036905966987449e-05,
      "loss": 0.5239,
      "step": 1335
    },
    {
      "epoch": 0.625,
      "grad_norm": 0.3169543755211694,
      "learning_rate": 2.0204632349841667e-05,
      "loss": 0.509,
      "step": 1340
    },
    {
      "epoch": 0.6273320895522388,
      "grad_norm": 0.2992926255801442,
      "learning_rate": 2.0040639269218532e-05,
      "loss": 0.5133,
      "step": 1345
    },
    {
      "epoch": 0.6296641791044776,
      "grad_norm": 0.3356585986496801,
      "learning_rate": 1.9877090189303182e-05,
      "loss": 0.534,
      "step": 1350
    },
    {
      "epoch": 0.6319962686567164,
      "grad_norm": 0.3119369573784347,
      "learning_rate": 1.9713994844965657e-05,
      "loss": 0.512,
      "step": 1355
    },
    {
      "epoch": 0.6343283582089553,
      "grad_norm": 0.3135149938329128,
      "learning_rate": 1.9551362944068462e-05,
      "loss": 0.5212,
      "step": 1360
    },
    {
      "epoch": 0.636660447761194,
      "grad_norm": 0.3294648855188193,
      "learning_rate": 1.938920416688874e-05,
      "loss": 0.5193,
      "step": 1365
    },
    {
      "epoch": 0.6389925373134329,
      "grad_norm": 0.32539933928427905,
      "learning_rate": 1.922752816554204e-05,
      "loss": 0.5125,
      "step": 1370
    },
    {
      "epoch": 0.6413246268656716,
      "grad_norm": 0.31904607968778304,
      "learning_rate": 1.9066344563407856e-05,
      "loss": 0.5208,
      "step": 1375
    },
    {
      "epoch": 0.6436567164179104,
      "grad_norm": 0.3111959523904429,
      "learning_rate": 1.890566295455678e-05,
      "loss": 0.4994,
      "step": 1380
    },
    {
      "epoch": 0.6459888059701493,
      "grad_norm": 0.3275382050930323,
      "learning_rate": 1.874549290317946e-05,
      "loss": 0.5024,
      "step": 1385
    },
    {
      "epoch": 0.648320895522388,
      "grad_norm": 0.31189886724164534,
      "learning_rate": 1.858584394301728e-05,
      "loss": 0.5262,
      "step": 1390
    },
    {
      "epoch": 0.6506529850746269,
      "grad_norm": 0.342658228039591,
      "learning_rate": 1.8426725576794918e-05,
      "loss": 0.5429,
      "step": 1395
    },
    {
      "epoch": 0.6529850746268657,
      "grad_norm": 0.32330617361285674,
      "learning_rate": 1.8268147275654707e-05,
      "loss": 0.5189,
      "step": 1400
    },
    {
      "epoch": 0.6553171641791045,
      "grad_norm": 0.29687021132266317,
      "learning_rate": 1.8110118478592915e-05,
      "loss": 0.5065,
      "step": 1405
    },
    {
      "epoch": 0.6576492537313433,
      "grad_norm": 0.30079266830219414,
      "learning_rate": 1.7952648591897858e-05,
      "loss": 0.5206,
      "step": 1410
    },
    {
      "epoch": 0.659981343283582,
      "grad_norm": 0.3096917346984741,
      "learning_rate": 1.7795746988590027e-05,
      "loss": 0.5098,
      "step": 1415
    },
    {
      "epoch": 0.6623134328358209,
      "grad_norm": 0.3241152516622816,
      "learning_rate": 1.7639423007864252e-05,
      "loss": 0.5017,
      "step": 1420
    },
    {
      "epoch": 0.6646455223880597,
      "grad_norm": 0.30514910474437695,
      "learning_rate": 1.7483685954533692e-05,
      "loss": 0.5108,
      "step": 1425
    },
    {
      "epoch": 0.6669776119402985,
      "grad_norm": 0.32357387534514104,
      "learning_rate": 1.7328545098476106e-05,
      "loss": 0.5075,
      "step": 1430
    },
    {
      "epoch": 0.6693097014925373,
      "grad_norm": 0.3125919347018114,
      "learning_rate": 1.717400967408196e-05,
      "loss": 0.5129,
      "step": 1435
    },
    {
