{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 4819,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00020751193193608634,
      "grad_norm": 23.81510217286444,
      "learning_rate": 2.0746887966804982e-08,
      "loss": 1.3923,
      "step": 1
    },
    {
      "epoch": 0.0010375596596804316,
      "grad_norm": 23.986426679578184,
      "learning_rate": 1.037344398340249e-07,
      "loss": 1.4149,
      "step": 5
    },
    {
      "epoch": 0.002075119319360863,
      "grad_norm": 21.906176707284583,
      "learning_rate": 2.074688796680498e-07,
      "loss": 1.4046,
      "step": 10
    },
    {
      "epoch": 0.003112678979041295,
      "grad_norm": 15.167476223997951,
      "learning_rate": 3.112033195020747e-07,
      "loss": 1.3592,
      "step": 15
    },
    {
      "epoch": 0.004150238638721726,
      "grad_norm": 9.140589856353586,
      "learning_rate": 4.149377593360996e-07,
      "loss": 1.2548,
      "step": 20
    },
    {
      "epoch": 0.005187798298402158,
      "grad_norm": 10.826361201441884,
      "learning_rate": 5.186721991701245e-07,
      "loss": 1.1628,
      "step": 25
    },
    {
      "epoch": 0.00622535795808259,
      "grad_norm": 9.062979414726044,
      "learning_rate": 6.224066390041494e-07,
      "loss": 1.0633,
      "step": 30
    },
    {
      "epoch": 0.007262917617763021,
      "grad_norm": 3.7471401092788903,
      "learning_rate": 7.261410788381744e-07,
      "loss": 1.0125,
      "step": 35
    },
    {
      "epoch": 0.008300477277443452,
      "grad_norm": 3.4052281488293517,
      "learning_rate": 8.298755186721992e-07,
      "loss": 0.9673,
      "step": 40
    },
    {
      "epoch": 0.009338036937123885,
      "grad_norm": 3.4155319860158584,
      "learning_rate": 9.336099585062241e-07,
      "loss": 0.9582,
      "step": 45
    },
    {
      "epoch": 0.010375596596804317,
      "grad_norm": 3.0460439284849947,
      "learning_rate": 1.037344398340249e-06,
      "loss": 0.9393,
      "step": 50
    },
    {
      "epoch": 0.011413156256484747,
      "grad_norm": 3.0573711664380117,
      "learning_rate": 1.141078838174274e-06,
      "loss": 0.928,
      "step": 55
    },
    {
      "epoch": 0.01245071591616518,
      "grad_norm": 3.104368834698698,
      "learning_rate": 1.2448132780082988e-06,
      "loss": 0.9128,
      "step": 60
    },
    {
      "epoch": 0.013488275575845612,
      "grad_norm": 3.1380190764833094,
      "learning_rate": 1.3485477178423237e-06,
      "loss": 0.9156,
      "step": 65
    },
    {
      "epoch": 0.014525835235526042,
      "grad_norm": 3.1006570296836182,
      "learning_rate": 1.4522821576763488e-06,
      "loss": 0.905,
      "step": 70
    },
    {
      "epoch": 0.015563394895206474,
      "grad_norm": 3.127744100257649,
      "learning_rate": 1.5560165975103735e-06,
      "loss": 0.9163,
      "step": 75
    },
    {
      "epoch": 0.016600954554886905,
      "grad_norm": 3.038437602227199,
      "learning_rate": 1.6597510373443984e-06,
      "loss": 0.8884,
      "step": 80
    },
    {
      "epoch": 0.017638514214567337,
      "grad_norm": 2.9596644094641413,
      "learning_rate": 1.7634854771784235e-06,
      "loss": 0.8923,
      "step": 85
    },
    {
      "epoch": 0.01867607387424777,
      "grad_norm": 2.958917381373649,
      "learning_rate": 1.8672199170124482e-06,
      "loss": 0.8939,
      "step": 90
    },
    {
      "epoch": 0.0197136335339282,
      "grad_norm": 3.13129582575627,
      "learning_rate": 1.970954356846473e-06,
      "loss": 0.8671,
      "step": 95
    },
    {
      "epoch": 0.020751193193608634,
      "grad_norm": 3.212698731191082,
      "learning_rate": 2.074688796680498e-06,
      "loss": 0.8846,
      "step": 100
    },
    {
      "epoch": 0.021788752853289062,
      "grad_norm": 2.9778959899019273,
      "learning_rate": 2.178423236514523e-06,
      "loss": 0.8747,
      "step": 105
    },
    {
      "epoch": 0.022826312512969495,
      "grad_norm": 3.02722706158588,
      "learning_rate": 2.282157676348548e-06,
      "loss": 0.8841,
      "step": 110
    },
    {
      "epoch": 0.023863872172649927,
      "grad_norm": 3.104436315782777,
      "learning_rate": 2.385892116182573e-06,
      "loss": 0.8646,
      "step": 115
    },
    {
      "epoch": 0.02490143183233036,
      "grad_norm": 3.0389426041639847,
      "learning_rate": 2.4896265560165977e-06,
      "loss": 0.8769,
      "step": 120
    },
    {
      "epoch": 0.02593899149201079,
      "grad_norm": 3.0481198307330772,
      "learning_rate": 2.5933609958506228e-06,
      "loss": 0.8597,
      "step": 125
    },
    {
      "epoch": 0.026976551151691223,
      "grad_norm": 3.02592811695882,
      "learning_rate": 2.6970954356846475e-06,
      "loss": 0.8668,
      "step": 130
    },
    {
      "epoch": 0.028014110811371652,
      "grad_norm": 3.2120467143006954,
      "learning_rate": 2.8008298755186726e-06,
      "loss": 0.8651,
      "step": 135
    },
    {
      "epoch": 0.029051670471052084,
      "grad_norm": 2.9463572893956633,
      "learning_rate": 2.9045643153526977e-06,
      "loss": 0.8648,
      "step": 140
    },
    {
      "epoch": 0.030089230130732517,
      "grad_norm": 3.2030553684980054,
      "learning_rate": 3.008298755186722e-06,
      "loss": 0.861,
      "step": 145
    },
    {
      "epoch": 0.03112678979041295,
      "grad_norm": 3.134802130941582,
      "learning_rate": 3.112033195020747e-06,
      "loss": 0.8572,
      "step": 150
    },
    {
      "epoch": 0.03216434945009338,
      "grad_norm": 3.0051811744191075,
      "learning_rate": 3.215767634854772e-06,
      "loss": 0.8548,
      "step": 155
    },
    {
      "epoch": 0.03320190910977381,
      "grad_norm": 2.9087769254486013,
      "learning_rate": 3.319502074688797e-06,
      "loss": 0.857,
      "step": 160
    },
    {
      "epoch": 0.034239468769454245,
      "grad_norm": 3.251836422833884,
      "learning_rate": 3.423236514522822e-06,
      "loss": 0.8479,
      "step": 165
    },
    {
      "epoch": 0.035277028429134674,
      "grad_norm": 2.988883877751971,
      "learning_rate": 3.526970954356847e-06,
      "loss": 0.8573,
      "step": 170
    },
    {
      "epoch": 0.03631458808881511,
      "grad_norm": 2.9670335974516155,
      "learning_rate": 3.6307053941908714e-06,
      "loss": 0.8509,
      "step": 175
    },
    {
      "epoch": 0.03735214774849554,
      "grad_norm": 3.0847182089806333,
      "learning_rate": 3.7344398340248965e-06,
      "loss": 0.8421,
      "step": 180
    },
    {
      "epoch": 0.03838970740817597,
      "grad_norm": 2.8264906420734843,
      "learning_rate": 3.838174273858922e-06,
      "loss": 0.8484,
      "step": 185
    },
    {
      "epoch": 0.0394272670678564,
      "grad_norm": 2.974170420918456,
      "learning_rate": 3.941908713692946e-06,
      "loss": 0.8474,
      "step": 190
    },
    {
      "epoch": 0.04046482672753683,
      "grad_norm": 3.077208581236518,
      "learning_rate": 4.045643153526971e-06,
      "loss": 0.8275,
      "step": 195
    },
    {
      "epoch": 0.04150238638721727,
      "grad_norm": 3.134791812920161,
      "learning_rate": 4.149377593360996e-06,
      "loss": 0.8419,
      "step": 200
    },
    {
      "epoch": 0.042539946046897696,
      "grad_norm": 3.2479409082629935,
      "learning_rate": 4.253112033195021e-06,
      "loss": 0.8466,
      "step": 205
    },
    {
      "epoch": 0.043577505706578125,
      "grad_norm": 2.8988040172466767,
      "learning_rate": 4.356846473029046e-06,
      "loss": 0.8489,
      "step": 210
    },
    {
      "epoch": 0.04461506536625856,
      "grad_norm": 3.0923511376582797,
      "learning_rate": 4.460580912863071e-06,
      "loss": 0.8455,
      "step": 215
    },
    {
      "epoch": 0.04565262502593899,
      "grad_norm": 3.1659114421543895,
      "learning_rate": 4.564315352697096e-06,
      "loss": 0.8366,
      "step": 220
    },
    {
      "epoch": 0.046690184685619425,
      "grad_norm": 2.9510079953090953,
      "learning_rate": 4.66804979253112e-06,
      "loss": 0.8492,
      "step": 225
    },
    {
      "epoch": 0.047727744345299854,
      "grad_norm": 2.9528441973334503,
      "learning_rate": 4.771784232365146e-06,
      "loss": 0.8317,
      "step": 230
    },
    {
      "epoch": 0.04876530400498029,
      "grad_norm": 3.009602261579148,
      "learning_rate": 4.875518672199171e-06,
      "loss": 0.8282,
      "step": 235
    },
    {
      "epoch": 0.04980286366466072,
      "grad_norm": 3.0538602837457702,
      "learning_rate": 4.979253112033195e-06,
      "loss": 0.8379,
      "step": 240
    },
    {
      "epoch": 0.05084042332434115,
      "grad_norm": 3.067939083048346,
      "learning_rate": 5.08298755186722e-06,
      "loss": 0.8255,
      "step": 245
    },
    {
      "epoch": 0.05187798298402158,
      "grad_norm": 2.8381217796695233,
      "learning_rate": 5.1867219917012455e-06,
      "loss": 0.8424,
      "step": 250
    },
    {
      "epoch": 0.05291554264370201,
      "grad_norm": 3.4784967911008247,
      "learning_rate": 5.29045643153527e-06,
      "loss": 0.8288,
      "step": 255
    },
    {
      "epoch": 0.05395310230338245,
      "grad_norm": 3.308889130459959,
      "learning_rate": 5.394190871369295e-06,
      "loss": 0.8458,
      "step": 260
    },
    {
      "epoch": 0.054990661963062876,
      "grad_norm": 3.043690822549995,
      "learning_rate": 5.4979253112033204e-06,
      "loss": 0.8233,
      "step": 265
    },
    {
      "epoch": 0.056028221622743304,
      "grad_norm": 3.168880712945065,
      "learning_rate": 5.601659751037345e-06,
      "loss": 0.831,
      "step": 270
    },
    {
      "epoch": 0.05706578128242374,
      "grad_norm": 3.1149299465392644,
      "learning_rate": 5.70539419087137e-06,
      "loss": 0.8363,
      "step": 275
    },
    {
      "epoch": 0.05810334094210417,
      "grad_norm": 3.1047167518774814,
      "learning_rate": 5.809128630705395e-06,
      "loss": 0.8391,
      "step": 280
    },
    {
      "epoch": 0.059140900601784605,
      "grad_norm": 2.8999047365073927,
      "learning_rate": 5.91286307053942e-06,
      "loss": 0.8368,
      "step": 285
    },
    {
      "epoch": 0.06017846026146503,
      "grad_norm": 3.066022536608264,
      "learning_rate": 6.016597510373444e-06,
      "loss": 0.8379,
      "step": 290
    },
    {
      "epoch": 0.06121601992114547,
      "grad_norm": 3.116983544786236,
      "learning_rate": 6.1203319502074694e-06,
      "loss": 0.8169,
      "step": 295
    },
    {
      "epoch": 0.0622535795808259,
      "grad_norm": 7.508628361145243,
      "learning_rate": 6.224066390041494e-06,
      "loss": 0.8186,
      "step": 300
    },
    {
      "epoch": 0.06329113924050633,
      "grad_norm": 2.959364581244697,
      "learning_rate": 6.327800829875519e-06,
      "loss": 0.823,
      "step": 305
    },
    {
      "epoch": 0.06432869890018676,
      "grad_norm": 3.3247616385013554,
      "learning_rate": 6.431535269709544e-06,
      "loss": 0.8336,
      "step": 310
    },
    {
      "epoch": 0.06536625855986719,
      "grad_norm": 2.97755825450701,
      "learning_rate": 6.535269709543569e-06,
      "loss": 0.8403,
      "step": 315
    },
    {
      "epoch": 0.06640381821954762,
      "grad_norm": 2.920924218399819,
      "learning_rate": 6.639004149377594e-06,
      "loss": 0.8352,
      "step": 320
    },
    {
      "epoch": 0.06744137787922806,
      "grad_norm": 2.996226962381477,
      "learning_rate": 6.742738589211619e-06,
      "loss": 0.8149,
      "step": 325
    },
    {
      "epoch": 0.06847893753890849,
      "grad_norm": 2.9599991745332743,
      "learning_rate": 6.846473029045644e-06,
      "loss": 0.8227,
      "step": 330
    },
    {
      "epoch": 0.06951649719858892,
      "grad_norm": 3.3083929869948374,
      "learning_rate": 6.950207468879669e-06,
      "loss": 0.8293,
      "step": 335
    },
    {
      "epoch": 0.07055405685826935,
      "grad_norm": 2.921592069295814,
      "learning_rate": 7.053941908713694e-06,
      "loss": 0.8063,
      "step": 340
    },
    {
      "epoch": 0.07159161651794978,
      "grad_norm": 2.9356160908302296,
      "learning_rate": 7.157676348547719e-06,
      "loss": 0.8249,
      "step": 345
    },
    {
      "epoch": 0.07262917617763022,
      "grad_norm": 3.127667490017175,
      "learning_rate": 7.261410788381743e-06,
      "loss": 0.805,
      "step": 350
    },
    {
      "epoch": 0.07366673583731065,
      "grad_norm": 2.811701116040028,
      "learning_rate": 7.365145228215769e-06,
      "loss": 0.8141,
      "step": 355
    },
    {
      "epoch": 0.07470429549699108,
      "grad_norm": 3.428565494126869,
      "learning_rate": 7.468879668049793e-06,
      "loss": 0.8137,
      "step": 360
    },
    {
      "epoch": 0.0757418551566715,
      "grad_norm": 3.0132405721717794,
      "learning_rate": 7.572614107883818e-06,
      "loss": 0.8121,
      "step": 365
    },
    {
      "epoch": 0.07677941481635193,
      "grad_norm": 3.0044504000522045,
      "learning_rate": 7.676348547717844e-06,
      "loss": 0.8112,
      "step": 370
    },
    {
      "epoch": 0.07781697447603238,
      "grad_norm": 2.938773166801736,
      "learning_rate": 7.780082987551869e-06,
      "loss": 0.8071,
      "step": 375
    },
    {
      "epoch": 0.0788545341357128,
      "grad_norm": 2.816233639978962,
      "learning_rate": 7.883817427385892e-06,
      "loss": 0.8269,
      "step": 380
    },
    {
      "epoch": 0.07989209379539323,
      "grad_norm": 3.2016060279962297,
      "learning_rate": 7.987551867219918e-06,
      "loss": 0.8173,
      "step": 385
    },
    {
      "epoch": 0.08092965345507366,
      "grad_norm": 3.0338701288383567,
      "learning_rate": 8.091286307053943e-06,
      "loss": 0.7997,
      "step": 390
    },
    {
      "epoch": 0.08196721311475409,
      "grad_norm": 2.8900162172973025,
      "learning_rate": 8.195020746887967e-06,
      "loss": 0.8296,
      "step": 395
    },
    {
      "epoch": 0.08300477277443453,
      "grad_norm": 2.775870448004777,
      "learning_rate": 8.298755186721992e-06,
      "loss": 0.8139,
      "step": 400
    },
    {
      "epoch": 0.08404233243411496,
      "grad_norm": 2.997946086685837,
      "learning_rate": 8.402489626556017e-06,
      "loss": 0.8244,
      "step": 405
    },
    {
      "epoch": 0.08507989209379539,
      "grad_norm": 2.79051986833281,
      "learning_rate": 8.506224066390042e-06,
      "loss": 0.8207,
      "step": 410
    },
    {
      "epoch": 0.08611745175347582,
      "grad_norm": 2.8248639051971334,
      "learning_rate": 8.609958506224068e-06,
      "loss": 0.805,
      "step": 415
    },
    {
      "epoch": 0.08715501141315625,
      "grad_norm": 2.9096736006801485,
      "learning_rate": 8.713692946058093e-06,
      "loss": 0.8037,
      "step": 420
    },
    {
      "epoch": 0.08819257107283669,
      "grad_norm": 2.940919950600375,
      "learning_rate": 8.817427385892117e-06,
      "loss": 0.8186,
      "step": 425
    },
    {
      "epoch": 0.08923013073251712,
      "grad_norm": 2.896130509805648,
      "learning_rate": 8.921161825726142e-06,
      "loss": 0.8245,
      "step": 430
    },
    {
      "epoch": 0.09026769039219755,
      "grad_norm": 2.8401858033389713,
      "learning_rate": 9.024896265560167e-06,
      "loss": 0.8081,
      "step": 435
    },
    {
      "epoch": 0.09130525005187798,
      "grad_norm": 2.86912546506336,
      "learning_rate": 9.128630705394191e-06,
      "loss": 0.8164,
      "step": 440
    },
    {
      "epoch": 0.09234280971155842,
      "grad_norm": 2.7242600466916453,
      "learning_rate": 9.232365145228218e-06,
      "loss": 0.8073,
      "step": 445
    },
    {
      "epoch": 0.09338036937123885,
      "grad_norm": 3.044646841531881,
      "learning_rate": 9.33609958506224e-06,
      "loss": 0.8145,
      "step": 450
    },
    {
      "epoch": 0.09441792903091928,
      "grad_norm": 2.8295153855299198,
      "learning_rate": 9.439834024896265e-06,
      "loss": 0.7957,
      "step": 455
    },
    {
      "epoch": 0.09545548869059971,
      "grad_norm": 2.910948125766166,
      "learning_rate": 9.543568464730292e-06,
      "loss": 0.8053,
      "step": 460
    },
    {
      "epoch": 0.09649304835028014,
      "grad_norm": 2.8219575214924597,
      "learning_rate": 9.647302904564317e-06,
      "loss": 0.8128,
      "step": 465
    },
    {
      "epoch": 0.09753060800996058,
      "grad_norm": 2.9677851670315167,
      "learning_rate": 9.751037344398341e-06,
      "loss": 0.8061,
      "step": 470
    },
    {
      "epoch": 0.09856816766964101,
      "grad_norm": 3.1531286571785326,
      "learning_rate": 9.854771784232366e-06,
      "loss": 0.7962,
      "step": 475
    },
    {
      "epoch": 0.09960572732932144,
      "grad_norm": 3.1075603430133105,
      "learning_rate": 9.95850622406639e-06,
      "loss": 0.8072,
      "step": 480
    },
    {
      "epoch": 0.10064328698900186,
      "grad_norm": 2.844518989686556,
      "learning_rate": 9.99998819398724e-06,
      "loss": 0.8198,
      "step": 485
    },
    {
      "epoch": 0.1016808466486823,
      "grad_norm": 2.8894422413511127,
      "learning_rate": 9.999916046333384e-06,
      "loss": 0.8146,
      "step": 490
    },
    {
      "epoch": 0.10271840630836274,
      "grad_norm": 2.8815754440809878,
      "learning_rate": 9.999778310866921e-06,
      "loss": 0.7899,
      "step": 495
    },
    {
      "epoch": 0.10375596596804317,
      "grad_norm": 2.885119932296298,
      "learning_rate": 9.999574989394634e-06,
      "loss": 0.8013,
      "step": 500
    },
    {
      "epoch": 0.1047935256277236,
      "grad_norm": 2.8642122606932174,
      "learning_rate": 9.99930608458365e-06,
      "loss": 0.805,
      "step": 505
    },
    {
      "epoch": 0.10583108528740402,
      "grad_norm": 2.9310957652228082,
      "learning_rate": 9.998971599961405e-06,
      "loss": 0.7915,
      "step": 510
    },
    {
      "epoch": 0.10686864494708445,
      "grad_norm": 2.8531320874329746,
      "learning_rate": 9.998571539915592e-06,
      "loss": 0.7981,
      "step": 515
    },
    {
      "epoch": 0.1079062046067649,
      "grad_norm": 2.839970691252372,
      "learning_rate": 9.998105909694117e-06,
      "loss": 0.7999,
      "step": 520
    },
    {
      "epoch": 0.10894376426644532,
      "grad_norm": 2.853880263858815,
      "learning_rate": 9.997574715405011e-06,
      "loss": 0.8311,
      "step": 525
    },
    {
      "epoch": 0.10998132392612575,
      "grad_norm": 2.9338168378898337,
      "learning_rate": 9.996977964016371e-06,
      "loss": 0.8005,
      "step": 530
    },
    {
      "epoch": 0.11101888358580618,
      "grad_norm": 2.7032896393481254,
      "learning_rate": 9.996315663356247e-06,
      "loss": 0.8003,
      "step": 535
    },
    {
      "epoch": 0.11205644324548661,
      "grad_norm": 2.761074984380135,
      "learning_rate": 9.995587822112558e-06,
      "loss": 0.8044,
      "step": 540
    },
    {
      "epoch": 0.11309400290516705,
      "grad_norm": 2.6815711646329556,
      "learning_rate": 9.994794449832966e-06,
      "loss": 0.7887,
      "step": 545
    },
    {
      "epoch": 0.11413156256484748,
      "grad_norm": 2.8649570007362657,
      "learning_rate": 9.993935556924756e-06,
      "loss": 0.7776,
      "step": 550
    },
    {
      "epoch": 0.11516912222452791,
      "grad_norm": 2.9788872729907547,
      "learning_rate": 9.993011154654702e-06,
      "loss": 0.7778,
      "step": 555
    },
    {
      "epoch": 0.11620668188420834,
      "grad_norm": 2.771583627699142,
      "learning_rate": 9.992021255148907e-06,
      "loss": 0.7876,
      "step": 560
    },
    {
      "epoch": 0.11724424154388878,
      "grad_norm": 3.156864019145989,
      "learning_rate": 9.990965871392662e-06,
      "loss": 0.7924,
      "step": 565
    },
    {
      "epoch": 0.11828180120356921,
      "grad_norm": 2.6854537961356653,
      "learning_rate": 9.989845017230258e-06,
      "loss": 0.7841,
      "step": 570
    },
    {
      "epoch": 0.11931936086324964,
      "grad_norm": 3.194287668643439,
      "learning_rate": 9.988658707364819e-06,
      "loss": 0.7807,
      "step": 575
    },
    {
      "epoch": 0.12035692052293007,
      "grad_norm": 2.7638444846888173,
      "learning_rate": 9.9874069573581e-06,
      "loss": 0.7846,
      "step": 580
    },
    {
      "epoch": 0.1213944801826105,
      "grad_norm": 2.7802968983308936,
      "learning_rate": 9.986089783630286e-06,
      "loss": 0.775,
      "step": 585
    },
    {
      "epoch": 0.12243203984229094,
      "grad_norm": 2.8469947394820103,
      "learning_rate": 9.984707203459774e-06,
      "loss": 0.7672,
      "step": 590
    },
    {
      "epoch": 0.12346959950197137,
      "grad_norm": 2.9990572740062493,
      "learning_rate": 9.983259234982951e-06,
      "loss": 0.7779,
      "step": 595
    },
    {
      "epoch": 0.1245071591616518,
      "grad_norm": 3.1849226285333345,
      "learning_rate": 9.981745897193955e-06,
      "loss": 0.7714,
      "step": 600
    },
    {
      "epoch": 0.12554471882133222,
      "grad_norm": 2.835199675947756,
      "learning_rate": 9.98016720994442e-06,
      "loss": 0.7784,
      "step": 605
    },
    {
      "epoch": 0.12658227848101267,
      "grad_norm": 2.8294757615641886,
      "learning_rate": 9.978523193943222e-06,
      "loss": 0.7905,
      "step": 610
    },
    {
      "epoch": 0.12761983814069308,
      "grad_norm": 2.6228325540550657,
      "learning_rate": 9.976813870756209e-06,
      "loss": 0.7695,
      "step": 615
    },
    {
      "epoch": 0.12865739780037352,
      "grad_norm": 2.6754726256364902,
      "learning_rate": 9.975039262805907e-06,
      "loss": 0.7784,
      "step": 620
    },
    {
      "epoch": 0.12969495746005397,
      "grad_norm": 2.6428193830815663,
      "learning_rate": 9.973199393371242e-06,
      "loss": 0.7768,
      "step": 625
    },
    {
      "epoch": 0.13073251711973438,
      "grad_norm": 2.7019164377169824,
      "learning_rate": 9.97129428658722e-06,
      "loss": 0.7787,
      "step": 630
    },
    {
      "epoch": 0.13177007677941482,
      "grad_norm": 2.7193094088840635,
      "learning_rate": 9.969323967444616e-06,
      "loss": 0.7691,
      "step": 635
    },
    {
      "epoch": 0.13280763643909524,
      "grad_norm": 2.9547462574056227,
      "learning_rate": 9.96728846178965e-06,
      "loss": 0.7791,
      "step": 640
    },
    {
      "epoch": 0.13384519609877568,
      "grad_norm": 2.7070337933579673,
      "learning_rate": 9.965187796323643e-06,
      "loss": 0.7793,
      "step": 645
    },
    {
      "epoch": 0.13488275575845612,
      "grad_norm": 2.846568092264262,
      "learning_rate": 9.96302199860267e-06,
      "loss": 0.7657,
      "step": 650
    },
    {
      "epoch": 0.13592031541813654,
      "grad_norm": 2.6430732770071885,
      "learning_rate": 9.96079109703719e-06,
      "loss": 0.7613,
      "step": 655
    },
    {
      "epoch": 0.13695787507781698,
      "grad_norm": 2.564352165321478,
      "learning_rate": 9.95849512089169e-06,
      "loss": 0.7716,
      "step": 660
    },
    {
      "epoch": 0.1379954347374974,
      "grad_norm": 2.787666582812226,
      "learning_rate": 9.956134100284285e-06,
      "loss": 0.7788,
      "step": 665
    },
    {
      "epoch": 0.13903299439717784,
      "grad_norm": 2.565898584923949,
      "learning_rate": 9.95370806618633e-06,
      "loss": 0.7612,
      "step": 670
    },
    {
      "epoch": 0.14007055405685828,
      "grad_norm": 2.9186277859262244,
      "learning_rate": 9.951217050422013e-06,
      "loss": 0.7787,
      "step": 675
    },
    {
      "epoch": 0.1411081137165387,
      "grad_norm": 2.7573443860042546,
      "learning_rate": 9.94866108566794e-06,
      "loss": 0.7556,
      "step": 680
    },
    {
      "epoch": 0.14214567337621914,
      "grad_norm": 2.840469194753,
      "learning_rate": 9.946040205452699e-06,
      "loss": 0.7456,
      "step": 685
    },
    {
      "epoch": 0.14318323303589955,
      "grad_norm": 2.55261542401345,
      "learning_rate": 9.943354444156428e-06,
      "loss": 0.7789,
      "step": 690
    },
    {
      "epoch": 0.14422079269558,
      "grad_norm": 2.6744948199976535,
      "learning_rate": 9.940603837010358e-06,
      "loss": 0.773,
      "step": 695
    },
    {
      "epoch": 0.14525835235526044,
      "grad_norm": 2.7013083342024107,
      "learning_rate": 9.937788420096362e-06,
      "loss": 0.7735,
      "step": 700
    },
    {
      "epoch": 0.14629591201494085,
      "grad_norm": 2.731487086002391,
      "learning_rate": 9.934908230346462e-06,
      "loss": 0.7523,
      "step": 705
    },
    {
      "epoch": 0.1473334716746213,
      "grad_norm": 2.6033671380903334,
      "learning_rate": 9.931963305542363e-06,
      "loss": 0.7517,
      "step": 710
    },
    {
      "epoch": 0.1483710313343017,
      "grad_norm": 2.662624062037032,
      "learning_rate": 9.92895368431495e-06,
      "loss": 0.7659,
      "step": 715
    },
    {
      "epoch": 0.14940859099398215,
      "grad_norm": 3.714026552589638,
      "learning_rate": 9.925879406143779e-06,
      "loss": 0.7646,
      "step": 720
    },
    {
      "epoch": 0.1504461506536626,
      "grad_norm": 3.3049933455540073,
      "learning_rate": 9.922740511356565e-06,
      "loss": 0.7681,
      "step": 725
    },
    {
      "epoch": 0.151483710313343,
      "grad_norm": 2.7320515915400816,
      "learning_rate": 9.919537041128647e-06,
      "loss": 0.746,
      "step": 730
    },
    {
      "epoch": 0.15252126997302345,
      "grad_norm": 2.635817269274888,
      "learning_rate": 9.916269037482452e-06,
      "loss": 0.7306,
      "step": 735
    },
    {
      "epoch": 0.15355882963270387,
      "grad_norm": 2.662103702116443,
      "learning_rate": 9.912936543286939e-06,
      "loss": 0.7536,
      "step": 740
    },
    {
      "epoch": 0.1545963892923843,
      "grad_norm": 2.85326362399482,
      "learning_rate": 9.909539602257048e-06,
      "loss": 0.7673,
      "step": 745
    },
    {
      "epoch": 0.15563394895206475,
      "grad_norm": 2.567231081949611,
      "learning_rate": 9.90607825895311e-06,
      "loss": 0.738,
      "step": 750
    },
    {
      "epoch": 0.15667150861174517,
      "grad_norm": 2.869479144385386,
      "learning_rate": 9.902552558780276e-06,
      "loss": 0.7598,
      "step": 755
    },
    {
      "epoch": 0.1577090682714256,
      "grad_norm": 2.730929830901487,
      "learning_rate": 9.898962547987913e-06,
      "loss": 0.748,
      "step": 760
    },
    {
      "epoch": 0.15874662793110603,
      "grad_norm": 2.793985046920194,
      "learning_rate": 9.895308273669007e-06,
      "loss": 0.7328,
      "step": 765
    },
    {
      "epoch": 0.15978418759078647,
      "grad_norm": 2.7966738638089246,
      "learning_rate": 9.89158978375953e-06,
      "loss": 0.7676,
      "step": 770
    },
    {
      "epoch": 0.1608217472504669,
      "grad_norm": 2.6679897290967576,
      "learning_rate": 9.887807127037827e-06,
      "loss": 0.7295,
      "step": 775
    },
    {
      "epoch": 0.16185930691014733,
      "grad_norm": 2.5457872650558455,
      "learning_rate": 9.88396035312397e-06,
      "loss": 0.728,
      "step": 780
    },
    {
      "epoch": 0.16289686656982777,
      "grad_norm": 2.8332820570127626,
      "learning_rate": 9.880049512479097e-06,
      "loss": 0.7421,
      "step": 785
    },
    {
      "epoch": 0.16393442622950818,
      "grad_norm": 2.7639539461730114,
      "learning_rate": 9.876074656404773e-06,
      "loss": 0.7534,
      "step": 790
    },
    {
      "epoch": 0.16497198588918863,
      "grad_norm": 2.7213075641667928,
      "learning_rate": 9.872035837042292e-06,
      "loss": 0.7363,
      "step": 795
    },
    {
      "epoch": 0.16600954554886907,
      "grad_norm": 2.6178852546410183,
      "learning_rate": 9.86793310737201e-06,
      "loss": 0.7318,
      "step": 800
    },
    {
      "epoch": 0.16704710520854948,
      "grad_norm": 2.7566382614918585,
      "learning_rate": 9.863766521212646e-06,
      "loss": 0.7507,
      "step": 805
    },
    {
      "epoch": 0.16808466486822993,
      "grad_norm": 2.75430515948912,
      "learning_rate": 9.859536133220569e-06,
      "loss": 0.7481,
      "step": 810
    },
    {
      "epoch": 0.16912222452791034,
      "grad_norm": 2.7773285238684813,
      "learning_rate": 9.855241998889091e-06,
      "loss": 0.7456,
      "step": 815
    },
    {
      "epoch": 0.17015978418759078,
      "grad_norm": 2.9942393187975904,
      "learning_rate": 9.850884174547734e-06,
      "loss": 0.7512,
      "step": 820
    },
    {
      "epoch": 0.17119734384727123,
      "grad_norm": 2.8176703833149706,
      "learning_rate": 9.846462717361489e-06,
      "loss": 0.7229,
      "step": 825
    },
    {
      "epoch": 0.17223490350695164,
      "grad_norm": 2.606112734187681,
      "learning_rate": 9.841977685330074e-06,
      "loss": 0.7544,
      "step": 830
    },
    {
      "epoch": 0.17327246316663208,
      "grad_norm": 2.7643282324128324,
      "learning_rate": 9.837429137287164e-06,
      "loss": 0.7233,
      "step": 835
    },
    {
      "epoch": 0.1743100228263125,
      "grad_norm": 2.58906230921786,
      "learning_rate": 9.832817132899622e-06,
      "loss": 0.7496,
      "step": 840
    },
    {
      "epoch": 0.17534758248599294,
      "grad_norm": 2.9173083959154913,
      "learning_rate": 9.828141732666722e-06,
      "loss": 0.7405,
      "step": 845
    },
    {
      "epoch": 0.17638514214567338,
      "grad_norm": 2.807434396354819,
      "learning_rate": 9.823402997919346e-06,
      "loss": 0.7032,
      "step": 850
    },
    {
      "epoch": 0.1774227018053538,
      "grad_norm": 2.7235521219589343,
      "learning_rate": 9.818600990819193e-06,
      "loss": 0.7162,
      "step": 855
    },
    {
      "epoch": 0.17846026146503424,
      "grad_norm": 3.0405865223857247,
      "learning_rate": 9.813735774357942e-06,
      "loss": 0.7286,
      "step": 860
    },
    {
      "epoch": 0.17949782112471468,
      "grad_norm": 2.6580664453159786,
      "learning_rate": 9.80880741235645e-06,
      "loss": 0.7153,
      "step": 865
    },
    {
      "epoch": 0.1805353807843951,
      "grad_norm": 2.6234079932920764,
      "learning_rate": 9.803815969463898e-06,
      "loss": 0.7267,
      "step": 870
    },
    {
      "epoch": 0.18157294044407554,
      "grad_norm": 2.7632489453595457,
      "learning_rate": 9.798761511156948e-06,
      "loss": 0.7198,
      "step": 875
    },
    {
      "epoch": 0.18261050010375596,
      "grad_norm": 2.607364882820461,
      "learning_rate": 9.79364410373889e-06,
      "loss": 0.7197,
      "step": 880
    },
    {
      "epoch": 0.1836480597634364,
      "grad_norm": 2.7165078647946856,
      "learning_rate": 9.78846381433876e-06,
      "loss": 0.7311,
      "step": 885
    },
    {
      "epoch": 0.18468561942311684,
      "grad_norm": 2.510454703638443,
      "learning_rate": 9.783220710910471e-06,
      "loss": 0.7318,
      "step": 890
    },
    {
      "epoch": 0.18572317908279726,
      "grad_norm": 2.638416867941834,
      "learning_rate": 9.777914862231912e-06,
      "loss": 0.73,
      "step": 895
    },
    {
      "epoch": 0.1867607387424777,
      "grad_norm": 2.7242863471244148,
      "learning_rate": 9.772546337904054e-06,
      "loss": 0.7191,
      "step": 900
    },
    {
      "epoch": 0.18779829840215811,
      "grad_norm": 2.6641307804734806,
      "learning_rate": 9.767115208350035e-06,
      "loss": 0.7207,
      "step": 905
    },
    {
      "epoch": 0.18883585806183856,
      "grad_norm": 2.4803694238796905,
      "learning_rate": 9.761621544814232e-06,
      "loss": 0.7366,
      "step": 910
    },
    {
      "epoch": 0.189873417721519,
      "grad_norm": 2.54201756384705,
      "learning_rate": 9.756065419361329e-06,
      "loss": 0.6971,
      "step": 915
    },
    {
      "epoch": 0.19091097738119941,
      "grad_norm": 2.663357207599627,
      "learning_rate": 9.750446904875374e-06,
      "loss": 0.7093,
      "step": 920
    },
    {
      "epoch": 0.19194853704087986,
      "grad_norm": 2.665290508123204,
      "learning_rate": 9.744766075058817e-06,
      "loss": 0.7092,
      "step": 925
    },
    {
      "epoch": 0.19298609670056027,
      "grad_norm": 2.561102249391226,
      "learning_rate": 9.739023004431553e-06,
      "loss": 0.7022,
      "step": 930
    },
    {
      "epoch": 0.19402365636024071,
      "grad_norm": 2.8532699081444344,
      "learning_rate": 9.733217768329934e-06,
      "loss": 0.7125,
      "step": 935
    },
    {
      "epoch": 0.19506121601992116,
      "grad_norm": 2.658165458053829,
      "learning_rate": 9.727350442905786e-06,
      "loss": 0.713,
      "step": 940
    },
    {
      "epoch": 0.19609877567960157,
      "grad_norm": 2.8346689183428233,
      "learning_rate": 9.721421105125409e-06,
      "loss": 0.7111,
      "step": 945
    },
    {
      "epoch": 0.19713633533928202,
      "grad_norm": 2.691363256166988,
      "learning_rate": 9.715429832768566e-06,
      "loss": 0.6997,
      "step": 950
    },
    {
      "epoch": 0.19817389499896243,
      "grad_norm": 2.670459556620767,
      "learning_rate": 9.709376704427471e-06,
      "loss": 0.7002,
      "step": 955
    },
    {
      "epoch": 0.19921145465864287,
      "grad_norm": 2.8298140051123624,
      "learning_rate": 9.703261799505743e-06,
      "loss": 0.6919,
      "step": 960
    },
    {
      "epoch": 0.20024901431832332,
      "grad_norm": 2.690706447862383,
      "learning_rate": 9.697085198217378e-06,
      "loss": 0.6951,
      "step": 965
    },
    {
      "epoch": 0.20128657397800373,
      "grad_norm": 2.6190530004747536,
      "learning_rate": 9.690846981585689e-06,
      "loss": 0.7088,
      "step": 970
    },
    {
      "epoch": 0.20232413363768417,
      "grad_norm": 2.896807508207353,
      "learning_rate": 9.684547231442248e-06,
      "loss": 0.7036,
      "step": 975
    },
    {
      "epoch": 0.2033616932973646,