      "epoch": 0.6716417910447762,
      "grad_norm": 0.3017747173268348,
      "learning_rate": 1.702008887970491e-05,
      "loss": 0.51,
      "step": 1440
    },
    {
      "epoch": 0.6739738805970149,
      "grad_norm": 0.3038952541237383,
      "learning_rate": 1.6866791877114165e-05,
      "loss": 0.5165,
      "step": 1445
    },
    {
      "epoch": 0.6763059701492538,
      "grad_norm": 0.301385951224946,
      "learning_rate": 1.671412779094926e-05,
      "loss": 0.5137,
      "step": 1450
    },
    {
      "epoch": 0.6786380597014925,
      "grad_norm": 0.29572393831965715,
      "learning_rate": 1.656210570817685e-05,
      "loss": 0.5145,
      "step": 1455
    },
    {
      "epoch": 0.6809701492537313,
      "grad_norm": 0.3342752067990135,
      "learning_rate": 1.6410734677549872e-05,
      "loss": 0.508,
      "step": 1460
    },
    {
      "epoch": 0.6833022388059702,
      "grad_norm": 0.3453403341497158,
      "learning_rate": 1.6260023709068932e-05,
      "loss": 0.5146,
      "step": 1465
    },
    {
      "epoch": 0.6856343283582089,
      "grad_norm": 0.3171036782377084,
      "learning_rate": 1.6109981773446036e-05,
      "loss": 0.5102,
      "step": 1470
    },
    {
      "epoch": 0.6879664179104478,
      "grad_norm": 0.29917424397521875,
      "learning_rate": 1.5960617801570555e-05,
      "loss": 0.5086,
      "step": 1475
    },
    {
      "epoch": 0.6902985074626866,
      "grad_norm": 0.30931917309200585,
      "learning_rate": 1.58119406839777e-05,
      "loss": 0.5131,
      "step": 1480
    },
    {
      "epoch": 0.6926305970149254,
      "grad_norm": 0.29550887254843045,
      "learning_rate": 1.566395927031932e-05,
      "loss": 0.5029,
      "step": 1485
    },
    {
      "epoch": 0.6949626865671642,
      "grad_norm": 0.30550523923571943,
      "learning_rate": 1.5516682368837133e-05,
      "loss": 0.5096,
      "step": 1490
    },
    {
      "epoch": 0.6972947761194029,
      "grad_norm": 0.31911694408903835,
      "learning_rate": 1.5370118745838453e-05,
      "loss": 0.5287,
      "step": 1495
    },
    {
      "epoch": 0.6996268656716418,
      "grad_norm": 0.3050301639561124,
      "learning_rate": 1.5224277125174388e-05,
      "loss": 0.5162,
      "step": 1500
    },
    {
      "epoch": 0.7019589552238806,
      "grad_norm": 0.3041739601013223,
      "learning_rate": 1.5079166187720561e-05,
      "loss": 0.5148,
      "step": 1505
    },
    {
      "epoch": 0.7042910447761194,
      "grad_norm": 0.32709042336575056,
      "learning_rate": 1.4934794570860416e-05,
      "loss": 0.5314,
      "step": 1510
    },
    {
      "epoch": 0.7066231343283582,
      "grad_norm": 0.3271425613420151,
      "learning_rate": 1.4791170867971132e-05,
      "loss": 0.5096,
      "step": 1515
    },
    {
      "epoch": 0.7089552238805971,
      "grad_norm": 0.31173665895625186,
      "learning_rate": 1.464830362791204e-05,
      "loss": 0.5089,
      "step": 1520
    },
    {
      "epoch": 0.7112873134328358,
      "grad_norm": 0.31280945225700124,
      "learning_rate": 1.450620135451585e-05,
      "loss": 0.5141,
      "step": 1525
    },
    {
      "epoch": 0.7136194029850746,
      "grad_norm": 0.3029633492207749,
      "learning_rate": 1.4364872506082425e-05,
      "loss": 0.5206,
      "step": 1530
    },
    {
      "epoch": 0.7159514925373134,
      "grad_norm": 0.3120512711303495,
      "learning_rate": 1.4224325494875385e-05,
      "loss": 0.5288,
      "step": 1535
    },
    {
      "epoch": 0.7182835820895522,
      "grad_norm": 0.31204261621369056,