      "grad_norm": 2.653583975726927,
      "learning_rate": 9.678186030425806e-06,
      "loss": 0.7014,
      "step": 980
    },
    {
      "epoch": 0.20439925295704503,
      "grad_norm": 2.7468040879725857,
      "learning_rate": 9.67176346198122e-06,
      "loss": 0.6887,
      "step": 985
    },
    {
      "epoch": 0.20543681261672547,
      "grad_norm": 2.6425850355422607,
      "learning_rate": 9.665279610358347e-06,
      "loss": 0.6912,
      "step": 990
    },
    {
      "epoch": 0.2064743722764059,
      "grad_norm": 2.607466586766701,
      "learning_rate": 9.658734560610942e-06,
      "loss": 0.6986,
      "step": 995
    },
    {
      "epoch": 0.20751193193608633,
      "grad_norm": 2.5797036833444684,
      "learning_rate": 9.652128398595548e-06,
      "loss": 0.6893,
      "step": 1000
    },
    {
      "epoch": 0.20854949159576675,
      "grad_norm": 2.627466198960634,
      "learning_rate": 9.645461210970363e-06,
      "loss": 0.6939,
      "step": 1005
    },
    {
      "epoch": 0.2095870512554472,
      "grad_norm": 2.7514054708664486,
      "learning_rate": 9.638733085194105e-06,
      "loss": 0.6879,
      "step": 1010
    },
    {
      "epoch": 0.21062461091512763,
      "grad_norm": 2.762232274342595,
      "learning_rate": 9.631944109524867e-06,
      "loss": 0.7206,
      "step": 1015
    },
    {
      "epoch": 0.21166217057480805,
      "grad_norm": 2.573942783710098,
      "learning_rate": 9.625094373018957e-06,
      "loss": 0.672,
      "step": 1020
    },
    {
      "epoch": 0.2126997302344885,
      "grad_norm": 2.9862062253264154,
      "learning_rate": 9.61818396552973e-06,
      "loss": 0.7027,
      "step": 1025
    },
    {
      "epoch": 0.2137372898941689,
      "grad_norm": 2.604158278569754,
      "learning_rate": 9.61121297770641e-06,
      "loss": 0.6832,
      "step": 1030
    },
    {
      "epoch": 0.21477484955384935,
      "grad_norm": 2.847316696943195,
      "learning_rate": 9.604181500992904e-06,
      "loss": 0.6799,
      "step": 1035
    },
    {
      "epoch": 0.2158124092135298,
      "grad_norm": 2.779271975810847,
      "learning_rate": 9.597089627626594e-06,
      "loss": 0.6804,
      "step": 1040
    },
    {
      "epoch": 0.2168499688732102,
      "grad_norm": 2.5638323571099546,
      "learning_rate": 9.589937450637134e-06,
      "loss": 0.6837,
      "step": 1045
    },
    {
      "epoch": 0.21788752853289065,
      "grad_norm": 2.7020801988998384,
      "learning_rate": 9.58272506384523e-06,
      "loss": 0.684,
      "step": 1050
    },
    {
      "epoch": 0.21892508819257106,
      "grad_norm": 2.549003888633134,
      "learning_rate": 9.5754525618614e-06,
      "loss": 0.6871,
      "step": 1055
    },
    {
      "epoch": 0.2199626478522515,
      "grad_norm": 2.8154214183624284,
      "learning_rate": 9.568120040084752e-06,
      "loss": 0.6652,
      "step": 1060
    },
    {
      "epoch": 0.22100020751193195,
      "grad_norm": 2.6474463880567884,
      "learning_rate": 9.56072759470171e-06,
      "loss": 0.6896,
      "step": 1065
    },
    {
      "epoch": 0.22203776717161236,
      "grad_norm": 2.623449606437885,
      "learning_rate": 9.553275322684769e-06,
      "loss": 0.6731,
      "step": 1070
    },
    {
      "epoch": 0.2230753268312928,
      "grad_norm": 2.5852337807326324,
      "learning_rate": 9.545763321791213e-06,
      "loss": 0.6914,
      "step": 1075
    },
    {
      "epoch": 0.22411288649097322,
      "grad_norm": 2.5768442254313424,
      "learning_rate": 9.538191690561838e-06,
      "loss": 0.6827,
      "step": 1080
    },
    {
      "epoch": 0.22515044615065366,
      "grad_norm": 2.616849381186977,
      "learning_rate": 9.530560528319657e-06,
      "loss": 0.6861,
      "step": 1085
    },
    {
      "epoch": 0.2261880058103341,
      "grad_norm": 2.542283706279765,
      "learning_rate": 9.522869935168601e-06,
      "loss": 0.6673,
      "step": 1090
    },
    {
      "epoch": 0.22722556547001452,
      "grad_norm": 2.6428940149632916,
      "learning_rate": 9.515120011992199e-06,
      "loss": 0.6595,
      "step": 1095
    },
    {
      "epoch": 0.22826312512969496,
      "grad_norm": 2.6516305122608324,
      "learning_rate": 9.507310860452258e-06,
      "loss": 0.6508,
      "step": 1100
    },
    {
      "epoch": 0.22930068478937538,
      "grad_norm": 2.618199521456939,
      "learning_rate": 9.499442582987535e-06,
      "loss": 0.672,
      "step": 1105
    },
    {
      "epoch": 0.23033824444905582,
      "grad_norm": 2.580520201322647,
      "learning_rate": 9.491515282812383e-06,
      "loss": 0.6798,
      "step": 1110
    },
    {
      "epoch": 0.23137580410873626,
      "grad_norm": 2.5578402107468285,
      "learning_rate": 9.483529063915405e-06,
      "loss": 0.6575,
      "step": 1115
    },
    {
      "epoch": 0.23241336376841668,
      "grad_norm": 2.5898669244363743,
      "learning_rate": 9.475484031058081e-06,
      "loss": 0.6686,
      "step": 1120
    },
    {
      "epoch": 0.23345092342809712,
      "grad_norm": 2.581092169980836,
      "learning_rate": 9.46738028977341e-06,
      "loss": 0.676,
      "step": 1125
    },
    {
      "epoch": 0.23448848308777756,
      "grad_norm": 2.564857684910852,
      "learning_rate": 9.459217946364508e-06,
      "loss": 0.6603,
      "step": 1130
    },
    {
      "epoch": 0.23552604274745798,
      "grad_norm": 2.6352764131909394,
      "learning_rate": 9.450997107903222e-06,
      "loss": 0.673,
      "step": 1135
    },
    {
      "epoch": 0.23656360240713842,
      "grad_norm": 2.7623386766168494,
      "learning_rate": 9.442717882228727e-06,
      "loss": 0.6713,
      "step": 1140
    },
    {
      "epoch": 0.23760116206681883,
      "grad_norm": 2.646657632225148,
      "learning_rate": 9.434380377946104e-06,
      "loss": 0.6714,
      "step": 1145
    },
    {
      "epoch": 0.23863872172649928,
      "grad_norm": 2.547166874529112,
      "learning_rate": 9.425984704424927e-06,
      "loss": 0.6664,
      "step": 1150
    },
    {
      "epoch": 0.23967628138617972,
      "grad_norm": 2.5806952200679225,
      "learning_rate": 9.417530971797812e-06,
      "loss": 0.6733,
      "step": 1155
    },
    {
      "epoch": 0.24071384104586013,
      "grad_norm": 2.574640603606341,
      "learning_rate": 9.409019290958993e-06,
      "loss": 0.6737,
      "step": 1160
    },
    {
      "epoch": 0.24175140070554058,
      "grad_norm": 2.5060886593359113,
      "learning_rate": 9.400449773562849e-06,
      "loss": 0.6762,
      "step": 1165
    },
    {
      "epoch": 0.242788960365221,
      "grad_norm": 2.716564112603813,
      "learning_rate": 9.391822532022445e-06,
      "loss": 0.6551,
      "step": 1170
    },
    {
      "epoch": 0.24382652002490143,
      "grad_norm": 2.7171895315845216,
      "learning_rate": 9.383137679508063e-06,
      "loss": 0.6561,
      "step": 1175
    },
    {
      "epoch": 0.24486407968458188,
      "grad_norm": 2.58866525923876,
      "learning_rate": 9.374395329945714e-06,
      "loss": 0.6586,
      "step": 1180
    },
    {
      "epoch": 0.2459016393442623,
      "grad_norm": 2.641911339118785,
      "learning_rate": 9.365595598015635e-06,
      "loss": 0.6879,
      "step": 1185
    },
    {
      "epoch": 0.24693919900394273,
      "grad_norm": 2.6209891495095716,
      "learning_rate": 9.356738599150805e-06,
      "loss": 0.6562,
      "step": 1190
    },
    {
      "epoch": 0.24797675866362315,
      "grad_norm": 2.537931458295151,
      "learning_rate": 9.347824449535406e-06,
      "loss": 0.671,
      "step": 1195
    },
    {
      "epoch": 0.2490143183233036,
      "grad_norm": 2.6373525596001777,
      "learning_rate": 9.338853266103318e-06,
      "loss": 0.6469,
      "step": 1200
    },
    {
      "epoch": 0.250051877982984,
      "grad_norm": 2.49609290349947,
      "learning_rate": 9.329825166536578e-06,
      "loss": 0.6494,
      "step": 1205
    },
    {
      "epoch": 0.25108943764266445,
      "grad_norm": 2.597719369713314,
      "learning_rate": 9.32074026926383e-06,
      "loss": 0.6644,
      "step": 1210
    },
    {
      "epoch": 0.2521269973023449,
      "grad_norm": 2.622902413291871,
      "learning_rate": 9.31159869345879e-06,
      "loss": 0.6277,
      "step": 1215
    },
    {
      "epoch": 0.25316455696202533,
      "grad_norm": 2.4773937545219615,
      "learning_rate": 9.302400559038658e-06,
      "loss": 0.6435,
      "step": 1220
    },
    {
      "epoch": 0.2542021166217057,
      "grad_norm": 2.417435834815959,
      "learning_rate": 9.293145986662567e-06,
      "loss": 0.6551,
      "step": 1225
    },
    {
      "epoch": 0.25523967628138616,
      "grad_norm": 2.7620863467259134,
      "learning_rate": 9.283835097729984e-06,
      "loss": 0.6524,
      "step": 1230
    },
    {
      "epoch": 0.2562772359410666,
      "grad_norm": 2.661471370041919,
      "learning_rate": 9.27446801437913e-06,
      "loss": 0.6564,
      "step": 1235
    },
    {
      "epoch": 0.25731479560074705,
      "grad_norm": 2.6022353359733996,
      "learning_rate": 9.265044859485369e-06,
      "loss": 0.6504,
      "step": 1240
    },
    {
      "epoch": 0.2583523552604275,
      "grad_norm": 2.67544454301597,
      "learning_rate": 9.2555657566596e-06,
      "loss": 0.6379,
      "step": 1245
    },
    {
      "epoch": 0.25938991492010793,
      "grad_norm": 2.7294894592664742,
      "learning_rate": 9.246030830246633e-06,
      "loss": 0.653,
      "step": 1250
    },
    {
      "epoch": 0.2604274745797883,
      "grad_norm": 2.4934502710929642,
      "learning_rate": 9.236440205323564e-06,
      "loss": 0.6504,
      "step": 1255
    },
    {
      "epoch": 0.26146503423946876,
      "grad_norm": 2.732333826367612,
      "learning_rate": 9.226794007698128e-06,
      "loss": 0.6417,
      "step": 1260
    },
    {
      "epoch": 0.2625025938991492,
      "grad_norm": 2.6288344277507023,
      "learning_rate": 9.217092363907047e-06,
      "loss": 0.6193,
      "step": 1265
    },
    {
      "epoch": 0.26354015355882965,
      "grad_norm": 2.7169648920990483,
      "learning_rate": 9.207335401214379e-06,
      "loss": 0.6536,
      "step": 1270
    },
    {
      "epoch": 0.2645777132185101,
      "grad_norm": 2.520924205856965,
      "learning_rate": 9.197523247609839e-06,
      "loss": 0.6375,
      "step": 1275
    },
    {
      "epoch": 0.2656152728781905,
      "grad_norm": 2.515144443600501,
      "learning_rate": 9.187656031807129e-06,
      "loss": 0.6442,
      "step": 1280
    },
    {
      "epoch": 0.2666528325378709,
      "grad_norm": 2.5704491825236473,
      "learning_rate": 9.177733883242244e-06,
      "loss": 0.6586,
      "step": 1285
    },
    {
      "epoch": 0.26769039219755136,
      "grad_norm": 2.8282989819254634,
      "learning_rate": 9.167756932071769e-06,
      "loss": 0.6609,
      "step": 1290
    },
    {
      "epoch": 0.2687279518572318,
      "grad_norm": 2.485389299186558,
      "learning_rate": 9.157725309171183e-06,
      "loss": 0.6459,
      "step": 1295
    },
    {
      "epoch": 0.26976551151691225,
      "grad_norm": 2.7788274917257327,
      "learning_rate": 9.147639146133142e-06,
      "loss": 0.6433,
      "step": 1300
    },
    {
      "epoch": 0.27080307117659264,
      "grad_norm": 2.521611630972388,
      "learning_rate": 9.137498575265736e-06,
      "loss": 0.6271,
      "step": 1305
    },
    {
      "epoch": 0.2718406308362731,
      "grad_norm": 2.67661014478125,
      "learning_rate": 9.12730372959077e-06,
      "loss": 0.6551,
      "step": 1310
    },
    {
      "epoch": 0.2728781904959535,
      "grad_norm": 2.678821798528819,
      "learning_rate": 9.11705474284202e-06,
      "loss": 0.6275,
      "step": 1315
    },
    {
      "epoch": 0.27391575015563396,
      "grad_norm": 2.5914261808444574,
      "learning_rate": 9.106751749463463e-06,
      "loss": 0.6401,
      "step": 1320
    },
    {
      "epoch": 0.2749533098153144,
      "grad_norm": 2.5454105011810664,
      "learning_rate": 9.09639488460753e-06,
      "loss": 0.6275,
      "step": 1325
    },
    {
      "epoch": 0.2759908694749948,
      "grad_norm": 2.746226513585246,
      "learning_rate": 9.08598428413333e-06,
      "loss": 0.6199,
      "step": 1330
    },
    {
      "epoch": 0.27702842913467524,
      "grad_norm": 2.4978350247576886,
      "learning_rate": 9.075520084604849e-06,
      "loss": 0.6081,
      "step": 1335
    },
    {
      "epoch": 0.2780659887943557,
      "grad_norm": 2.565415348716322,
      "learning_rate": 9.065002423289189e-06,
      "loss": 0.6117,
      "step": 1340
    },
    {
      "epoch": 0.2791035484540361,
      "grad_norm": 2.5742091961939892,
      "learning_rate": 9.054431438154745e-06,
      "loss": 0.613,
      "step": 1345
    },
    {
      "epoch": 0.28014110811371656,
      "grad_norm": 2.5626823209269642,
      "learning_rate": 9.043807267869403e-06,
      "loss": 0.624,
      "step": 1350
    },
    {
      "epoch": 0.28117866777339695,
      "grad_norm": 2.4910808560943942,
      "learning_rate": 9.033130051798725e-06,
      "loss": 0.6314,
      "step": 1355
    },
    {
      "epoch": 0.2822162274330774,
      "grad_norm": 2.4573499982125524,
      "learning_rate": 9.022399930004106e-06,
      "loss": 0.625,
      "step": 1360
    },
    {
      "epoch": 0.28325378709275784,
      "grad_norm": 2.531960067205625,
      "learning_rate": 9.011617043240956e-06,
      "loss": 0.6261,
      "step": 1365
    },
    {
      "epoch": 0.2842913467524383,
      "grad_norm": 2.6336708253712917,
      "learning_rate": 9.000781532956844e-06,
      "loss": 0.6057,
      "step": 1370
    },
    {
      "epoch": 0.2853289064121187,
      "grad_norm": 2.5416291249522063,
      "learning_rate": 8.989893541289636e-06,
      "loss": 0.6114,
      "step": 1375
    },
    {
      "epoch": 0.2863664660717991,
      "grad_norm": 2.6016836625846227,
      "learning_rate": 8.978953211065645e-06,
      "loss": 0.6308,
      "step": 1380
    },
    {
      "epoch": 0.28740402573147955,
      "grad_norm": 2.670421258864357,
      "learning_rate": 8.96796068579774e-06,
      "loss": 0.6373,
      "step": 1385
    },
    {
      "epoch": 0.28844158539116,
      "grad_norm": 2.6256018056635195,
      "learning_rate": 8.956916109683488e-06,
      "loss": 0.6136,
      "step": 1390
    },
    {
      "epoch": 0.28947914505084044,
      "grad_norm": 2.665986383822575,
      "learning_rate": 8.945819627603235e-06,
      "loss": 0.6294,
      "step": 1395
    },
    {
      "epoch": 0.2905167047105209,
      "grad_norm": 2.620184006600422,
      "learning_rate": 8.934671385118224e-06,
      "loss": 0.6154,
      "step": 1400
    },
    {
      "epoch": 0.29155426437020127,
      "grad_norm": 2.5773178200606797,
      "learning_rate": 8.923471528468675e-06,
      "loss": 0.6263,
      "step": 1405
    },
    {
      "epoch": 0.2925918240298817,
      "grad_norm": 2.528035051025028,
      "learning_rate": 8.912220204571878e-06,
      "loss": 0.6139,
      "step": 1410
    },
    {
      "epoch": 0.29362938368956215,
      "grad_norm": 2.6830704473334053,
      "learning_rate": 8.900917561020255e-06,
      "loss": 0.6256,
      "step": 1415
    },
    {
      "epoch": 0.2946669433492426,
      "grad_norm": 2.5701558374316256,
      "learning_rate": 8.889563746079428e-06,
      "loss": 0.6163,
      "step": 1420
    },
    {
      "epoch": 0.29570450300892304,
      "grad_norm": 2.542727546558585,
      "learning_rate": 8.878158908686276e-06,
      "loss": 0.6214,
      "step": 1425
    },
    {
      "epoch": 0.2967420626686034,
      "grad_norm": 2.6289522750495684,
      "learning_rate": 8.86670319844698e-06,
      "loss": 0.6184,
      "step": 1430
    },
    {
      "epoch": 0.29777962232828387,
      "grad_norm": 2.605979434187302,
      "learning_rate": 8.855196765635055e-06,
      "loss": 0.6148,
      "step": 1435
    },
    {
      "epoch": 0.2988171819879643,
      "grad_norm": 2.5378051705763136,
      "learning_rate": 8.843639761189392e-06,
      "loss": 0.6309,
      "step": 1440
    },
    {
      "epoch": 0.29985474164764475,
      "grad_norm": 3.122542869019648,
      "learning_rate": 8.83203233671226e-06,
      "loss": 0.6148,
      "step": 1445
    },
    {
      "epoch": 0.3008923013073252,
      "grad_norm": 2.7847080279432355,
      "learning_rate": 8.820374644467334e-06,
      "loss": 0.6149,
      "step": 1450
    },
    {
      "epoch": 0.3019298609670056,
      "grad_norm": 2.7571745781556176,
      "learning_rate": 8.808666837377688e-06,
      "loss": 0.6043,
      "step": 1455
    },
    {
      "epoch": 0.302967420626686,
      "grad_norm": 2.6729505157429645,
      "learning_rate": 8.796909069023793e-06,
      "loss": 0.6091,
      "step": 1460
    },
    {
      "epoch": 0.30400498028636647,
      "grad_norm": 2.5670622136089123,
      "learning_rate": 8.7851014936415e-06,
      "loss": 0.5973,
      "step": 1465
    },
    {
      "epoch": 0.3050425399460469,
      "grad_norm": 2.5809462309222586,