      "learning_rate": 1.4084568686621314e-05,
      "loss": 0.5117,
      "step": 1540
    },
    {
      "epoch": 0.7206156716417911,
      "grad_norm": 0.30972458717271706,
      "learning_rate": 1.3945610400011851e-05,
      "loss": 0.5234,
      "step": 1545
    },
    {
      "epoch": 0.7229477611940298,
      "grad_norm": 0.3038198863306939,
      "learning_rate": 1.3807458906208546e-05,
      "loss": 0.5309,
      "step": 1550
    },
    {
      "epoch": 0.7252798507462687,
      "grad_norm": 0.30351276635634267,
      "learning_rate": 1.3670122428350521e-05,
      "loss": 0.5096,
      "step": 1555
    },
    {
      "epoch": 0.7276119402985075,
      "grad_norm": 0.3009416107835106,
      "learning_rate": 1.3533609141065008e-05,
      "loss": 0.5037,
      "step": 1560
    },
    {
      "epoch": 0.7299440298507462,
      "grad_norm": 0.3115101175293925,
      "learning_rate": 1.3397927169980773e-05,
      "loss": 0.5092,
      "step": 1565
    },
    {
      "epoch": 0.7322761194029851,
      "grad_norm": 0.3050023046127322,
      "learning_rate": 1.326308459124447e-05,
      "loss": 0.5093,
      "step": 1570
    },
    {
      "epoch": 0.7346082089552238,
      "grad_norm": 0.31586860196893074,
      "learning_rate": 1.3129089431039931e-05,
      "loss": 0.5122,
      "step": 1575
    },
    {
      "epoch": 0.7369402985074627,
      "grad_norm": 0.31512946760011284,
      "learning_rate": 1.299594966511038e-05,
      "loss": 0.5238,
      "step": 1580
    },
    {
      "epoch": 0.7392723880597015,
      "grad_norm": 0.31489040782644456,
      "learning_rate": 1.2863673218283783e-05,
      "loss": 0.5091,
      "step": 1585
    },
    {
      "epoch": 0.7416044776119403,
      "grad_norm": 0.3190396380815969,
      "learning_rate": 1.2732267964001033e-05,
      "loss": 0.503,
      "step": 1590
    },
    {
      "epoch": 0.7439365671641791,
      "grad_norm": 0.31926216711658545,
      "learning_rate": 1.26017417238474e-05,
      "loss": 0.5138,
      "step": 1595
    },
    {
      "epoch": 0.746268656716418,
      "grad_norm": 0.30703349595902396,
      "learning_rate": 1.2472102267086904e-05,
      "loss": 0.5111,
      "step": 1600
    },
    {
      "epoch": 0.7486007462686567,
      "grad_norm": 0.3303603139088595,
      "learning_rate": 1.2343357310199925e-05,
      "loss": 0.5283,
      "step": 1605
    },
    {
      "epoch": 0.7509328358208955,
      "grad_norm": 0.32035527267812985,
      "learning_rate": 1.2215514516423813e-05,
      "loss": 0.5248,
      "step": 1610
    },
    {
      "epoch": 0.7532649253731343,
      "grad_norm": 0.2960384532117532,
      "learning_rate": 1.2088581495296852e-05,
      "loss": 0.5056,
      "step": 1615
    },
    {
      "epoch": 0.7555970149253731,
      "grad_norm": 0.2951510172831845,
      "learning_rate": 1.1962565802205255e-05,
      "loss": 0.505,
      "step": 1620
    },
    {
      "epoch": 0.757929104477612,
      "grad_norm": 0.29152330525669484,
      "learning_rate": 1.1837474937933464e-05,
      "loss": 0.5124,
      "step": 1625
    },
    {
      "epoch": 0.7602611940298507,
      "grad_norm": 0.29251949876960087,
      "learning_rate": 1.1713316348217673e-05,
      "loss": 0.5077,
      "step": 1630
    },
    {
      "epoch": 0.7625932835820896,
      "grad_norm": 0.29528274341042277,
      "learning_rate": 1.1590097423302684e-05,
      "loss": 0.4907,
      "step": 1635
    },
    {
      "epoch": 0.7649253731343284,
      "grad_norm": 0.30185100341646487,
      "learning_rate": 1.1467825497501954e-05,
      "loss": 0.4986,