      "learning_rate": 8.77324426612002e-06,
      "loss": 0.601,
      "step": 1470
    },
    {
      "epoch": 0.30608009960572735,
      "grad_norm": 2.6928125038468695,
      "learning_rate": 8.761337541999884e-06,
      "loss": 0.603,
      "step": 1475
    },
    {
      "epoch": 0.30711765926540774,
      "grad_norm": 2.5859004844354887,
      "learning_rate": 8.749381477470915e-06,
      "loss": 0.5902,
      "step": 1480
    },
    {
      "epoch": 0.3081552189250882,
      "grad_norm": 2.706857250295533,
      "learning_rate": 8.73737622937017e-06,
      "loss": 0.6068,
      "step": 1485
    },
    {
      "epoch": 0.3091927785847686,
      "grad_norm": 2.5825402406347995,
      "learning_rate": 8.725321955179886e-06,
      "loss": 0.5943,
      "step": 1490
    },
    {
      "epoch": 0.31023033824444907,
      "grad_norm": 2.564859545646307,
      "learning_rate": 8.713218813025412e-06,
      "loss": 0.6166,
      "step": 1495
    },
    {
      "epoch": 0.3112678979041295,
      "grad_norm": 2.4788537811048514,
      "learning_rate": 8.70106696167314e-06,
      "loss": 0.6107,
      "step": 1500
    },
    {
      "epoch": 0.3123054575638099,
      "grad_norm": 2.5086302932261857,
      "learning_rate": 8.688866560528414e-06,
      "loss": 0.5953,
      "step": 1505
    },
    {
      "epoch": 0.31334301722349034,
      "grad_norm": 2.641157195225993,
      "learning_rate": 8.676617769633449e-06,
      "loss": 0.5942,
      "step": 1510
    },
    {
      "epoch": 0.3143805768831708,
      "grad_norm": 2.5041601695953952,
      "learning_rate": 8.66432074966522e-06,
      "loss": 0.614,
      "step": 1515
    },
    {
      "epoch": 0.3154181365428512,
      "grad_norm": 2.6463153946217988,
      "learning_rate": 8.651975661933368e-06,
      "loss": 0.6046,
      "step": 1520
    },
    {
      "epoch": 0.31645569620253167,
      "grad_norm": 2.5329735196813186,
      "learning_rate": 8.639582668378068e-06,
      "loss": 0.5939,
      "step": 1525
    },
    {
      "epoch": 0.31749325586221205,
      "grad_norm": 2.626444758743183,
      "learning_rate": 8.627141931567918e-06,
      "loss": 0.5955,
      "step": 1530
    },
    {
      "epoch": 0.3185308155218925,
      "grad_norm": 2.4901380695636672,
      "learning_rate": 8.614653614697804e-06,
      "loss": 0.5887,
      "step": 1535
    },
    {
      "epoch": 0.31956837518157294,
      "grad_norm": 2.562038119980849,
      "learning_rate": 8.602117881586748e-06,
      "loss": 0.5887,
      "step": 1540
    },
    {
      "epoch": 0.3206059348412534,
      "grad_norm": 2.705668712762943,
      "learning_rate": 8.589534896675782e-06,
      "loss": 0.6155,
      "step": 1545
    },
    {
      "epoch": 0.3216434945009338,
      "grad_norm": 2.4700710695591144,
      "learning_rate": 8.576904825025763e-06,
      "loss": 0.5805,
      "step": 1550
    },
    {
      "epoch": 0.3226810541606142,
      "grad_norm": 2.569997778071618,
      "learning_rate": 8.56422783231523e-06,
      "loss": 0.5861,
      "step": 1555
    },
    {
      "epoch": 0.32371861382029465,
      "grad_norm": 2.6800463634463454,
      "learning_rate": 8.551504084838217e-06,
      "loss": 0.5888,
      "step": 1560
    },
    {
      "epoch": 0.3247561734799751,
      "grad_norm": 2.614913869620652,
      "learning_rate": 8.538733749502084e-06,
      "loss": 0.5916,
      "step": 1565
    },
    {
      "epoch": 0.32579373313965554,
      "grad_norm": 2.5392913094762015,
      "learning_rate": 8.525916993825312e-06,
      "loss": 0.5845,
      "step": 1570
    },
    {
      "epoch": 0.326831292799336,
      "grad_norm": 2.5970061681752714,
      "learning_rate": 8.51305398593532e-06,
      "loss": 0.5885,
      "step": 1575
    },
    {
      "epoch": 0.32786885245901637,
      "grad_norm": 2.892436571898512,
      "learning_rate": 8.50014489456625e-06,
      "loss": 0.5735,
      "step": 1580
    },
    {
      "epoch": 0.3289064121186968,
      "grad_norm": 2.6138781972835874,
      "learning_rate": 8.487189889056758e-06,
      "loss": 0.5559,
      "step": 1585
    },
    {
      "epoch": 0.32994397177837725,
      "grad_norm": 2.623576409499823,
      "learning_rate": 8.474189139347795e-06,
      "loss": 0.5846,
      "step": 1590
    },
    {
      "epoch": 0.3309815314380577,
      "grad_norm": 2.468823382349316,
      "learning_rate": 8.461142815980368e-06,
      "loss": 0.5986,
      "step": 1595
    },
    {
      "epoch": 0.33201909109773814,
      "grad_norm": 2.672237640145444,
      "learning_rate": 8.448051090093315e-06,
      "loss": 0.591,
      "step": 1600
    },
    {
      "epoch": 0.3330566507574185,
      "grad_norm": 2.4728672068736937,
      "learning_rate": 8.434914133421053e-06,
      "loss": 0.5845,
      "step": 1605
    },
    {
      "epoch": 0.33409421041709897,
      "grad_norm": 2.5895072383457087,
      "learning_rate": 8.421732118291326e-06,
      "loss": 0.5782,
      "step": 1610
    },
    {
      "epoch": 0.3351317700767794,
      "grad_norm": 2.5040782851199137,
      "learning_rate": 8.408505217622942e-06,
      "loss": 0.5815,
      "step": 1615
    },
    {
      "epoch": 0.33616932973645985,
      "grad_norm": 3.6988626884139846,
      "learning_rate": 8.395233604923515e-06,
      "loss": 0.5843,
      "step": 1620
    },
    {
      "epoch": 0.3372068893961403,
      "grad_norm": 2.5828684207028547,
      "learning_rate": 8.381917454287175e-06,
      "loss": 0.5793,
      "step": 1625
    },
    {
      "epoch": 0.3382444490558207,
      "grad_norm": 2.638686184144453,
      "learning_rate": 8.368556940392295e-06,
      "loss": 0.5841,
      "step": 1630
    },
    {
      "epoch": 0.3392820087155011,
      "grad_norm": 2.650538574655116,
      "learning_rate": 8.355152238499192e-06,
      "loss": 0.5875,
      "step": 1635
    },
    {
      "epoch": 0.34031956837518157,
      "grad_norm": 2.486723068143089,
      "learning_rate": 8.341703524447834e-06,
      "loss": 0.5752,
      "step": 1640
    },
    {
      "epoch": 0.341357128034862,
      "grad_norm": 2.542548425745693,
      "learning_rate": 8.328210974655534e-06,
      "loss": 0.582,
      "step": 1645
    },
    {
      "epoch": 0.34239468769454245,
      "grad_norm": 2.5238474016255346,
      "learning_rate": 8.314674766114625e-06,
      "loss": 0.5886,
      "step": 1650
    },
    {
      "epoch": 0.34343224735422284,
      "grad_norm": 2.6663652283930164,
      "learning_rate": 8.301095076390151e-06,
      "loss": 0.5703,
      "step": 1655
    },
    {
      "epoch": 0.3444698070139033,
      "grad_norm": 2.5916048021671765,
      "learning_rate": 8.287472083617534e-06,
      "loss": 0.5578,
      "step": 1660
    },
    {
      "epoch": 0.3455073666735837,
      "grad_norm": 2.64819187278271,
      "learning_rate": 8.273805966500233e-06,
      "loss": 0.566,
      "step": 1665
    },
    {
      "epoch": 0.34654492633326417,
      "grad_norm": 2.7355426760241115,
      "learning_rate": 8.260096904307404e-06,
      "loss": 0.5724,
      "step": 1670
    },
    {
      "epoch": 0.3475824859929446,
      "grad_norm": 2.5288788248400422,
      "learning_rate": 8.246345076871548e-06,
      "loss": 0.5852,
      "step": 1675
    },
    {
      "epoch": 0.348620045652625,
      "grad_norm": 2.4333668936065727,
      "learning_rate": 8.232550664586145e-06,
      "loss": 0.562,
      "step": 1680
    },
    {
      "epoch": 0.34965760531230544,
      "grad_norm": 2.666095931772747,
      "learning_rate": 8.218713848403306e-06,
      "loss": 0.5761,
      "step": 1685
    },
    {
      "epoch": 0.3506951649719859,
      "grad_norm": 2.585591858891525,
      "learning_rate": 8.204834809831377e-06,
      "loss": 0.579,
      "step": 1690
    },
    {
      "epoch": 0.3517327246316663,
      "grad_norm": 2.5419925763820457,
      "learning_rate": 8.190913730932567e-06,
      "loss": 0.5792,
      "step": 1695
    },
    {
      "epoch": 0.35277028429134677,
      "grad_norm": 2.5691748657197575,
      "learning_rate": 8.176950794320572e-06,
      "loss": 0.5647,
      "step": 1700
    },
    {
      "epoch": 0.35380784395102716,
      "grad_norm": 2.5977820134449603,
      "learning_rate": 8.16294618315816e-06,
      "loss": 0.5708,
      "step": 1705
    },
    {
      "epoch": 0.3548454036107076,
      "grad_norm": 2.423522377595987,
      "learning_rate": 8.148900081154773e-06,
      "loss": 0.5666,
      "step": 1710
    },
    {
      "epoch": 0.35588296327038804,
      "grad_norm": 2.569485264156882,
      "learning_rate": 8.134812672564131e-06,
      "loss": 0.5504,
      "step": 1715
    },
    {
      "epoch": 0.3569205229300685,
      "grad_norm": 2.758198111046606,
      "learning_rate": 8.1206841421818e-06,
      "loss": 0.5691,
      "step": 1720
    },
    {
      "epoch": 0.3579580825897489,
      "grad_norm": 2.5786351112814545,
      "learning_rate": 8.10651467534277e-06,
      "loss": 0.57,
      "step": 1725
    },
    {
      "epoch": 0.35899564224942937,
      "grad_norm": 2.4962078269510015,
      "learning_rate": 8.092304457919028e-06,
      "loss": 0.557,
      "step": 1730
    },
    {
      "epoch": 0.36003320190910976,
      "grad_norm": 2.5552381714765384,
      "learning_rate": 8.078053676317124e-06,
      "loss": 0.5673,
      "step": 1735
    },
    {
      "epoch": 0.3610707615687902,
      "grad_norm": 2.611892754049072,
      "learning_rate": 8.06376251747571e-06,
      "loss": 0.5535,
      "step": 1740
    },
    {
      "epoch": 0.36210832122847064,
      "grad_norm": 2.4981862280677545,
      "learning_rate": 8.049431168863107e-06,
      "loss": 0.5543,
      "step": 1745
    },
    {
      "epoch": 0.3631458808881511,
      "grad_norm": 2.590859008799774,
      "learning_rate": 8.035059818474833e-06,
      "loss": 0.5688,
      "step": 1750
    },
    {
      "epoch": 0.3641834405478315,
      "grad_norm": 2.7049066460444657,
      "learning_rate": 8.02064865483114e-06,
      "loss": 0.5666,
      "step": 1755
    },
    {
      "epoch": 0.3652210002075119,
      "grad_norm": 2.50555658991844,
      "learning_rate": 8.00619786697454e-06,
      "loss": 0.553,
      "step": 1760
    },
    {
      "epoch": 0.36625855986719236,
      "grad_norm": 2.4491820476275805,
      "learning_rate": 7.991707644467335e-06,
      "loss": 0.5635,
      "step": 1765
    },
    {
      "epoch": 0.3672961195268728,
      "grad_norm": 2.560789498381096,
      "learning_rate": 7.97717817738911e-06,
      "loss": 0.5408,
      "step": 1770
    },
    {
      "epoch": 0.36833367918655324,
      "grad_norm": 2.6052621436325416,
      "learning_rate": 7.962609656334262e-06,
      "loss": 0.5488,
      "step": 1775
    },
    {
      "epoch": 0.3693712388462337,
      "grad_norm": 2.5278436185793582,
      "learning_rate": 7.94800227240948e-06,
      "loss": 0.5573,
      "step": 1780
    },
    {
      "epoch": 0.37040879850591407,
      "grad_norm": 2.5168973470980607,
      "learning_rate": 7.933356217231261e-06,
      "loss": 0.5358,
      "step": 1785
    },
    {
      "epoch": 0.3714463581655945,
      "grad_norm": 2.5427476959285475,
      "learning_rate": 7.918671682923371e-06,
      "loss": 0.557,
      "step": 1790
    },
    {
      "epoch": 0.37248391782527496,
      "grad_norm": 2.5597051499240315,
      "learning_rate": 7.90394886211434e-06,
      "loss": 0.5443,
      "step": 1795
    },
    {
      "epoch": 0.3735214774849554,
      "grad_norm": 2.6134854223700033,
      "learning_rate": 7.889187947934939e-06,
      "loss": 0.5643,
      "step": 1800
    },
    {
      "epoch": 0.37455903714463584,
      "grad_norm": 2.5638490649548507,
      "learning_rate": 7.874389134015627e-06,
      "loss": 0.5515,
      "step": 1805
    },
    {
      "epoch": 0.37559659680431623,
      "grad_norm": 2.638245607620569,
      "learning_rate": 7.859552614484035e-06,
      "loss": 0.5512,
      "step": 1810
    },
    {
      "epoch": 0.37663415646399667,
      "grad_norm": 2.5439284853350683,
      "learning_rate": 7.844678583962403e-06,
      "loss": 0.5357,
      "step": 1815
    },
    {
      "epoch": 0.3776717161236771,
      "grad_norm": 2.7382120697501264,
      "learning_rate": 7.829767237565027e-06,
      "loss": 0.5499,
      "step": 1820
    },
    {
      "epoch": 0.37870927578335756,
      "grad_norm": 2.693790343217592,
      "learning_rate": 7.814818770895718e-06,
      "loss": 0.5447,
      "step": 1825
    },
    {
      "epoch": 0.379746835443038,
      "grad_norm": 2.5002163516357685,
      "learning_rate": 7.79983338004521e-06,
      "loss": 0.5548,
      "step": 1830
    },
    {
      "epoch": 0.3807843951027184,
      "grad_norm": 2.5140469437451514,
      "learning_rate": 7.784811261588605e-06,
      "loss": 0.5396,
      "step": 1835
    },
    {
      "epoch": 0.38182195476239883,
      "grad_norm": 2.3632709986503384,
      "learning_rate": 7.769752612582793e-06,
      "loss": 0.5455,
      "step": 1840
    },
    {
      "epoch": 0.3828595144220793,
      "grad_norm": 2.6174115493874415,
      "learning_rate": 7.754657630563855e-06,
      "loss": 0.5501,
      "step": 1845
    },
    {
      "epoch": 0.3838970740817597,
      "grad_norm": 2.662869705594198,
      "learning_rate": 7.739526513544492e-06,
      "loss": 0.5458,
      "step": 1850
    },
    {
      "epoch": 0.38493463374144016,
      "grad_norm": 2.595342018363226,
      "learning_rate": 7.724359460011406e-06,
      "loss": 0.5484,
      "step": 1855
    },
    {
      "epoch": 0.38597219340112054,
      "grad_norm": 2.574988659923898,
      "learning_rate": 7.709156668922715e-06,
      "loss": 0.5465,
      "step": 1860
    },
    {
      "epoch": 0.387009753060801,
      "grad_norm": 2.696253501363756,
      "learning_rate": 7.693918339705327e-06,
      "loss": 0.5416,
      "step": 1865
    },
    {
      "epoch": 0.38804731272048143,
      "grad_norm": 2.571358317228633,
      "learning_rate": 7.678644672252334e-06,
      "loss": 0.5432,
      "step": 1870
    },
    {
      "epoch": 0.3890848723801619,
      "grad_norm": 2.6773383342412007,
      "learning_rate": 7.663335866920389e-06,
      "loss": 0.5435,
      "step": 1875
    },
    {
      "epoch": 0.3901224320398423,
      "grad_norm": 2.605715400204335,
      "learning_rate": 7.647992124527076e-06,
      "loss": 0.5394,
      "step": 1880
    },
    {
      "epoch": 0.3911599916995227,
      "grad_norm": 2.542289996117672,
      "learning_rate": 7.632613646348273e-06,
      "loss": 0.5365,
      "step": 1885
    },
    {
      "epoch": 0.39219755135920314,
      "grad_norm": 2.640322094229756,
      "learning_rate": 7.617200634115516e-06,
      "loss": 0.5473,
      "step": 1890
    },
    {
      "epoch": 0.3932351110188836,
      "grad_norm": 2.5708221170868355,
      "learning_rate": 7.601753290013353e-06,
      "loss": 0.5209,
      "step": 1895
    },
    {
      "epoch": 0.39427267067856403,
      "grad_norm": 2.525915351063816,
      "learning_rate": 7.586271816676687e-06,
      "loss": 0.5288,
      "step": 1900
    },
    {
      "epoch": 0.3953102303382445,
      "grad_norm": 2.5364307727837234,
      "learning_rate": 7.570756417188123e-06,
      "loss": 0.5429,
      "step": 1905
    },
    {
      "epoch": 0.39634778999792486,
      "grad_norm": 2.559190451038623,
      "learning_rate": 7.555207295075303e-06,
      "loss": 0.5128,
      "step": 1910
    },
    {
      "epoch": 0.3973853496576053,
      "grad_norm": 2.7079589585625095,
      "learning_rate": 7.539624654308231e-06,
      "loss": 0.5333,
      "step": 1915
    },
    {
      "epoch": 0.39842290931728574,
      "grad_norm": 2.4387259072975267,
      "learning_rate": 7.5240086992966045e-06,
      "loss": 0.5334,
      "step": 1920
    },
    {
      "epoch": 0.3994604689769662,
      "grad_norm": 2.5588774252471818,
      "learning_rate": 7.508359634887128e-06,
      "loss": 0.5429,
      "step": 1925
    },
    {
      "epoch": 0.40049802863664663,
      "grad_norm": 2.50003757681576,
      "learning_rate": 7.4926776663608305e-06,
      "loss": 0.5353,
      "step": 1930
    },
    {
      "epoch": 0.401535588296327,
      "grad_norm": 2.471663404094866,
      "learning_rate": 7.476962999430368e-06,
      "loss": 0.5373,
      "step": 1935
    },
    {
      "epoch": 0.40257314795600746,
      "grad_norm": 2.4541757324292215,
      "learning_rate": 7.461215840237329e-06,
      "loss": 0.5278,
      "step": 1940
    },
    {
      "epoch": 0.4036107076156879,
      "grad_norm": 2.5133909898638738,
      "learning_rate": 7.4454363953495255e-06,
      "loss": 0.5224,
      "step": 1945
    },
    {
      "epoch": 0.40464826727536835,
      "grad_norm": 2.456759039414412,
      "learning_rate": 7.429624871758289e-06,
      "loss": 0.5274,
      "step": 1950
    },
    {
      "epoch": 0.4056858269350488,
      "grad_norm": 2.6254399143011025,
      "learning_rate": 7.41378147687575e-06,
      "loss": 0.539,
      "step": 1955
    },
    {
      "epoch": 0.4067233865947292,
      "grad_norm": 2.77782256038311,
      "learning_rate": 7.397906418532124e-06,
      "loss": 0.5255,
      "step": 1960
    },
    {
      "epoch": 0.4077609462544096,
      "grad_norm": 2.401784319404263,
      "learning_rate": 7.381999904972974e-06,
      "loss": 0.5373,
      "step": 1965
    },
    {
      "epoch": 0.40879850591409006,
      "grad_norm": 2.4340739184784996,
      "learning_rate": 7.366062144856494e-06,
      "loss": 0.5292,
      "step": 1970
    },
    {
      "epoch": 0.4098360655737705,
      "grad_norm": 2.5139056739660104,
      "learning_rate": 7.350093347250754e-06,
      "loss": 0.524,
      "step": 1975
    },
    {
      "epoch": 0.41087362523345095,
      "grad_norm": 2.513276255127495,
      "learning_rate": 7.334093721630976e-06,
      "loss": 0.5231,
      "step": 1980
    },
    {
      "epoch": 0.41191118489313133,
      "grad_norm": 2.6310824706160636,
      "learning_rate": 7.318063477876775e-06,
      "loss": 0.5233,
      "step": 1985
    },
    {
      "epoch": 0.4129487445528118,
      "grad_norm": 2.491475901603011,
      "learning_rate": 7.302002826269401e-06,
      "loss": 0.5341,
      "step": 1990
    },
    {
      "epoch": 0.4139863042124922,
      "grad_norm": 2.506660610799657,
      "learning_rate": 7.285911977488995e-06,
      "loss": 0.5182,
      "step": 1995
    },
    {
      "epoch": 0.41502386387217266,
      "grad_norm": 2.4794467040525507,
      "learning_rate": 7.269791142611819e-06,
      "loss": 0.5305,
      "step": 2000
    },
    {
      "epoch": 0.4160614235318531,
      "grad_norm": 2.4409844087069197,
      "learning_rate": 7.253640533107482e-06,
      "loss": 0.5097,
      "step": 2005
    },
    {
      "epoch": 0.4170989831915335,
      "grad_norm": 2.5831589194501334,
      "learning_rate": 7.23746036083617e-06,
      "loss": 0.509,
      "step": 2010
    },
    {
      "epoch": 0.41813654285121393,
      "grad_norm": 2.5848519997923134,
      "learning_rate": 7.221250838045866e-06,
      "loss": 0.5212,
      "step": 2015
    },
    {
      "epoch": 0.4191741025108944,
      "grad_norm": 2.4649428551198507,
      "learning_rate": 7.205012177369573e-06,
      "loss": 0.5097,
      "step": 2020
    },
    {
      "epoch": 0.4202116621705748,
      "grad_norm": 2.6470337729349334,
      "learning_rate": 7.188744591822514e-06,
      "loss": 0.5265,
      "step": 2025
    },
    {
      "epoch": 0.42124922183025526,
      "grad_norm": 2.6777002487915427,
      "learning_rate": 7.17244829479934e-06,
      "loss": 0.5132,
      "step": 2030
    },
    {
      "epoch": 0.42228678148993565,
      "grad_norm": 2.628723636023145,
      "learning_rate": 7.156123500071337e-06,
      "loss": 0.5383,
      "step": 2035
    },
    {
      "epoch": 0.4233243411496161,
      "grad_norm": 2.6077956409975322,
      "learning_rate": 7.139770421783616e-06,
      "loss": 0.5143,
      "step": 2040
    },
    {
      "epoch": 0.42436190080929653,
      "grad_norm": 2.5195148593267165,
      "learning_rate": 7.1233892744523055e-06,
      "loss": 0.5292,
      "step": 2045
    },
    {
      "epoch": 0.425399460468977,
      "grad_norm": 2.4869546800490276,
      "learning_rate": 7.1069802729617385e-06,
      "loss": 0.5219,
      "step": 2050
    },
    {
      "epoch": 0.4264370201286574,
      "grad_norm": 2.498457038824186,
      "learning_rate": 7.090543632561632e-06,
      "loss": 0.5227,
      "step": 2055
    },
    {
      "epoch": 0.4274745797883378,
      "grad_norm": 2.5334280776647855,
      "learning_rate": 7.0740795688642635e-06,
      "loss": 0.5174,
      "step": 2060
    },
    {
      "epoch": 0.42851213944801825,
      "grad_norm": 2.4142684408837667,
      "learning_rate": 7.057588297841645e-06,
      "loss": 0.5154,
      "step": 2065
    },
    {
      "epoch": 0.4295496991076987,
      "grad_norm": 2.7421341811351967,
      "learning_rate": 7.041070035822687e-06,
      "loss": 0.4983,
      "step": 2070
    },
    {
      "epoch": 0.43058725876737913,
      "grad_norm": 2.385233500154077,
      "learning_rate": 7.024524999490364e-06,
      "loss": 0.535,
      "step": 2075
    },
    {
      "epoch": 0.4316248184270596,
      "grad_norm": 2.493022029900477,
      "learning_rate": 7.007953405878867e-06,
      "loss": 0.5036,
      "step": 2080
    },
    {
      "epoch": 0.43266237808673996,
      "grad_norm": 2.5007769964543507,
      "learning_rate": 6.991355472370762e-06,
      "loss": 0.5288,
      "step": 2085
    },
    {
      "epoch": 0.4336999377464204,
      "grad_norm": 2.4139386188608842,
      "learning_rate": 6.974731416694135e-06,
      "loss": 0.5142,
      "step": 2090
    },
    {
      "epoch": 0.43473749740610085,
      "grad_norm": 2.5053762871669303,
      "learning_rate": 6.958081456919737e-06,
      "loss": 0.502,
      "step": 2095
    },
    {
      "epoch": 0.4357750570657813,
      "grad_norm": 2.5446406360038267,
      "learning_rate": 6.941405811458126e-06,
      "loss": 0.5079,
      "step": 2100
    },
    {
      "epoch": 0.43681261672546173,
      "grad_norm": 2.5024786886158465,
      "learning_rate": 6.924704699056792e-06,
      "loss": 0.5102,
      "step": 2105
    },
    {
      "epoch": 0.4378501763851421,
      "grad_norm": 2.5991293471527754,
      "learning_rate": 6.907978338797304e-06,
      "loss": 0.5033,
      "step": 2110
    },
    {
      "epoch": 0.43888773604482256,
      "grad_norm": 2.45095417799665,
      "learning_rate": 6.891226950092422e-06,
      "loss": 0.5033,
      "step": 2115
    },
    {
      "epoch": 0.439925295704503,
      "grad_norm": 2.5068078607240762,
      "learning_rate": 6.874450752683223e-06,
      "loss": 0.5131,
      "step": 2120
    },
    {
      "epoch": 0.44096285536418345,
      "grad_norm": 2.615502411671901,
      "learning_rate": 6.85764996663622e-06,
      "loss": 0.514,
      "step": 2125
    },
    {
      "epoch": 0.4420004150238639,
      "grad_norm": 2.483179103106064,
      "learning_rate": 6.840824812340476e-06,
      "loss": 0.482,
      "step": 2130
    },
    {
      "epoch": 0.4430379746835443,
      "grad_norm": 2.615056810287462,
      "learning_rate": 6.82397551050471e-06,
      "loss": 0.4981,
      "step": 2135
    },
    {
      "epoch": 0.4440755343432247,
      "grad_norm": 2.5081191557582962,
      "learning_rate": 6.807102282154406e-06,
      "loss": 0.5038,
      "step": 2140
    },
    {
      "epoch": 0.44511309400290516,
      "grad_norm": 2.4228335632525857,
      "learning_rate": 6.790205348628902e-06,
      "loss": 0.5116,
      "step": 2145
    },
    {
      "epoch": 0.4461506536625856,
      "grad_norm": 2.4865567217795834,
      "learning_rate": 6.773284931578508e-06,
      "loss": 0.4923,
      "step": 2150
    },
    {
      "epoch": 0.44718821332226605,
      "grad_norm": 2.5028391081664707,
      "learning_rate": 6.756341252961575e-06,
      "loss": 0.507,
      "step": 2155
    },
    {
      "epoch": 0.44822577298194644,
      "grad_norm": 2.4437292124023755,
      "learning_rate": 6.739374535041601e-06,
      "loss": 0.5041,
      "step": 2160
    },
    {
      "epoch": 0.4492633326416269,
      "grad_norm": 2.4804781219836705,
      "learning_rate": 6.722385000384305e-06,
      "loss": 0.5071,
      "step": 2165
    },
    {
      "epoch": 0.4503008923013073,
      "grad_norm": 2.613493029619529,
      "learning_rate": 6.705372871854713e-06,
      "loss": 0.5045,
      "step": 2170
    },
    {
      "epoch": 0.45133845196098776,
      "grad_norm": 2.5550452506734613,
      "learning_rate": 6.688338372614232e-06,
      "loss": 0.4954,
      "step": 2175
    },
    {
      "epoch": 0.4523760116206682,
      "grad_norm": 2.5823699134656795,
      "learning_rate": 6.671281726117721e-06,
      "loss": 0.5029,
      "step": 2180
    },
    {
      "epoch": 0.4534135712803486,
      "grad_norm": 2.6550826128938287,
      "learning_rate": 6.654203156110565e-06,
      "loss": 0.4942,
      "step": 2185
    },
    {
      "epoch": 0.45445113094002904,
      "grad_norm": 2.4268474548849976,
      "learning_rate": 6.6371028866257355e-06,
      "loss": 0.5027,
      "step": 2190
    },
    {
      "epoch": 0.4554886905997095,
      "grad_norm": 2.437142297358699,
      "learning_rate": 6.6199811419808525e-06,
      "loss": 0.4949,
      "step": 2195
    },
    {
      "epoch": 0.4565262502593899,
      "grad_norm": 2.494477319925722,
      "learning_rate": 6.602838146775243e-06,
      "loss": 0.4796,
      "step": 2200
    },
    {
      "epoch": 0.45756380991907036,
      "grad_norm": 2.5244330207655223,
      "learning_rate": 6.585674125886996e-06,
      "loss": 0.5066,
      "step": 2205
    },
    {
      "epoch": 0.45860136957875075,
      "grad_norm": 2.6127909654783608,
      "learning_rate": 6.568489304470007e-06,
      "loss": 0.4909,
      "step": 2210
    },
    {
      "epoch": 0.4596389292384312,
      "grad_norm": 2.5733537948603105,
      "learning_rate": 6.551283907951031e-06,
      "loss": 0.4886,
      "step": 2215
    },
    {
      "epoch": 0.46067648889811164,
      "grad_norm": 2.627653050074194,
      "learning_rate": 6.534058162026724e-06,
      "loss": 0.4871,
      "step": 2220
    },
    {
      "epoch": 0.4617140485577921,
      "grad_norm": 2.493707567539853,
      "learning_rate": 6.516812292660675e-06,
      "loss": 0.5115,
      "step": 2225
    },
    {
      "epoch": 0.4627516082174725,
      "grad_norm": 2.576298574074138,
      "learning_rate": 6.499546526080457e-06,
      "loss": 0.4935,
      "step": 2230
    },
    {
      "epoch": 0.46378916787715296,
      "grad_norm": 2.3441840780878227,
      "learning_rate": 6.482261088774642e-06,
      "loss": 0.4918,
      "step": 2235
    },
    {
      "epoch": 0.46482672753683335,
      "grad_norm": 2.5855232225238702,
      "learning_rate": 6.464956207489843e-06,
      "loss": 0.5009,
      "step": 2240
    },
    {
      "epoch": 0.4658642871965138,
      "grad_norm": 2.578617818062702,
      "learning_rate": 6.447632109227735e-06,
      "loss": 0.4931,
      "step": 2245
    },
    {
      "epoch": 0.46690184685619424,
      "grad_norm": 2.6300654942955752,
      "learning_rate": 6.4302890212420735e-06,
      "loss": 0.4924,
      "step": 2250
    },
    {
      "epoch": 0.4679394065158747,
      "grad_norm": 2.535272976711584,
      "learning_rate": 6.412927171035721e-06,
      "loss": 0.4864,
      "step": 2255
    },
    {
      "epoch": 0.4689769661755551,
      "grad_norm": 2.497392960102334,
      "learning_rate": 6.3955467863576555e-06,
      "loss": 0.502,
      "step": 2260
    },
    {
      "epoch": 0.4700145258352355,
      "grad_norm": 2.476885100955309,
      "learning_rate": 6.37814809519999e-06,
      "loss": 0.498,
      "step": 2265
    },
    {
      "epoch": 0.47105208549491595,
      "grad_norm": 2.3650744048382726,
      "learning_rate": 6.360731325794975e-06,
      "loss": 0.486,
      "step": 2270
    },
    {
      "epoch": 0.4720896451545964,
      "grad_norm": 2.469078585494707,
      "learning_rate": 6.343296706612008e-06,
      "loss": 0.4745,
      "step": 2275
    },
    {
      "epoch": 0.47312720481427684,
      "grad_norm": 2.5527304594928326,
      "learning_rate": 6.325844466354637e-06,
      "loss": 0.4959,
      "step": 2280
    },
    {
      "epoch": 0.4741647644739573,
      "grad_norm": 2.388876155874213,
      "learning_rate": 6.308374833957556e-06,
      "loss": 0.4787,
      "step": 2285
    },
    {
      "epoch": 0.47520232413363767,
      "grad_norm": 2.642090924554838,
      "learning_rate": 6.290888038583611e-06,
      "loss": 0.4951,
      "step": 2290
    },
    {
      "epoch": 0.4762398837933181,
      "grad_norm": 2.5493215911731557,
      "learning_rate": 6.273384309620785e-06,
      "loss": 0.4799,
      "step": 2295
    },
    {
      "epoch": 0.47727744345299855,
      "grad_norm": 2.3751142920110944,
      "learning_rate": 6.25586387667919e-06,
      "loss": 0.4841,
      "step": 2300
    },
    {
      "epoch": 0.478315003112679,
      "grad_norm": 2.4695590079808625,
      "learning_rate": 6.238326969588062e-06,
      "loss": 0.4739,
      "step": 2305
    },
    {
      "epoch": 0.47935256277235944,
      "grad_norm": 2.6302561474616937,
      "learning_rate": 6.220773818392738e-06,
      "loss": 0.4809,
      "step": 2310
    },
    {
      "epoch": 0.4803901224320398,
      "grad_norm": 2.5254595904398744,
      "learning_rate": 6.203204653351642e-06,
      "loss": 0.4964,
      "step": 2315
    },
    {
      "epoch": 0.48142768209172027,
      "grad_norm": 2.4902195244274288,
      "learning_rate": 6.185619704933267e-06,
      "loss": 0.4654,
      "step": 2320
    },
    {
      "epoch": 0.4824652417514007,
      "grad_norm": 2.5224684668035007,
      "learning_rate": 6.168019203813143e-06,
      "loss": 0.479,
      "step": 2325
    },
    {
      "epoch": 0.48350280141108115,
      "grad_norm": 2.531197814844583,
      "learning_rate": 6.15040338087082e-06,
      "loss": 0.4756,
      "step": 2330
    },
    {
      "epoch": 0.4845403610707616,
      "grad_norm": 2.3546555565826135,
      "learning_rate": 6.132772467186841e-06,
      "loss": 0.4649,
      "step": 2335
    },
    {
      "epoch": 0.485577920730442,
      "grad_norm": 2.3784120564924622,
      "learning_rate": 6.115126694039699e-06,
      "loss": 0.4709,
      "step": 2340
    },
    {
      "epoch": 0.4866154803901224,
      "grad_norm": 2.56815940386139,
      "learning_rate": 6.097466292902815e-06,
      "loss": 0.486,
      "step": 2345
    },
    {
      "epoch": 0.48765304004980287,
      "grad_norm": 2.5301845486075103,
      "learning_rate": 6.079791495441491e-06,
      "loss": 0.4754,
      "step": 2350
    },
    {
      "epoch": 0.4886905997094833,
      "grad_norm": 2.4650710165826806,
      "learning_rate": 6.062102533509886e-06,
      "loss": 0.4663,
      "step": 2355
    },
    {
      "epoch": 0.48972815936916375,
      "grad_norm": 2.7587452612942145,
      "learning_rate": 6.044399639147957e-06,
      "loss": 0.4632,
      "step": 2360
    },
    {
      "epoch": 0.49076571902884414,
      "grad_norm": 2.543480240365517,
      "learning_rate": 6.026683044578427e-06,
      "loss": 0.4689,
      "step": 2365
    },
    {
      "epoch": 0.4918032786885246,
      "grad_norm": 2.4525809117166366,
      "learning_rate": 6.008952982203737e-06,
      "loss": 0.4843,
      "step": 2370
    },
    {
      "epoch": 0.492840838348205,
      "grad_norm": 2.417071327082838,
      "learning_rate": 5.991209684602991e-06,
      "loss": 0.4677,
      "step": 2375
    },
    {
      "epoch": 0.49387839800788547,
      "grad_norm": 2.8159423032357394,
      "learning_rate": 5.9734533845289144e-06,
      "loss": 0.466,
      "step": 2380
    },
    {
      "epoch": 0.4949159576675659,
      "grad_norm": 2.5331145358141023,
      "learning_rate": 5.955684314904795e-06,
      "loss": 0.491,
      "step": 2385
    },
    {
      "epoch": 0.4959535173272463,
      "grad_norm": 2.5850294725860103,
      "learning_rate": 5.937902708821427e-06,
      "loss": 0.4727,
      "step": 2390
    },
    {
      "epoch": 0.49699107698692674,
      "grad_norm": 2.3455874492137627,
      "learning_rate": 5.920108799534059e-06,
      "loss": 0.4699,
      "step": 2395
    },
    {
      "epoch": 0.4980286366466072,
      "grad_norm": 2.496007013879939,
      "learning_rate": 5.902302820459324e-06,
      "loss": 0.4599,
      "step": 2400
    },
    {
      "epoch": 0.4990661963062876,
      "grad_norm": 2.412699996479322,
      "learning_rate": 5.884485005172189e-06,
      "loss": 0.474,
      "step": 2405
    },
    {
      "epoch": 0.500103755965968,
      "grad_norm": 2.6197335120279805,
      "learning_rate": 5.866655587402886e-06,
      "loss": 0.4815,
      "step": 2410
    },
    {
      "epoch": 0.5011413156256485,
      "grad_norm": 2.632568563481647,
      "learning_rate": 5.8488148010338445e-06,
      "loss": 0.474,
      "step": 2415
    },
    {
      "epoch": 0.5021788752853289,
      "grad_norm": 2.383359491818944,
      "learning_rate": 5.8309628800966225e-06,
      "loss": 0.464,
      "step": 2420
    },
    {
      "epoch": 0.5032164349450093,
      "grad_norm": 2.687191301722998,
      "learning_rate": 5.813100058768841e-06,
      "loss": 0.4671,
      "step": 2425
    },
    {
      "epoch": 0.5042539946046898,
      "grad_norm": 2.455559246352499,
      "learning_rate": 5.795226571371114e-06,
      "loss": 0.4682,
      "step": 2430
    },
    {
      "epoch": 0.5052915542643702,
      "grad_norm": 2.589166158482144,
      "learning_rate": 5.777342652363963e-06,
      "loss": 0.4756,
      "step": 2435
    },
    {
      "epoch": 0.5063291139240507,
      "grad_norm": 2.5173329990266295,
      "learning_rate": 5.759448536344753e-06,
      "loss": 0.4849,
      "step": 2440
    },
    {
      "epoch": 0.5073666735837311,
      "grad_norm": 2.3813334648732227,
      "learning_rate": 5.741544458044611e-06,
      "loss": 0.4725,
      "step": 2445
    },
    {
      "epoch": 0.5084042332434114,
      "grad_norm": 2.4645035637647683,
      "learning_rate": 5.723630652325349e-06,