      "step": 1640
    },
    {
      "epoch": 0.7672574626865671,
      "grad_norm": 0.32878665265624396,
      "learning_rate": 1.1346507848761077e-05,
      "loss": 0.5096,
      "step": 1645
    },
    {
      "epoch": 0.769589552238806,
      "grad_norm": 0.33863639658890127,
      "learning_rate": 1.1226151698224597e-05,
      "loss": 0.5211,
      "step": 1650
    },
    {
      "epoch": 0.7719216417910447,
      "grad_norm": 0.2962877335836816,
      "learning_rate": 1.1106764209806127e-05,
      "loss": 0.5125,
      "step": 1655
    },
    {
      "epoch": 0.7742537313432836,
      "grad_norm": 0.3158878505358477,
      "learning_rate": 1.0988352489762006e-05,
      "loss": 0.5259,
      "step": 1660
    },
    {
      "epoch": 0.7765858208955224,
      "grad_norm": 0.2881382914118729,
      "learning_rate": 1.0870923586268245e-05,
      "loss": 0.5075,
      "step": 1665
    },
    {
      "epoch": 0.7789179104477612,
      "grad_norm": 0.31719158675261994,
      "learning_rate": 1.0754484489001085e-05,
      "loss": 0.5112,
      "step": 1670
    },
    {
      "epoch": 0.78125,
      "grad_norm": 0.30964447847533355,
      "learning_rate": 1.0639042128720847e-05,
      "loss": 0.517,
      "step": 1675
    },
    {
      "epoch": 0.7835820895522388,
      "grad_norm": 0.3001723609524999,
      "learning_rate": 1.052460337685951e-05,
      "loss": 0.4999,
      "step": 1680
    },
    {
      "epoch": 0.7859141791044776,
      "grad_norm": 0.29546723704143385,
      "learning_rate": 1.0411175045111602e-05,
      "loss": 0.4981,
      "step": 1685
    },
    {
      "epoch": 0.7882462686567164,
      "grad_norm": 0.3075763476131655,
      "learning_rate": 1.0298763885028839e-05,
      "loss": 0.5112,
      "step": 1690
    },
    {
      "epoch": 0.7905783582089553,
      "grad_norm": 0.29591361609011235,
      "learning_rate": 1.018737658761817e-05,
      "loss": 0.5044,
      "step": 1695
    },
    {
      "epoch": 0.792910447761194,
      "grad_norm": 0.2933462196983581,
      "learning_rate": 1.0077019782943584e-05,
      "loss": 0.5151,
      "step": 1700
    },
    {
      "epoch": 0.7952425373134329,
      "grad_norm": 0.3018435924556085,
      "learning_rate": 9.967700039731427e-06,
      "loss": 0.5062,
      "step": 1705
    },
    {
      "epoch": 0.7975746268656716,
      "grad_norm": 0.28856623086950894,
      "learning_rate": 9.859423864979441e-06,
      "loss": 0.491,
      "step": 1710
    },
    {
      "epoch": 0.7999067164179104,
      "grad_norm": 0.29990971775965447,
      "learning_rate": 9.752197703569422e-06,
      "loss": 0.5248,
      "step": 1715
    },
    {
      "epoch": 0.8022388059701493,
      "grad_norm": 0.3031154798133516,
      "learning_rate": 9.646027937883622e-06,
      "loss": 0.5099,
      "step": 1720
    },
    {
      "epoch": 0.804570895522388,
      "grad_norm": 0.29656753451458473,
      "learning_rate": 9.54092088742485e-06,
      "loss": 0.5278,
      "step": 1725
    },
    {
      "epoch": 0.8069029850746269,
      "grad_norm": 0.3087030815128696,
      "learning_rate": 9.436882808440334e-06,
      "loss": 0.5091,
      "step": 1730
    },
    {
      "epoch": 0.8092350746268657,
      "grad_norm": 0.30044194171207367,
      "learning_rate": 9.333919893549294e-06,
      "loss": 0.5049,
      "step": 1735
    },
    {
      "epoch": 0.8115671641791045,
      "grad_norm": 0.30610765770429715,
      "learning_rate": 9.232038271374377e-06,
      "loss": 0.4965,
      "step": 1740
    },
    {
      "epoch": 0.8138992537313433,