      "loss": 0.4523,
      "step": 2450
    },
    {
      "epoch": 0.5094417929030919,
      "grad_norm": 2.5330130007551848,
      "learning_rate": 5.705707354176377e-06,
      "loss": 0.4655,
      "step": 2455
    },
    {
      "epoch": 0.5104793525627723,
      "grad_norm": 2.3938015606881886,
      "learning_rate": 5.687774798711627e-06,
      "loss": 0.468,
      "step": 2460
    },
    {
      "epoch": 0.5115169122224528,
      "grad_norm": 2.647072227845224,
      "learning_rate": 5.669833221166469e-06,
      "loss": 0.4695,
      "step": 2465
    },
    {
      "epoch": 0.5125544718821332,
      "grad_norm": 2.4713289117715544,
      "learning_rate": 5.651882856894615e-06,
      "loss": 0.4617,
      "step": 2470
    },
    {
      "epoch": 0.5135920315418137,
      "grad_norm": 2.494015444510411,
      "learning_rate": 5.633923941365049e-06,
      "loss": 0.4659,
      "step": 2475
    },
    {
      "epoch": 0.5146295912014941,
      "grad_norm": 2.3455915787636807,
      "learning_rate": 5.615956710158921e-06,
      "loss": 0.4563,
      "step": 2480
    },
    {
      "epoch": 0.5156671508611745,
      "grad_norm": 2.453795440026443,
      "learning_rate": 5.597981398966468e-06,
      "loss": 0.4698,
      "step": 2485
    },
    {
      "epoch": 0.516704710520855,
      "grad_norm": 2.3862841672431143,
      "learning_rate": 5.579998243583919e-06,
      "loss": 0.4583,
      "step": 2490
    },
    {
      "epoch": 0.5177422701805354,
      "grad_norm": 2.412824706136417,
      "learning_rate": 5.562007479910396e-06,
      "loss": 0.4714,
      "step": 2495
    },
    {
      "epoch": 0.5187798298402159,
      "grad_norm": 2.5121998422157543,
      "learning_rate": 5.544009343944834e-06,
      "loss": 0.4597,
      "step": 2500
    },
    {
      "epoch": 0.5198173894998962,
      "grad_norm": 2.479588579717285,
      "learning_rate": 5.526004071782868e-06,
      "loss": 0.461,
      "step": 2505
    },
    {
      "epoch": 0.5208549491595766,
      "grad_norm": 2.4676776770935365,
      "learning_rate": 5.507991899613746e-06,
      "loss": 0.4632,
      "step": 2510
    },
    {
      "epoch": 0.5218925088192571,
      "grad_norm": 2.4315208171491838,
      "learning_rate": 5.489973063717233e-06,
      "loss": 0.4702,
      "step": 2515
    },
    {
      "epoch": 0.5229300684789375,
      "grad_norm": 2.430974288239865,
      "learning_rate": 5.471947800460502e-06,
      "loss": 0.4389,
      "step": 2520
    },
    {
      "epoch": 0.523967628138618,
      "grad_norm": 2.481418769699562,
      "learning_rate": 5.453916346295043e-06,
      "loss": 0.4516,
      "step": 2525
    },
    {
      "epoch": 0.5250051877982984,
      "grad_norm": 2.454530070786792,
      "learning_rate": 5.435878937753553e-06,
      "loss": 0.4461,
      "step": 2530
    },
    {
      "epoch": 0.5260427474579789,
      "grad_norm": 2.412430721171224,
      "learning_rate": 5.417835811446839e-06,
      "loss": 0.4516,
      "step": 2535
    },
    {
      "epoch": 0.5270803071176593,
      "grad_norm": 2.5095500970858566,
      "learning_rate": 5.3997872040607154e-06,
      "loss": 0.4647,
      "step": 2540
    },
    {
      "epoch": 0.5281178667773397,
      "grad_norm": 2.47241909883305,
      "learning_rate": 5.3817333523528895e-06,
      "loss": 0.4529,
      "step": 2545
    },
    {
      "epoch": 0.5291554264370202,
      "grad_norm": 2.5164134259250677,
      "learning_rate": 5.363674493149868e-06,
      "loss": 0.4584,
      "step": 2550
    },
    {
      "epoch": 0.5301929860967005,
      "grad_norm": 2.4698416937451073,
      "learning_rate": 5.345610863343843e-06,
      "loss": 0.4479,
      "step": 2555
    },
    {
      "epoch": 0.531230545756381,
      "grad_norm": 2.388109532933269,
      "learning_rate": 5.327542699889586e-06,
      "loss": 0.4527,
      "step": 2560
    },
    {
      "epoch": 0.5322681054160614,
      "grad_norm": 2.532256574198705,
      "learning_rate": 5.309470239801343e-06,
      "loss": 0.4541,
      "step": 2565
    },
    {
      "epoch": 0.5333056650757418,
      "grad_norm": 2.414906181196464,
      "learning_rate": 5.291393720149716e-06,
      "loss": 0.4415,
      "step": 2570
    },
    {
      "epoch": 0.5343432247354223,
      "grad_norm": 2.5860824474566146,
      "learning_rate": 5.273313378058566e-06,
      "loss": 0.4377,
      "step": 2575
    },
    {
      "epoch": 0.5353807843951027,
      "grad_norm": 2.6418841514226967,
      "learning_rate": 5.255229450701893e-06,
      "loss": 0.4342,
      "step": 2580
    },
    {
      "epoch": 0.5364183440547832,
      "grad_norm": 2.4497315903217247,
      "learning_rate": 5.237142175300726e-06,
      "loss": 0.4533,
      "step": 2585
    },
    {
      "epoch": 0.5374559037144636,
      "grad_norm": 2.4935146868944265,
      "learning_rate": 5.219051789120015e-06,
      "loss": 0.44,
      "step": 2590
    },
    {
      "epoch": 0.538493463374144,
      "grad_norm": 2.4091632402760346,
      "learning_rate": 5.200958529465517e-06,
      "loss": 0.454,
      "step": 2595
    },
    {
      "epoch": 0.5395310230338245,
      "grad_norm": 2.460401303833881,
      "learning_rate": 5.182862633680683e-06,
      "loss": 0.4512,
      "step": 2600
    },
    {
      "epoch": 0.5405685826935048,
      "grad_norm": 2.49980869010523,
      "learning_rate": 5.164764339143542e-06,
      "loss": 0.4531,
      "step": 2605
    },
    {
      "epoch": 0.5416061423531853,
      "grad_norm": 2.3750857632390767,
      "learning_rate": 5.14666388326359e-06,
      "loss": 0.4548,
      "step": 2610
    },
    {
      "epoch": 0.5426437020128657,
      "grad_norm": 2.402759567147983,
      "learning_rate": 5.128561503478676e-06,
      "loss": 0.4582,
      "step": 2615
    },
    {
      "epoch": 0.5436812616725462,
      "grad_norm": 2.387820424005501,
      "learning_rate": 5.110457437251886e-06,
      "loss": 0.4413,
      "step": 2620
    },
    {
      "epoch": 0.5447188213322266,
      "grad_norm": 2.3796099453754764,
      "learning_rate": 5.092351922068427e-06,
      "loss": 0.4524,
      "step": 2625
    },
    {
      "epoch": 0.545756380991907,
      "grad_norm": 2.3669548551534563,
      "learning_rate": 5.0742451954325156e-06,
      "loss": 0.4473,
      "step": 2630
    },
    {
      "epoch": 0.5467939406515875,
      "grad_norm": 2.463149244250711,
      "learning_rate": 5.056137494864259e-06,
      "loss": 0.447,
      "step": 2635
    },
    {
      "epoch": 0.5478315003112679,
      "grad_norm": 2.3744815980543494,
      "learning_rate": 5.0380290578965375e-06,
      "loss": 0.4404,
      "step": 2640
    },
    {
      "epoch": 0.5488690599709484,
      "grad_norm": 2.3391410347348796,
      "learning_rate": 5.019920122071896e-06,
      "loss": 0.4388,
      "step": 2645
    },
    {
      "epoch": 0.5499066196306288,
      "grad_norm": 2.3768157404751085,
      "learning_rate": 5.00181092493942e-06,
      "loss": 0.4386,
      "step": 2650
    },
    {
      "epoch": 0.5509441792903091,
      "grad_norm": 2.6059542471826824,
      "learning_rate": 4.983701704051625e-06,
      "loss": 0.4528,
      "step": 2655
    },
    {
      "epoch": 0.5519817389499896,
      "grad_norm": 2.417585242968673,
      "learning_rate": 4.965592696961335e-06,
      "loss": 0.4501,
      "step": 2660
    },
    {
      "epoch": 0.55301929860967,
      "grad_norm": 2.4947079086664123,
      "learning_rate": 4.947484141218572e-06,
      "loss": 0.4385,
      "step": 2665
    },
    {
      "epoch": 0.5540568582693505,
      "grad_norm": 2.5191730683436604,
      "learning_rate": 4.929376274367438e-06,
      "loss": 0.4324,
      "step": 2670
    },
    {
      "epoch": 0.5550944179290309,
      "grad_norm": 2.4257268049363336,
      "learning_rate": 4.911269333942994e-06,
      "loss": 0.4388,
      "step": 2675
    },
    {
      "epoch": 0.5561319775887114,
      "grad_norm": 2.5184876591899648,
      "learning_rate": 4.893163557468155e-06,
      "loss": 0.4316,
      "step": 2680
    },
    {
      "epoch": 0.5571695372483918,
      "grad_norm": 2.380468990224694,
      "learning_rate": 4.87505918245056e-06,
      "loss": 0.437,
      "step": 2685
    },
    {
      "epoch": 0.5582070969080722,
      "grad_norm": 2.3265663935377727,
      "learning_rate": 4.856956446379472e-06,
      "loss": 0.4307,
      "step": 2690
    },
    {
      "epoch": 0.5592446565677527,
      "grad_norm": 2.3400451782909646,
      "learning_rate": 4.838855586722647e-06,
      "loss": 0.4351,
      "step": 2695
    },
    {
      "epoch": 0.5602822162274331,
      "grad_norm": 2.4589440190512675,
      "learning_rate": 4.820756840923232e-06,
      "loss": 0.4447,
      "step": 2700
    },
    {
      "epoch": 0.5613197758871135,
      "grad_norm": 2.4951528488343184,
      "learning_rate": 4.802660446396642e-06,
      "loss": 0.4445,
      "step": 2705
    },
    {
      "epoch": 0.5623573355467939,
      "grad_norm": 2.666275326085189,
      "learning_rate": 4.784566640527451e-06,
      "loss": 0.4406,
      "step": 2710
    },
    {
      "epoch": 0.5633948952064743,
      "grad_norm": 2.4165213736699522,
      "learning_rate": 4.766475660666271e-06,
      "loss": 0.4222,
      "step": 2715
    },
    {
      "epoch": 0.5644324548661548,
      "grad_norm": 2.345464539142852,
      "learning_rate": 4.748387744126649e-06,
      "loss": 0.4355,
      "step": 2720
    },
    {
      "epoch": 0.5654700145258352,
      "grad_norm": 2.5576836547836215,
      "learning_rate": 4.730303128181944e-06,
      "loss": 0.4289,
      "step": 2725
    },
    {
      "epoch": 0.5665075741855157,
      "grad_norm": 2.5283882321035858,
      "learning_rate": 4.712222050062219e-06,
      "loss": 0.4283,
      "step": 2730
    },
    {
      "epoch": 0.5675451338451961,
      "grad_norm": 2.556873104995108,
      "learning_rate": 4.694144746951131e-06,
      "loss": 0.434,
      "step": 2735
    },
    {
      "epoch": 0.5685826935048766,
      "grad_norm": 2.280037463289983,
      "learning_rate": 4.676071455982811e-06,
      "loss": 0.4294,
      "step": 2740
    },
    {
      "epoch": 0.569620253164557,
      "grad_norm": 2.3570715386535555,
      "learning_rate": 4.658002414238771e-06,
      "loss": 0.4357,
      "step": 2745
    },
    {
      "epoch": 0.5706578128242374,
      "grad_norm": 2.4558176026821217,
      "learning_rate": 4.63993785874477e-06,
      "loss": 0.4421,
      "step": 2750
    },
    {
      "epoch": 0.5716953724839178,
      "grad_norm": 2.4779834247504082,
      "learning_rate": 4.621878026467725e-06,
      "loss": 0.4336,
      "step": 2755
    },
    {
      "epoch": 0.5727329321435982,
      "grad_norm": 2.3737617720589483,
      "learning_rate": 4.603823154312593e-06,
      "loss": 0.4263,
      "step": 2760
    },
    {
      "epoch": 0.5737704918032787,
      "grad_norm": 2.4648165069495147,
      "learning_rate": 4.585773479119265e-06,
      "loss": 0.4487,
      "step": 2765
    },
    {
      "epoch": 0.5748080514629591,
      "grad_norm": 2.3105842246846064,
      "learning_rate": 4.567729237659459e-06,
      "loss": 0.4252,
      "step": 2770
    },
    {
      "epoch": 0.5758456111226395,
      "grad_norm": 2.560822029486094,
      "learning_rate": 4.549690666633615e-06,
      "loss": 0.4432,
      "step": 2775
    },
    {
      "epoch": 0.57688317078232,
      "grad_norm": 2.6130981185416,
      "learning_rate": 4.531658002667787e-06,
      "loss": 0.4402,
      "step": 2780
    },
    {
      "epoch": 0.5779207304420004,
      "grad_norm": 2.363414455159581,
      "learning_rate": 4.51363148231055e-06,
      "loss": 0.4351,
      "step": 2785
    },
    {
      "epoch": 0.5789582901016809,
      "grad_norm": 2.363075041688293,
      "learning_rate": 4.495611342029875e-06,
      "loss": 0.428,
      "step": 2790
    },
    {
      "epoch": 0.5799958497613613,
      "grad_norm": 2.4754540310783733,
      "learning_rate": 4.477597818210054e-06,
      "loss": 0.4246,
      "step": 2795
    },
    {
      "epoch": 0.5810334094210418,
      "grad_norm": 2.6816555150046755,
      "learning_rate": 4.459591147148575e-06,
      "loss": 0.4253,
      "step": 2800
    },
    {
      "epoch": 0.5820709690807221,
      "grad_norm": 2.4036860880645894,
      "learning_rate": 4.441591565053041e-06,
      "loss": 0.4272,
      "step": 2805
    },
    {
      "epoch": 0.5831085287404025,
      "grad_norm": 2.4802074251251156,
      "learning_rate": 4.423599308038057e-06,
      "loss": 0.4209,
      "step": 2810
    },
    {
      "epoch": 0.584146088400083,
      "grad_norm": 2.415585886184946,
      "learning_rate": 4.405614612122145e-06,
      "loss": 0.4226,
      "step": 2815
    },
    {
      "epoch": 0.5851836480597634,
      "grad_norm": 2.4533356227902816,
      "learning_rate": 4.387637713224638e-06,
      "loss": 0.4343,
      "step": 2820
    },
    {
      "epoch": 0.5862212077194439,
      "grad_norm": 2.7240629036546284,
      "learning_rate": 4.36966884716259e-06,
      "loss": 0.4287,
      "step": 2825
    },
    {
      "epoch": 0.5872587673791243,
      "grad_norm": 2.3878324516316276,
      "learning_rate": 4.3517082496476845e-06,
      "loss": 0.4184,
      "step": 2830
    },
    {
      "epoch": 0.5882963270388047,
      "grad_norm": 2.422038027754476,
      "learning_rate": 4.333756156283136e-06,
      "loss": 0.4209,
      "step": 2835
    },
    {
      "epoch": 0.5893338866984852,
      "grad_norm": 2.5027178602080697,
      "learning_rate": 4.315812802560609e-06,
      "loss": 0.4151,
      "step": 2840
    },
    {
      "epoch": 0.5903714463581656,
      "grad_norm": 2.440654195509635,
      "learning_rate": 4.2978784238571145e-06,
      "loss": 0.424,
      "step": 2845
    },
    {
      "epoch": 0.5914090060178461,
      "grad_norm": 2.4534970838374206,
      "learning_rate": 4.279953255431944e-06,
      "loss": 0.427,
      "step": 2850
    },
    {
      "epoch": 0.5924465656775264,
      "grad_norm": 2.4846011931620087,
      "learning_rate": 4.262037532423556e-06,
      "loss": 0.4376,
      "step": 2855
    },
    {
      "epoch": 0.5934841253372068,
      "grad_norm": 2.4555820059315,
      "learning_rate": 4.244131489846519e-06,
      "loss": 0.4102,
      "step": 2860
    },
    {
      "epoch": 0.5945216849968873,
      "grad_norm": 2.443517931338813,
      "learning_rate": 4.2262353625884054e-06,
      "loss": 0.4138,
      "step": 2865
    },
    {
      "epoch": 0.5955592446565677,
      "grad_norm": 2.369110831081435,
      "learning_rate": 4.208349385406729e-06,
      "loss": 0.4364,
      "step": 2870
    },
    {
      "epoch": 0.5965968043162482,
      "grad_norm": 2.3567775981536423,
      "learning_rate": 4.190473792925851e-06,
      "loss": 0.4277,
      "step": 2875
    },
    {
      "epoch": 0.5976343639759286,
      "grad_norm": 2.3822530664354216,
      "learning_rate": 4.1726088196339106e-06,
      "loss": 0.4266,
      "step": 2880
    },
    {
      "epoch": 0.5986719236356091,
      "grad_norm": 2.3958405655171293,
      "learning_rate": 4.154754699879748e-06,
      "loss": 0.4177,
      "step": 2885
    },
    {
      "epoch": 0.5997094832952895,
      "grad_norm": 2.3762721908121573,
      "learning_rate": 4.136911667869827e-06,
      "loss": 0.4146,
      "step": 2890
    },
    {
      "epoch": 0.60074704295497,
      "grad_norm": 2.2932296484786137,
      "learning_rate": 4.119079957665163e-06,
      "loss": 0.4074,
      "step": 2895
    },
    {
      "epoch": 0.6017846026146504,
      "grad_norm": 2.446668454803684,
      "learning_rate": 4.101259803178265e-06,
      "loss": 0.4318,
      "step": 2900
    },
    {
      "epoch": 0.6028221622743307,
      "grad_norm": 2.361867602889005,
      "learning_rate": 4.083451438170039e-06,
      "loss": 0.4098,
      "step": 2905
    },
    {
      "epoch": 0.6038597219340112,
      "grad_norm": 2.323202296910631,
      "learning_rate": 4.065655096246755e-06,
      "loss": 0.408,
      "step": 2910
    },
    {
      "epoch": 0.6048972815936916,
      "grad_norm": 2.3845839128387722,
      "learning_rate": 4.047871010856959e-06,
      "loss": 0.4071,
      "step": 2915
    },
    {
      "epoch": 0.605934841253372,
      "grad_norm": 2.317266250356725,
      "learning_rate": 4.03009941528842e-06,
      "loss": 0.4051,
      "step": 2920
    },
    {
      "epoch": 0.6069724009130525,
      "grad_norm": 2.4278074780829852,
      "learning_rate": 4.012340542665067e-06,
      "loss": 0.4002,
      "step": 2925
    },
    {
      "epoch": 0.6080099605727329,
      "grad_norm": 2.4948657955155853,
      "learning_rate": 3.994594625943936e-06,
      "loss": 0.4103,
      "step": 2930
    },
    {
      "epoch": 0.6090475202324134,
      "grad_norm": 2.4097506379402596,
      "learning_rate": 3.976861897912106e-06,
      "loss": 0.4137,
      "step": 2935
    },
    {
      "epoch": 0.6100850798920938,
      "grad_norm": 2.695103229335665,
      "learning_rate": 3.959142591183652e-06,
      "loss": 0.4184,
      "step": 2940
    },
    {
      "epoch": 0.6111226395517743,
      "grad_norm": 2.433881870125308,
      "learning_rate": 3.9414369381965904e-06,
      "loss": 0.4084,
      "step": 2945
    },
    {
      "epoch": 0.6121601992114547,
      "grad_norm": 2.424853521797893,
      "learning_rate": 3.92374517120983e-06,
      "loss": 0.4157,
      "step": 2950
    },
    {
      "epoch": 0.6131977588711351,
      "grad_norm": 2.3289318124697926,
      "learning_rate": 3.90606752230013e-06,
      "loss": 0.4002,
      "step": 2955
    },
    {
      "epoch": 0.6142353185308155,
      "grad_norm": 2.4465640629382186,
      "learning_rate": 3.888404223359045e-06,
      "loss": 0.4057,
      "step": 2960
    },
    {
      "epoch": 0.6152728781904959,
      "grad_norm": 2.572631649766761,
      "learning_rate": 3.870755506089899e-06,
      "loss": 0.4144,
      "step": 2965
    },
    {
      "epoch": 0.6163104378501764,
      "grad_norm": 2.4132420332744204,
      "learning_rate": 3.8531216020047246e-06,
      "loss": 0.4116,
      "step": 2970
    },
    {
      "epoch": 0.6173479975098568,
      "grad_norm": 2.3892278386798593,
      "learning_rate": 3.835502742421251e-06,
      "loss": 0.4093,
      "step": 2975
    },
    {
      "epoch": 0.6183855571695372,
      "grad_norm": 2.2641401615974406,
      "learning_rate": 3.8178991584598474e-06,
      "loss": 0.4131,
      "step": 2980
    },
    {
      "epoch": 0.6194231168292177,
      "grad_norm": 2.438762066594083,
      "learning_rate": 3.8003110810405065e-06,
      "loss": 0.4064,
      "step": 2985
    },
    {
      "epoch": 0.6204606764888981,
      "grad_norm": 2.4506479186056396,
      "learning_rate": 3.782738740879806e-06,
      "loss": 0.4052,
      "step": 2990
    },
    {
      "epoch": 0.6214982361485786,
      "grad_norm": 2.5283170692610617,
      "learning_rate": 3.7651823684878884e-06,
      "loss": 0.396,
      "step": 2995
    },
    {
      "epoch": 0.622535795808259,
      "grad_norm": 2.4749410345861453,
      "learning_rate": 3.7476421941654318e-06,
      "loss": 0.4193,
      "step": 3000
    },
    {
      "epoch": 0.6235733554679395,
      "grad_norm": 2.317549282593909,
      "learning_rate": 3.7301184480006337e-06,
      "loss": 0.3973,
      "step": 3005
    },
    {
      "epoch": 0.6246109151276198,
      "grad_norm": 2.5252500112854275,
      "learning_rate": 3.712611359866188e-06,
      "loss": 0.4147,
      "step": 3010
    },
    {
      "epoch": 0.6256484747873002,
      "grad_norm": 2.4400394840947475,
      "learning_rate": 3.6951211594162784e-06,
      "loss": 0.4089,
      "step": 3015
    },
    {
      "epoch": 0.6266860344469807,
      "grad_norm": 2.349957863539589,
      "learning_rate": 3.677648076083549e-06,
      "loss": 0.3992,
      "step": 3020
    },
    {
      "epoch": 0.6277235941066611,
      "grad_norm": 2.3832033455743833,
      "learning_rate": 3.6601923390761156e-06,
      "loss": 0.4131,
      "step": 3025
    },
    {
      "epoch": 0.6287611537663416,
      "grad_norm": 2.576394180910392,
      "learning_rate": 3.6427541773745433e-06,
      "loss": 0.3968,
      "step": 3030
    },
    {
      "epoch": 0.629798713426022,
      "grad_norm": 2.4507894648738877,
      "learning_rate": 3.6253338197288505e-06,
      "loss": 0.4023,
      "step": 3035
    },
    {
      "epoch": 0.6308362730857024,
      "grad_norm": 2.3382065135829997,
      "learning_rate": 3.607931494655504e-06,
      "loss": 0.3918,
      "step": 3040
    },
    {
      "epoch": 0.6318738327453829,
      "grad_norm": 2.4244914619480373,
      "learning_rate": 3.5905474304344225e-06,
      "loss": 0.4117,
      "step": 3045
    },
    {
      "epoch": 0.6329113924050633,
      "grad_norm": 2.567189154890026,
      "learning_rate": 3.573181855105986e-06,
      "loss": 0.42,
      "step": 3050
    },
    {
      "epoch": 0.6339489520647438,
      "grad_norm": 2.278238684757535,
      "learning_rate": 3.555834996468039e-06,
      "loss": 0.4041,
      "step": 3055
    },
    {
      "epoch": 0.6349865117244241,
      "grad_norm": 2.504690078277028,
      "learning_rate": 3.538507082072905e-06,
      "loss": 0.3944,
      "step": 3060
    },
    {
      "epoch": 0.6360240713841046,
      "grad_norm": 2.508152236947373,
      "learning_rate": 3.5211983392243996e-06,
      "loss": 0.4,
      "step": 3065
    },
    {
      "epoch": 0.637061631043785,
      "grad_norm": 2.3910619945271683,
      "learning_rate": 3.503908994974856e-06,
      "loss": 0.4093,
      "step": 3070
    },
    {
      "epoch": 0.6380991907034654,
      "grad_norm": 2.3248985179294652,
      "learning_rate": 3.4866392761221303e-06,
      "loss": 0.4065,
      "step": 3075
    },
    {
      "epoch": 0.6391367503631459,
      "grad_norm": 2.361597466745494,
      "learning_rate": 3.4693894092066483e-06,
      "loss": 0.3907,
      "step": 3080
    },
    {
      "epoch": 0.6401743100228263,
      "grad_norm": 2.4949060134883743,
      "learning_rate": 3.452159620508414e-06,
      "loss": 0.4004,
      "step": 3085
    },
    {
      "epoch": 0.6412118696825068,
      "grad_norm": 2.3586999565378886,
      "learning_rate": 3.4349501360440556e-06,
      "loss": 0.3977,
      "step": 3090
    },
    {
      "epoch": 0.6422494293421872,
      "grad_norm": 2.2746540643844426,
      "learning_rate": 3.417761181563849e-06,
      "loss": 0.3949,
      "step": 3095
    },
    {
      "epoch": 0.6432869890018676,
      "grad_norm": 2.4940700061779064,
      "learning_rate": 3.4005929825487684e-06,
      "loss": 0.4011,
      "step": 3100
    },
    {
      "epoch": 0.6443245486615481,
      "grad_norm": 2.3950183894327934,
      "learning_rate": 3.383445764207516e-06,
      "loss": 0.408,
      "step": 3105
    },
    {
      "epoch": 0.6453621083212284,
      "grad_norm": 2.30382353091722,
      "learning_rate": 3.366319751473579e-06,
      "loss": 0.4022,
      "step": 3110
    },
    {
      "epoch": 0.6463996679809089,
      "grad_norm": 2.4321346720182766,
      "learning_rate": 3.3492151690022712e-06,
      "loss": 0.3986,
      "step": 3115
    },
    {
      "epoch": 0.6474372276405893,
      "grad_norm": 2.3061011632873285,
      "learning_rate": 3.332132241167793e-06,
      "loss": 0.3972,
      "step": 3120
    },
    {
      "epoch": 0.6484747873002698,
      "grad_norm": 2.4917204012266327,
      "learning_rate": 3.3150711920602765e-06,
      "loss": 0.4042,
      "step": 3125
    },
    {
      "epoch": 0.6495123469599502,
      "grad_norm": 2.190798555385459,
      "learning_rate": 3.2980322454828617e-06,
      "loss": 0.3917,
      "step": 3130
    },
    {
      "epoch": 0.6505499066196306,
      "grad_norm": 2.343964465011578,
      "learning_rate": 3.281015624948746e-06,
      "loss": 0.3893,
      "step": 3135
    },
    {
      "epoch": 0.6515874662793111,
      "grad_norm": 2.61426487111248,
      "learning_rate": 3.264021553678264e-06,
      "loss": 0.4087,
      "step": 3140
    },
    {
      "epoch": 0.6526250259389915,
      "grad_norm": 2.3381914892622553,
      "learning_rate": 3.247050254595947e-06,
      "loss": 0.3996,
      "step": 3145
    },
    {
      "epoch": 0.653662585598672,
      "grad_norm": 2.4097371109758945,
      "learning_rate": 3.2301019503276144e-06,
      "loss": 0.404,
      "step": 3150
    },
    {
      "epoch": 0.6547001452583524,
      "grad_norm": 2.3425157027203127,
      "learning_rate": 3.2131768631974375e-06,
      "loss": 0.4025,
      "step": 3155
    },
    {
      "epoch": 0.6557377049180327,
      "grad_norm": 2.405066414410578,
      "learning_rate": 3.196275215225032e-06,
      "loss": 0.4095,
      "step": 3160
    },
    {
      "epoch": 0.6567752645777132,
      "grad_norm": 2.332707293053418,
      "learning_rate": 3.179397228122547e-06,
      "loss": 0.4,
      "step": 3165
    },
    {
      "epoch": 0.6578128242373936,
      "grad_norm": 2.374933561243685,
      "learning_rate": 3.162543123291749e-06,
      "loss": 0.3887,
      "step": 3170
    },
    {
      "epoch": 0.6588503838970741,
      "grad_norm": 2.265067002723173,
      "learning_rate": 3.1457131218211263e-06,
      "loss": 0.3974,
      "step": 3175
    },
    {
      "epoch": 0.6598879435567545,
      "grad_norm": 2.363694486377073,
      "learning_rate": 3.1289074444829783e-06,
      "loss": 0.3932,
      "step": 3180
    },
    {
      "epoch": 0.660925503216435,
      "grad_norm": 2.5691413705019603,
      "learning_rate": 3.1121263117305355e-06,
      "loss": 0.3848,
      "step": 3185
    },
    {
      "epoch": 0.6619630628761154,
      "grad_norm": 2.3700655398167005,
      "learning_rate": 3.0953699436950464e-06,
      "loss": 0.3942,
      "step": 3190
    },
    {
      "epoch": 0.6630006225357958,
      "grad_norm": 2.5079197198404932,
      "learning_rate": 3.0786385601829114e-06,
      "loss": 0.3921,
      "step": 3195
    },
    {
      "epoch": 0.6640381821954763,
      "grad_norm": 2.3572529532760136,
      "learning_rate": 3.061932380672783e-06,
      "loss": 0.389,
      "step": 3200
    },
    {
      "epoch": 0.6650757418551567,
      "grad_norm": 2.3406368361183643,
      "learning_rate": 3.0452516243126955e-06,
      "loss": 0.3942,
      "step": 3205
    },
    {
      "epoch": 0.666113301514837,
      "grad_norm": 2.2910811346177953,
      "learning_rate": 3.0285965099171864e-06,
      "loss": 0.3954,
      "step": 3210
    },
    {
      "epoch": 0.6671508611745175,
      "grad_norm": 2.3711327251353254,
      "learning_rate": 3.0119672559644313e-06,
      "loss": 0.3825,
      "step": 3215
    },
    {
      "epoch": 0.6681884208341979,
      "grad_norm": 2.346432027672407,
      "learning_rate": 2.995364080593368e-06,
      "loss": 0.3862,
      "step": 3220
    },
    {
      "epoch": 0.6692259804938784,
      "grad_norm": 2.25488355968479,
      "learning_rate": 2.978787201600847e-06,
      "loss": 0.3869,
      "step": 3225
    },
    {
      "epoch": 0.6702635401535588,
      "grad_norm": 2.5145720297012204,
      "learning_rate": 2.9622368364387626e-06,
      "loss": 0.3979,
      "step": 3230
    },
    {
      "epoch": 0.6713010998132393,
      "grad_norm": 2.430288285642923,
      "learning_rate": 2.9457132022112156e-06,
      "loss": 0.3876,
      "step": 3235
    },
    {
      "epoch": 0.6723386594729197,
      "grad_norm": 2.52733248085438,
      "learning_rate": 2.9292165156716447e-06,
      "loss": 0.3918,
      "step": 3240
    },
    {
      "epoch": 0.6733762191326002,
      "grad_norm": 2.4117862477617464,
      "learning_rate": 2.9127469932200034e-06,
      "loss": 0.3904,
      "step": 3245
    },
    {
      "epoch": 0.6744137787922806,
      "grad_norm": 2.544838045591555,
      "learning_rate": 2.89630485089991e-06,
      "loss": 0.3968,
      "step": 3250
    },
    {
      "epoch": 0.675451338451961,
      "grad_norm": 2.3114731154281687,
      "learning_rate": 2.879890304395816e-06,
      "loss": 0.389,
      "step": 3255
    },
    {
      "epoch": 0.6764888981116414,
      "grad_norm": 2.214278091013657,
      "learning_rate": 2.8635035690301725e-06,
      "loss": 0.368,
      "step": 3260
    },
    {
      "epoch": 0.6775264577713218,
      "grad_norm": 2.4150879417081033,
      "learning_rate": 2.847144859760622e-06,
      "loss": 0.3997,
      "step": 3265
    },
    {
      "epoch": 0.6785640174310023,
      "grad_norm": 2.2503456214277766,
      "learning_rate": 2.8308143911771555e-06,
      "loss": 0.3785,
      "step": 3270
    },
    {
      "epoch": 0.6796015770906827,
      "grad_norm": 2.400256629243757,
      "learning_rate": 2.8145123774993075e-06,
      "loss": 0.3873,
      "step": 3275
    },
    {
      "epoch": 0.6806391367503631,
      "grad_norm": 2.4464548102871393,
      "learning_rate": 2.798239032573362e-06,
      "loss": 0.3811,
      "step": 3280
    },
    {
      "epoch": 0.6816766964100436,
      "grad_norm": 2.5094627463784973,
      "learning_rate": 2.7819945698695148e-06,
      "loss": 0.387,
      "step": 3285
    },
    {
      "epoch": 0.682714256069724,
      "grad_norm": 2.4588062818595944,
      "learning_rate": 2.765779202479103e-06,
      "loss": 0.3848,
      "step": 3290
    },
    {
      "epoch": 0.6837518157294045,
      "grad_norm": 2.2530973533871035,
      "learning_rate": 2.749593143111793e-06,
      "loss": 0.3641,
      "step": 3295
    },
    {
      "epoch": 0.6847893753890849,
      "grad_norm": 2.3169454419169684,
      "learning_rate": 2.733436604092797e-06,
      "loss": 0.378,
      "step": 3300
    },
    {
      "epoch": 0.6858269350487654,
      "grad_norm": 2.5056936406549917,
      "learning_rate": 2.7173097973600806e-06,
      "loss": 0.3837,
      "step": 3305
    },
    {
      "epoch": 0.6868644947084457,
      "grad_norm": 2.315516361770829,
      "learning_rate": 2.7012129344615933e-06,
      "loss": 0.3797,
      "step": 3310
    },
    {
      "epoch": 0.6879020543681261,
      "grad_norm": 2.3765574424439504,
      "learning_rate": 2.6851462265524862e-06,
      "loss": 0.3821,
      "step": 3315
    },
    {
      "epoch": 0.6889396140278066,
      "grad_norm": 2.632664268427571,
      "learning_rate": 2.6691098843923464e-06,
      "loss": 0.3869,
      "step": 3320
    },
    {
      "epoch": 0.689977173687487,
      "grad_norm": 2.4558633695260554,
      "learning_rate": 2.65310411834242e-06,
      "loss": 0.3778,
      "step": 3325
    },
    {
      "epoch": 0.6910147333471675,
      "grad_norm": 2.3802011004843053,
      "learning_rate": 2.637129138362877e-06,
      "loss": 0.3818,
      "step": 3330
    },
    {
      "epoch": 0.6920522930068479,
      "grad_norm": 2.3545532106676625,
      "learning_rate": 2.62118515401003e-06,
      "loss": 0.3745,
      "step": 3335
    },
    {
      "epoch": 0.6930898526665283,
      "grad_norm": 2.2370151266527913,
      "learning_rate": 2.6052723744336027e-06,
      "loss": 0.382,
      "step": 3340
    },
    {
      "epoch": 0.6941274123262088,
      "grad_norm": 2.3233925949072645,
      "learning_rate": 2.589391008373982e-06,
      "loss": 0.3901,
      "step": 3345
    },
    {
      "epoch": 0.6951649719858892,
      "grad_norm": 2.385447861484446,
      "learning_rate": 2.5735412641594804e-06,
      "loss": 0.3804,
      "step": 3350
    },
    {
      "epoch": 0.6962025316455697,
      "grad_norm": 2.3578389155683217,
      "learning_rate": 2.5577233497035943e-06,
      "loss": 0.3888,
      "step": 3355
    },
    {
      "epoch": 0.69724009130525,
      "grad_norm": 2.3006918024124965,
      "learning_rate": 2.541937472502293e-06,
      "loss": 0.3661,
      "step": 3360
    },
    {
      "epoch": 0.6982776509649304,
      "grad_norm": 2.303381728187516,
      "learning_rate": 2.526183839631283e-06,
      "loss": 0.3827,
      "step": 3365
    },
    {
      "epoch": 0.6993152106246109,
      "grad_norm": 2.324125134648238,
      "learning_rate": 2.5104626577433022e-06,
      "loss": 0.376,
      "step": 3370
    },
    {
      "epoch": 0.7003527702842913,
      "grad_norm": 2.4863826311733614,
      "learning_rate": 2.4947741330653942e-06,
      "loss": 0.3765,
      "step": 3375
    },
    {
      "epoch": 0.7013903299439718,
      "grad_norm": 2.309178058770916,
      "learning_rate": 2.4791184713962207e-06,
      "loss": 0.3665,
      "step": 3380
    },
    {
      "epoch": 0.7024278896036522,
      "grad_norm": 2.338910146993185,
      "learning_rate": 2.463495878103352e-06,
      "loss": 0.3601,
      "step": 3385
    },
    {
      "epoch": 0.7034654492633327,
      "grad_norm": 2.386442857917931,
      "learning_rate": 2.4479065581205673e-06,
      "loss": 0.3775,
      "step": 3390
    },
    {
      "epoch": 0.7045030089230131,
      "grad_norm": 2.4264131593130913,
      "learning_rate": 2.4323507159451887e-06,
      "loss": 0.3775,
      "step": 3395
    },
    {
      "epoch": 0.7055405685826935,
      "grad_norm": 35.69245904532589,
      "learning_rate": 2.416828555635368e-06,
      "loss": 0.3836,
      "step": 3400
    },
    {
      "epoch": 0.706578128242374,
      "grad_norm": 2.3981493771293385,
      "learning_rate": 2.4013402808074356e-06,
      "loss": 0.3648,
      "step": 3405
    },
    {
      "epoch": 0.7076156879020543,
      "grad_norm": 2.458681006711934,
      "learning_rate": 2.3858860946332148e-06,
      "loss": 0.3706,
      "step": 3410
    },
    {
      "epoch": 0.7086532475617348,
      "grad_norm": 2.5072358874605545,
      "learning_rate": 2.3704661998373652e-06,
      "loss": 0.3786,
      "step": 3415
    },
    {
      "epoch": 0.7096908072214152,
      "grad_norm": 2.4783372667192576,
      "learning_rate": 2.3550807986947133e-06,
      "loss": 0.3635,
      "step": 3420
    },
    {
      "epoch": 0.7107283668810956,
      "grad_norm": 2.583412985853997,
      "learning_rate": 2.3397300930276116e-06,
      "loss": 0.3574,
      "step": 3425
    },
    {
      "epoch": 0.7117659265407761,
      "grad_norm": 2.39052336222071,
      "learning_rate": 2.3244142842032823e-06,
      "loss": 0.3642,
      "step": 3430
    },
    {
      "epoch": 0.7128034862004565,