      "grad_norm": 0.3174108731530412,
      "learning_rate": 9.131244006176846e-06,
      "loss": 0.5118,
      "step": 1745
    },
    {
      "epoch": 0.816231343283582,
      "grad_norm": 0.30939560544024536,
      "learning_rate": 9.031543097495638e-06,
      "loss": 0.4984,
      "step": 1750
    },
    {
      "epoch": 0.8185634328358209,
      "grad_norm": 0.29094442523658076,
      "learning_rate": 8.93294147979023e-06,
      "loss": 0.4945,
      "step": 1755
    },
    {
      "epoch": 0.8208955223880597,
      "grad_norm": 0.2928095193861828,
      "learning_rate": 8.835445022087426e-06,
      "loss": 0.5115,
      "step": 1760
    },
    {
      "epoch": 0.8232276119402985,
      "grad_norm": 0.28058171311068625,
      "learning_rate": 8.739059527631999e-06,
      "loss": 0.5119,
      "step": 1765
    },
    {
      "epoch": 0.8255597014925373,
      "grad_norm": 0.3094890046076664,
      "learning_rate": 8.6437907335413e-06,
      "loss": 0.5141,
      "step": 1770
    },
    {
      "epoch": 0.8278917910447762,
      "grad_norm": 0.2810800661440279,
      "learning_rate": 8.549644310463717e-06,
      "loss": 0.5167,
      "step": 1775
    },
    {
      "epoch": 0.8302238805970149,
      "grad_norm": 0.29240889207305526,
      "learning_rate": 8.456625862241193e-06,
      "loss": 0.5135,
      "step": 1780
    },
    {
      "epoch": 0.8325559701492538,
      "grad_norm": 0.292684734782098,
      "learning_rate": 8.364740925575643e-06,
      "loss": 0.5125,
      "step": 1785
    },
    {
      "epoch": 0.8348880597014925,
      "grad_norm": 0.2959795436894085,
      "learning_rate": 8.273994969699394e-06,
      "loss": 0.5057,
      "step": 1790
    },
    {
      "epoch": 0.8372201492537313,
      "grad_norm": 0.2954465382619872,
      "learning_rate": 8.184393396049675e-06,
      "loss": 0.5069,
      "step": 1795
    },
    {
      "epoch": 0.8395522388059702,
      "grad_norm": 0.3011785107092073,
      "learning_rate": 8.095941537947057e-06,
      "loss": 0.5176,
      "step": 1800
    },
    {
      "epoch": 0.8418843283582089,
      "grad_norm": 0.2912325453859299,
      "learning_rate": 8.008644660278051e-06,
      "loss": 0.4892,
      "step": 1805
    },
    {
      "epoch": 0.8442164179104478,
      "grad_norm": 0.3048172340443361,
      "learning_rate": 7.922507959181673e-06,
      "loss": 0.512,
      "step": 1810
    },
    {
      "epoch": 0.8465485074626866,
      "grad_norm": 0.2930943668283754,
      "learning_rate": 7.837536561740225e-06,
      "loss": 0.5033,
      "step": 1815
    },
    {
      "epoch": 0.8488805970149254,
      "grad_norm": 0.29936839076020144,
      "learning_rate": 7.753735525674059e-06,
      "loss": 0.5104,
      "step": 1820
    },
    {
      "epoch": 0.8512126865671642,
      "grad_norm": 0.2951288880142476,
      "learning_rate": 7.671109839040547e-06,
      "loss": 0.5126,
      "step": 1825
    },
    {
      "epoch": 0.8535447761194029,
      "grad_norm": 0.2790075544816064,
      "learning_rate": 7.58966441993719e-06,
      "loss": 0.494,
      "step": 1830
    },
    {
      "epoch": 0.8558768656716418,
      "grad_norm": 0.29641024588363307,
      "learning_rate": 7.509404116208868e-06,
      "loss": 0.5077,
      "step": 1835
    },
    {
      "epoch": 0.8582089552238806,
      "grad_norm": 0.3230633177700583,
      "learning_rate": 7.430333705159286e-06,
      "loss": 0.5401,
      "step": 1840
    },
    {
      "epoch": 0.8605410447761194,
      "grad_norm": 0.302196461115966,
      "learning_rate": 7.352457893266627e-06,
      "loss": 0.531,
      "step": 1845
    },
    {
      "epoch": 0.8628731343283582,
      "grad_norm": 0.29794221249791447,
      "learning_rate": 7.275781315903374e-06,
      "loss": 0.5137,
      "step": 1850
    },
    {
      "epoch": 0.8652052238805971,
      "grad_norm": 0.29632779382150637,
      "learning_rate": 7.20030853706046e-06,
      "loss": 0.5108,
      "step": 1855
    },
    {
      "epoch": 0.8675373134328358,
      "grad_norm": 0.29830642010980857,
      "learning_rate": 7.126044049075548e-06,
      "loss": 0.5163,
      "step": 1860
    },
    {
      "epoch": 0.8698694029850746,
      "grad_norm": 0.3124152954120725,
      "learning_rate": 7.052992272365681e-06,
      "loss": 0.5073,
      "step": 1865
    },
    {
      "epoch": 0.8722014925373134,
      "grad_norm": 0.30181418504381746,
      "learning_rate": 6.9811575551641224e-06,
      "loss": 0.4966,
      "step": 1870
    },
    {
      "epoch": 0.8745335820895522,
      "grad_norm": 0.29846707519728344,
      "learning_rate": 6.910544173261588e-06,
      "loss": 0.5003,
      "step": 1875
    },
    {
      "epoch": 0.8768656716417911,
      "grad_norm": 0.2951342743446569,
      "learning_rate": 6.8411563297516995e-06,
      "loss": 0.4912,
      "step": 1880
    },
    {
      "epoch": 0.8791977611940298,
      "grad_norm": 0.30200825628656813,
      "learning_rate": 6.772998154780832e-06,
      "loss": 0.5068,
      "step": 1885
    },
    {
      "epoch": 0.8815298507462687,
      "grad_norm": 0.30227641385772225,
      "learning_rate": 6.706073705302254e-06,
      "loss": 0.5067,
      "step": 1890
    },
    {
      "epoch": 0.8838619402985075,
      "grad_norm": 0.300849841970548,
      "learning_rate": 6.6403869648346634e-06,
      "loss": 0.5056,
      "step": 1895
    },
    {
      "epoch": 0.8861940298507462,
      "grad_norm": 0.29748535547416677,
      "learning_rate": 6.575941843225068e-06,
      "loss": 0.4889,
      "step": 1900
    },
    {
      "epoch": 0.8885261194029851,
      "grad_norm": 0.3017981544449824,
      "learning_rate": 6.5127421764160685e-06,
      "loss": 0.4947,
      "step": 1905
    },
    {
      "epoch": 0.8908582089552238,
      "grad_norm": 0.3026018984518875,
      "learning_rate": 6.450791726217538e-06,
      "loss": 0.5149,
      "step": 1910
    },
    {
      "epoch": 0.8931902985074627,
      "grad_norm": 0.3081030135585167,
      "learning_rate": 6.390094180082694e-06,
      "loss": 0.5079,
      "step": 1915
    },
    {
      "epoch": 0.8955223880597015,
      "grad_norm": 0.3041553330020546,
      "learning_rate": 6.330653150888617e-06,
      "loss": 0.509,
      "step": 1920
    },
    {
      "epoch": 0.8978544776119403,
      "grad_norm": 0.30475667040751553,
      "learning_rate": 6.272472176721207e-06,
      "loss": 0.5124,
      "step": 1925
    },
    {
      "epoch": 0.9001865671641791,
      "grad_norm": 0.29416760390485075,
      "learning_rate": 6.215554720664598e-06,
      "loss": 0.5163,
      "step": 1930
    },
    {
      "epoch": 0.902518656716418,
      "grad_norm": 0.27884692839413544,
      "learning_rate": 6.159904170594982e-06,
      "loss": 0.5069,
      "step": 1935
    },
    {
      "epoch": 0.9048507462686567,
      "grad_norm": 0.30054576904592345,
      "learning_rate": 6.105523838979022e-06,
      "loss": 0.5212,
      "step": 1940
    },
    {
      "epoch": 0.9071828358208955,
      "grad_norm": 0.30887707397671693,
      "learning_rate": 6.052416962676621e-06,
      "loss": 0.5012,
      "step": 1945
    },
    {
      "epoch": 0.9095149253731343,
      "grad_norm": 0.2901477305872294,
      "learning_rate": 6.000586702748301e-06,
      "loss": 0.4987,
      "step": 1950
    },
    {
      "epoch": 0.9118470149253731,
      "grad_norm": 0.3119574760410935,
      "learning_rate": 5.950036144267021e-06,
      "loss": 0.504,
      "step": 1955
    },
    {
      "epoch": 0.914179104477612,
      "grad_norm": 0.27442255971280144,
      "learning_rate": 5.900768296134551e-06,
      "loss": 0.4997,
      "step": 1960
    },
    {
      "epoch": 0.9165111940298507,
      "grad_norm": 0.31250294548724994,
      "learning_rate": 5.852786090902383e-06,
      "loss": 0.5129,
      "step": 1965
    },
    {
      "epoch": 0.9188432835820896,
      "grad_norm": 0.293909840986113,
      "learning_rate": 5.8060923845971825e-06,
      "loss": 0.506,
      "step": 1970
    },
    {
      "epoch": 0.9211753731343284,
      "grad_norm": 0.3001956020740078,
      "learning_rate": 5.760689956550763e-06,
      "loss": 0.5,
      "step": 1975
    },
    {
      "epoch": 0.9235074626865671,
      "grad_norm": 0.28174848937085023,
      "learning_rate": 5.7165815092346825e-06,
      "loss": 0.4912,
      "step": 1980
    },
    {
      "epoch": 0.925839552238806,
      "grad_norm": 0.3020312165517983,
      "learning_rate": 5.673769668099364e-06,
      "loss": 0.5031,
      "step": 1985
    },
    {
      "epoch": 0.9281716417910447,
      "grad_norm": 0.2943021791326706,
      "learning_rate": 5.632256981417845e-06,
      "loss": 0.5051,
      "step": 1990
    },
    {
      "epoch": 0.9305037313432836,
      "grad_norm": 0.2876667152171628,
      "learning_rate": 5.59204592013407e-06,
      "loss": 0.4933,
      "step": 1995
    },
    {
      "epoch": 0.9328358208955224,
      "grad_norm": 0.31254382539787895,
      "learning_rate": 5.553138877715833e-06,
      "loss": 0.5189,
      "step": 2000
    },
    {
      "epoch": 0.9351679104477612,
      "grad_norm": 0.2927974593197605,
      "learning_rate": 5.515538170012309e-06,
      "loss": 0.5028,
      "step": 2005
    },
    {
      "epoch": 0.9375,
      "grad_norm": 0.2909113245834731,
      "learning_rate": 5.479246035116201e-06,
      "loss": 0.495,
      "step": 2010
    },
    {
      "epoch": 0.9398320895522388,
      "grad_norm": 0.3194578057854513,
      "learning_rate": 5.444264633230531e-06,
      "loss": 0.5072,
      "step": 2015
    },
    {
      "epoch": 0.9421641791044776,
      "grad_norm": 0.2944404152818133,
      "learning_rate": 5.410596046540051e-06,
      "loss": 0.4952,
      "step": 2020
    },
    {
      "epoch": 0.9444962686567164,
      "grad_norm": 0.3115430812347992,
      "learning_rate": 5.378242279087314e-06,
      "loss": 0.501,
      "step": 2025
    },
    {
      "epoch": 0.9468283582089553,
      "grad_norm": 0.29847488276777256,
      "learning_rate": 5.347205256653387e-06,
      "loss": 0.5064,
      "step": 2030
    },
    {
      "epoch": 0.949160447761194,
      "grad_norm": 0.2890456115548558,
      "learning_rate": 5.317486826643219e-06,
      "loss": 0.486,
      "step": 2035
    },
    {
      "epoch": 0.9514925373134329,
      "grad_norm": 0.29211240702780633,
      "learning_rate": 5.28908875797568e-06,
      "loss": 0.5082,
      "step": 2040
    },
    {
      "epoch": 0.9538246268656716,
      "grad_norm": 0.3070505504678916,
      "learning_rate": 5.262012740978269e-06,
      "loss": 0.5097,
      "step": 2045
    },
    {
      "epoch": 0.9561567164179104,
      "grad_norm": 0.2773411345027328,
      "learning_rate": 5.236260387286509e-06,
      "loss": 0.4962,
      "step": 2050
    },
    {
      "epoch": 0.9584888059701493,
      "grad_norm": 0.29799985399558016,
      "learning_rate": 5.2118332297480105e-06,
      "loss": 0.5128,
      "step": 2055
    },
    {
      "epoch": 0.960820895522388,
      "grad_norm": 0.287169325155614,
      "learning_rate": 5.1887327223312296e-06,
      "loss": 0.4866,
      "step": 2060
    },
    {
      "epoch": 0.9631529850746269,
      "grad_norm": 0.31835555206048355,
      "learning_rate": 5.166960240038937e-06,
      "loss": 0.504,
      "step": 2065
    },
    {
      "epoch": 0.9654850746268657,
      "grad_norm": 0.3095841539651552,
      "learning_rate": 5.1465170788263595e-06,
      "loss": 0.5117,
      "step": 2070
    },
    {
      "epoch": 0.9678171641791045,
      "grad_norm": 0.3262964021293147,
      "learning_rate": 5.1274044555240525e-06,
      "loss": 0.502,
      "step": 2075
    },
    {
      "epoch": 0.9701492537313433,
      "grad_norm": 0.29662543157502314,
      "learning_rate": 5.109623507765466e-06,
      "loss": 0.5042,
      "step": 2080
    },
    {
      "epoch": 0.972481343283582,
      "grad_norm": 0.29199443293251653,
      "learning_rate": 5.093175293919228e-06,
      "loss": 0.5081,
      "step": 2085
    },
    {
      "epoch": 0.9748134328358209,
      "grad_norm": 0.29856223815679944,
      "learning_rate": 5.07806079302615e-06,
      "loss": 0.5128,
      "step": 2090
    },
    {
      "epoch": 0.9771455223880597,
      "grad_norm": 0.2812213514715703,
      "learning_rate": 5.064280904740953e-06,
      "loss": 0.5097,
      "step": 2095
    },
    {
      "epoch": 0.9794776119402985,
      "grad_norm": 0.29489687258433694,
      "learning_rate": 5.051836449278715e-06,
      "loss": 0.4999,
      "step": 2100
    },
    {
      "epoch": 0.9818097014925373,
      "grad_norm": 0.30265017359808566,
      "learning_rate": 5.040728167366057e-06,
      "loss": 0.4966,
      "step": 2105
    },
    {
      "epoch": 0.9841417910447762,
      "grad_norm": 0.2991437683499533,
      "learning_rate": 5.030956720197035e-06,
      "loss": 0.4991,
      "step": 2110
    },
    {
      "epoch": 0.9864738805970149,
      "grad_norm": 0.2993425987702054,
      "learning_rate": 5.022522689393809e-06,
      "loss": 0.5269,
      "step": 2115
    },
    {
      "epoch": 0.9888059701492538,
      "grad_norm": 0.28083575851271286,
      "learning_rate": 5.015426576972003e-06,
      "loss": 0.5034,
      "step": 2120
    },
    {
      "epoch": 0.9911380597014925,
      "grad_norm": 0.2846014329272288,
      "learning_rate": 5.009668805310832e-06,
      "loss": 0.5001,
      "step": 2125
    },
    {
      "epoch": 0.9934701492537313,
      "grad_norm": 0.297526890423307,
      "learning_rate": 5.005249717127964e-06,
      "loss": 0.5124,
      "step": 2130
    },
    {
      "epoch": 0.9958022388059702,
      "grad_norm": 0.3006228142618389,
      "learning_rate": 5.002169575459111e-06,
      "loss": 0.5046,
      "step": 2135
    },
    {
      "epoch": 0.9981343283582089,
      "grad_norm": 0.3155643221180713,
      "learning_rate": 5.000428563642382e-06,
      "loss": 0.5086,
      "step": 2140
    },
    {
      "epoch": 1.0,
      "step": 2144,
      "total_flos": 488578517827584.0,
      "train_loss": 0.5429407808540473,
      "train_runtime": 24734.1783,
      "train_samples_per_second": 1.387,
      "train_steps_per_second": 0.087
    }
  ],
  "logging_steps": 5,
  "max_steps": 2144,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 488578517827584.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}