      "grad_norm": 2.2998480766526974,
      "learning_rate": 2.309133573131181e-06,
      "loss": 0.3659,
      "step": 3435
    },
    {
      "epoch": 0.713841045860137,
      "grad_norm": 2.452595068182353,
      "learning_rate": 2.2938881602603496e-06,
      "loss": 0.3767,
      "step": 3440
    },
    {
      "epoch": 0.7148786055198174,
      "grad_norm": 2.4427339177937597,
      "learning_rate": 2.2786782455768113e-06,
      "loss": 0.3701,
      "step": 3445
    },
    {
      "epoch": 0.7159161651794979,
      "grad_norm": 2.5344881209783647,
      "learning_rate": 2.2635040286009163e-06,
      "loss": 0.3636,
      "step": 3450
    },
    {
      "epoch": 0.7169537248391783,
      "grad_norm": 2.4517299777866914,
      "learning_rate": 2.2483657083847487e-06,
      "loss": 0.3638,
      "step": 3455
    },
    {
      "epoch": 0.7179912844988587,
      "grad_norm": 2.333164872282216,
      "learning_rate": 2.233263483509505e-06,
      "loss": 0.3652,
      "step": 3460
    },
    {
      "epoch": 0.7190288441585391,
      "grad_norm": 2.536346581261384,
      "learning_rate": 2.218197552082893e-06,
      "loss": 0.3843,
      "step": 3465
    },
    {
      "epoch": 0.7200664038182195,
      "grad_norm": 2.3170765412707444,
      "learning_rate": 2.203168111736524e-06,
      "loss": 0.3702,
      "step": 3470
    },
    {
      "epoch": 0.7211039634779,
      "grad_norm": 2.3720426965424926,
      "learning_rate": 2.1881753596233334e-06,
      "loss": 0.3661,
      "step": 3475
    },
    {
      "epoch": 0.7221415231375804,
      "grad_norm": 2.4784864232706427,
      "learning_rate": 2.173219492414988e-06,
      "loss": 0.3557,
      "step": 3480
    },
    {
      "epoch": 0.7231790827972608,
      "grad_norm": 2.2572170707401598,
      "learning_rate": 2.1583007062993037e-06,
      "loss": 0.3626,
      "step": 3485
    },
    {
      "epoch": 0.7242166424569413,
      "grad_norm": 2.3751613572262618,
      "learning_rate": 2.1434191969776787e-06,
      "loss": 0.372,
      "step": 3490
    },
    {
      "epoch": 0.7252542021166217,
      "grad_norm": 2.403916983756595,
      "learning_rate": 2.1285751596625153e-06,
      "loss": 0.3793,
      "step": 3495
    },
    {
      "epoch": 0.7262917617763022,
      "grad_norm": 2.422300937120251,
      "learning_rate": 2.1137687890746733e-06,
      "loss": 0.3757,
      "step": 3500
    },
    {
      "epoch": 0.7273293214359826,
      "grad_norm": 2.3278136492746904,
      "learning_rate": 2.099000279440905e-06,
      "loss": 0.3695,
      "step": 3505
    },
    {
      "epoch": 0.728366881095663,
      "grad_norm": 2.3521512237809774,
      "learning_rate": 2.0842698244913146e-06,
      "loss": 0.3577,
      "step": 3510
    },
    {
      "epoch": 0.7294044407553434,
      "grad_norm": 2.4420360123708837,
      "learning_rate": 2.0695776174568054e-06,
      "loss": 0.3617,
      "step": 3515
    },
    {
      "epoch": 0.7304420004150238,
      "grad_norm": 2.355326415228057,
      "learning_rate": 2.054923851066561e-06,
      "loss": 0.366,
      "step": 3520
    },
    {
      "epoch": 0.7314795600747043,
      "grad_norm": 2.408952585713157,
      "learning_rate": 2.0403087175455044e-06,
      "loss": 0.3592,
      "step": 3525
    },
    {
      "epoch": 0.7325171197343847,
      "grad_norm": 2.3814394489123556,
      "learning_rate": 2.025732408611786e-06,
      "loss": 0.3702,
      "step": 3530
    },
    {
      "epoch": 0.7335546793940652,
      "grad_norm": 2.505258380168417,
      "learning_rate": 2.0111951154742526e-06,
      "loss": 0.3541,
      "step": 3535
    },
    {
      "epoch": 0.7345922390537456,
      "grad_norm": 2.2378548590102962,
      "learning_rate": 1.9966970288299666e-06,
      "loss": 0.367,
      "step": 3540
    },
    {
      "epoch": 0.735629798713426,
      "grad_norm": 2.5282472044177657,
      "learning_rate": 1.982238338861673e-06,
      "loss": 0.372,
      "step": 3545
    },
    {
      "epoch": 0.7366673583731065,
      "grad_norm": 2.395346352470493,
      "learning_rate": 1.9678192352353282e-06,
      "loss": 0.3647,
      "step": 3550
    },
    {
      "epoch": 0.7377049180327869,
      "grad_norm": 2.377816912239941,
      "learning_rate": 1.9534399070976013e-06,
      "loss": 0.3694,
      "step": 3555
    },
    {
      "epoch": 0.7387424776924674,
      "grad_norm": 2.395069006251528,
      "learning_rate": 1.9391005430733973e-06,
      "loss": 0.3643,
      "step": 3560
    },
    {
      "epoch": 0.7397800373521477,
      "grad_norm": 2.360921498446891,
      "learning_rate": 1.924801331263375e-06,
      "loss": 0.3592,
      "step": 3565
    },
    {
      "epoch": 0.7408175970118281,
      "grad_norm": 2.4026862396255977,
      "learning_rate": 1.9105424592414905e-06,
      "loss": 0.3596,
      "step": 3570
    },
    {
      "epoch": 0.7418551566715086,
      "grad_norm": 2.3493974707324323,
      "learning_rate": 1.8963241140525302e-06,
      "loss": 0.352,
      "step": 3575
    },
    {
      "epoch": 0.742892716331189,
      "grad_norm": 2.255383769774033,
      "learning_rate": 1.8821464822096587e-06,
      "loss": 0.3545,
      "step": 3580
    },
    {
      "epoch": 0.7439302759908695,
      "grad_norm": 2.3439036527893804,
      "learning_rate": 1.8680097496919663e-06,
      "loss": 0.3614,
      "step": 3585
    },
    {
      "epoch": 0.7449678356505499,
      "grad_norm": 2.402995035755627,
      "learning_rate": 1.8539141019420459e-06,
      "loss": 0.3591,
      "step": 3590
    },
    {
      "epoch": 0.7460053953102304,
      "grad_norm": 2.3095366349656845,
      "learning_rate": 1.8398597238635375e-06,
      "loss": 0.3555,
      "step": 3595
    },
    {
      "epoch": 0.7470429549699108,
      "grad_norm": 2.454441768372595,
      "learning_rate": 1.825846799818722e-06,
      "loss": 0.3645,
      "step": 3600
    },
    {
      "epoch": 0.7480805146295912,
      "grad_norm": 2.3783282437356883,
      "learning_rate": 1.8118755136260941e-06,
      "loss": 0.364,
      "step": 3605
    },
    {
      "epoch": 0.7491180742892717,
      "grad_norm": 2.4804513717759975,
      "learning_rate": 1.7979460485579486e-06,
      "loss": 0.3577,
      "step": 3610
    },
    {
      "epoch": 0.750155633948952,
      "grad_norm": 2.399533389844388,
      "learning_rate": 1.784058587337984e-06,
      "loss": 0.3548,
      "step": 3615
    },
    {
      "epoch": 0.7511931936086325,
      "grad_norm": 2.5287966988574992,
      "learning_rate": 1.7702133121388999e-06,
      "loss": 0.3702,
      "step": 3620
    },
    {
      "epoch": 0.7522307532683129,
      "grad_norm": 2.5280174627826097,
      "learning_rate": 1.7564104045800101e-06,
      "loss": 0.3717,
      "step": 3625
    },
    {
      "epoch": 0.7532683129279933,
      "grad_norm": 2.454120509191578,
      "learning_rate": 1.7426500457248552e-06,
      "loss": 0.3508,
      "step": 3630
    },
    {
      "epoch": 0.7543058725876738,
      "grad_norm": 2.2917885535880176,
      "learning_rate": 1.7289324160788346e-06,
      "loss": 0.3509,
      "step": 3635
    },
    {
      "epoch": 0.7553434322473542,
      "grad_norm": 2.478759913910432,
      "learning_rate": 1.7152576955868338e-06,
      "loss": 0.3597,
      "step": 3640
    },
    {
      "epoch": 0.7563809919070347,
      "grad_norm": 2.3448679544687256,
      "learning_rate": 1.701626063630869e-06,
      "loss": 0.367,
      "step": 3645
    },
    {
      "epoch": 0.7574185515667151,
      "grad_norm": 2.3452502536112485,
      "learning_rate": 1.6880376990277202e-06,
      "loss": 0.3548,
      "step": 3650
    },
    {
      "epoch": 0.7584561112263956,
      "grad_norm": 2.3587859019469084,
      "learning_rate": 1.674492780026611e-06,
      "loss": 0.3635,
      "step": 3655
    },
    {
      "epoch": 0.759493670886076,
      "grad_norm": 2.4281427875718635,
      "learning_rate": 1.6609914843068403e-06,
      "loss": 0.3535,
      "step": 3660
    },
    {
      "epoch": 0.7605312305457563,
      "grad_norm": 2.644672013859547,
      "learning_rate": 1.6475339889754755e-06,
      "loss": 0.3546,
      "step": 3665
    },
    {
      "epoch": 0.7615687902054368,
      "grad_norm": 2.231863917742126,
      "learning_rate": 1.6341204705650155e-06,
      "loss": 0.3565,
      "step": 3670
    },
    {
      "epoch": 0.7626063498651172,
      "grad_norm": 2.3627064704364003,
      "learning_rate": 1.6207511050310842e-06,
      "loss": 0.3601,
      "step": 3675
    },
    {
      "epoch": 0.7636439095247977,
      "grad_norm": 2.5482612194562746,
      "learning_rate": 1.6074260677501102e-06,
      "loss": 0.3606,
      "step": 3680
    },
    {
      "epoch": 0.7646814691844781,
      "grad_norm": 2.4794463771443187,
      "learning_rate": 1.5941455335170408e-06,
      "loss": 0.3538,
      "step": 3685
    },
    {
      "epoch": 0.7657190288441585,
      "grad_norm": 2.6762628509028565,
      "learning_rate": 1.5809096765430387e-06,
      "loss": 0.3517,
      "step": 3690
    },
    {
      "epoch": 0.766756588503839,
      "grad_norm": 2.2429463729622565,
      "learning_rate": 1.5677186704532016e-06,
      "loss": 0.3457,
      "step": 3695
    },
    {
      "epoch": 0.7677941481635194,
      "grad_norm": 2.36484474879296,
      "learning_rate": 1.5545726882842782e-06,
      "loss": 0.3573,
      "step": 3700
    },
    {
      "epoch": 0.7688317078231999,
      "grad_norm": 2.294696024445774,
      "learning_rate": 1.5414719024824127e-06,
      "loss": 0.3509,
      "step": 3705
    },
    {
      "epoch": 0.7698692674828803,
      "grad_norm": 2.51233601030323,
      "learning_rate": 1.5284164849008648e-06,
      "loss": 0.3589,
      "step": 3710
    },
    {
      "epoch": 0.7709068271425606,
      "grad_norm": 2.473322215306534,
      "learning_rate": 1.515406606797763e-06,
      "loss": 0.3491,
      "step": 3715
    },
    {
      "epoch": 0.7719443868022411,
      "grad_norm": 2.2061318656296067,
      "learning_rate": 1.5024424388338682e-06,
      "loss": 0.3499,
      "step": 3720
    },
    {
      "epoch": 0.7729819464619215,
      "grad_norm": 2.435209027478781,
      "learning_rate": 1.4895241510703157e-06,
      "loss": 0.3643,
      "step": 3725
    },
    {
      "epoch": 0.774019506121602,
      "grad_norm": 2.369611296713981,
      "learning_rate": 1.4766519129663992e-06,
      "loss": 0.3483,
      "step": 3730
    },
    {
      "epoch": 0.7750570657812824,
      "grad_norm": 2.355834071175694,
      "learning_rate": 1.4638258933773425e-06,
      "loss": 0.3519,
      "step": 3735
    },
    {
      "epoch": 0.7760946254409629,
      "grad_norm": 2.469740057096309,
      "learning_rate": 1.451046260552086e-06,
      "loss": 0.3475,
      "step": 3740
    },
    {
      "epoch": 0.7771321851006433,
      "grad_norm": 2.4205796114015077,
      "learning_rate": 1.438313182131073e-06,
      "loss": 0.3523,
      "step": 3745
    },
    {
      "epoch": 0.7781697447603237,
      "grad_norm": 2.252004484682199,
      "learning_rate": 1.4256268251440631e-06,
      "loss": 0.3501,
      "step": 3750
    },
    {
      "epoch": 0.7792073044200042,
      "grad_norm": 2.3823736067008556,
      "learning_rate": 1.412987356007931e-06,
      "loss": 0.3439,
      "step": 3755
    },
    {
      "epoch": 0.7802448640796846,
      "grad_norm": 2.440498718677342,
      "learning_rate": 1.4003949405244888e-06,
      "loss": 0.3545,
      "step": 3760
    },
    {
      "epoch": 0.781282423739365,
      "grad_norm": 2.4557991100819905,
      "learning_rate": 1.3878497438783035e-06,
      "loss": 0.3476,
      "step": 3765
    },
    {
      "epoch": 0.7823199833990454,
      "grad_norm": 2.3768196794994956,
      "learning_rate": 1.3753519306345443e-06,
      "loss": 0.3566,
      "step": 3770
    },
    {
      "epoch": 0.7833575430587258,
      "grad_norm": 2.3763354198113427,
      "learning_rate": 1.3629016647368077e-06,
      "loss": 0.3508,
      "step": 3775
    },
    {
      "epoch": 0.7843951027184063,
      "grad_norm": 2.5905901892830725,
      "learning_rate": 1.3504991095049774e-06,
      "loss": 0.3499,
      "step": 3780
    },
    {
      "epoch": 0.7854326623780867,
      "grad_norm": 2.4256417735699407,
      "learning_rate": 1.338144427633079e-06,
      "loss": 0.3504,
      "step": 3785
    },
    {
      "epoch": 0.7864702220377672,
      "grad_norm": 2.4053687510971664,
      "learning_rate": 1.3258377811871481e-06,
      "loss": 0.3484,
      "step": 3790
    },
    {
      "epoch": 0.7875077816974476,
      "grad_norm": 2.4549151530337654,
      "learning_rate": 1.3135793316030958e-06,
      "loss": 0.3482,
      "step": 3795
    },
    {
      "epoch": 0.7885453413571281,
      "grad_norm": 2.4124091117521576,
      "learning_rate": 1.3013692396846028e-06,
      "loss": 0.3417,
      "step": 3800
    },
    {
      "epoch": 0.7895829010168085,
      "grad_norm": 2.492914244103795,
      "learning_rate": 1.2892076656010017e-06,
      "loss": 0.346,
      "step": 3805
    },
    {
      "epoch": 0.790620460676489,
      "grad_norm": 2.5012298456270807,
      "learning_rate": 1.277094768885182e-06,
      "loss": 0.3359,
      "step": 3810
    },
    {
      "epoch": 0.7916580203361693,
      "grad_norm": 2.3887971930046,
      "learning_rate": 1.2650307084314872e-06,
      "loss": 0.3434,
      "step": 3815
    },
    {
      "epoch": 0.7926955799958497,
      "grad_norm": 2.4930229591528605,
      "learning_rate": 1.2530156424936469e-06,
      "loss": 0.362,
      "step": 3820
    },
    {
      "epoch": 0.7937331396555302,
      "grad_norm": 2.386338375070925,
      "learning_rate": 1.241049728682684e-06,
      "loss": 0.3484,
      "step": 3825
    },
    {
      "epoch": 0.7947706993152106,
      "grad_norm": 2.303121828761511,
      "learning_rate": 1.229133123964853e-06,
      "loss": 0.355,
      "step": 3830
    },
    {
      "epoch": 0.795808258974891,
      "grad_norm": 2.418730350979906,
      "learning_rate": 1.2172659846595924e-06,
      "loss": 0.3534,
      "step": 3835
    },
    {
      "epoch": 0.7968458186345715,
      "grad_norm": 2.4002454150472143,
      "learning_rate": 1.2054484664374533e-06,
      "loss": 0.3484,
      "step": 3840
    },
    {
      "epoch": 0.7978833782942519,
      "grad_norm": 2.453421003455541,
      "learning_rate": 1.1936807243180743e-06,
      "loss": 0.3493,
      "step": 3845
    },
    {
      "epoch": 0.7989209379539324,
      "grad_norm": 2.4746199776766473,
      "learning_rate": 1.1819629126681398e-06,
      "loss": 0.3387,
      "step": 3850
    },
    {
      "epoch": 0.7999584976136128,
      "grad_norm": 2.38307853790052,
      "learning_rate": 1.1702951851993598e-06,
      "loss": 0.3483,
      "step": 3855
    },
    {
      "epoch": 0.8009960572732933,
      "grad_norm": 2.4103963437096114,
      "learning_rate": 1.1586776949664453e-06,
      "loss": 0.3423,
      "step": 3860
    },
    {
      "epoch": 0.8020336169329736,
      "grad_norm": 2.3551774825792924,
      "learning_rate": 1.1471105943651117e-06,
      "loss": 0.341,
      "step": 3865
    },
    {
      "epoch": 0.803071176592654,
      "grad_norm": 2.518707782804587,
      "learning_rate": 1.1355940351300715e-06,
      "loss": 0.349,
      "step": 3870
    },
    {
      "epoch": 0.8041087362523345,
      "grad_norm": 2.3344993783154964,
      "learning_rate": 1.1241281683330486e-06,
      "loss": 0.3519,
      "step": 3875
    },
    {
      "epoch": 0.8051462959120149,
      "grad_norm": 2.3344944133616665,
      "learning_rate": 1.1127131443807887e-06,
      "loss": 0.3412,
      "step": 3880
    },
    {
      "epoch": 0.8061838555716954,
      "grad_norm": 2.3019549791312754,
      "learning_rate": 1.1013491130131027e-06,
      "loss": 0.3346,
      "step": 3885
    },
    {
      "epoch": 0.8072214152313758,
      "grad_norm": 2.3790731919640415,
      "learning_rate": 1.0900362233008804e-06,
      "loss": 0.3414,
      "step": 3890
    },
    {
      "epoch": 0.8082589748910562,
      "grad_norm": 2.4317034804688453,
      "learning_rate": 1.0787746236441538e-06,
      "loss": 0.3521,
      "step": 3895
    },
    {
      "epoch": 0.8092965345507367,
      "grad_norm": 2.387356188471124,
      "learning_rate": 1.0675644617701402e-06,
      "loss": 0.3528,
      "step": 3900
    },
    {
      "epoch": 0.8103340942104171,
      "grad_norm": 2.53615626667828,
      "learning_rate": 1.0564058847313108e-06,
      "loss": 0.351,
      "step": 3905
    },
    {
      "epoch": 0.8113716538700976,
      "grad_norm": 2.4249799199654376,
      "learning_rate": 1.0452990389034507e-06,
      "loss": 0.3392,
      "step": 3910
    },
    {
      "epoch": 0.8124092135297779,
      "grad_norm": 2.3915606563087946,
      "learning_rate": 1.0342440699837537e-06,
      "loss": 0.3361,
      "step": 3915
    },
    {
      "epoch": 0.8134467731894583,
      "grad_norm": 2.5405240392917094,
      "learning_rate": 1.0232411229888994e-06,
      "loss": 0.3457,
      "step": 3920
    },
    {
      "epoch": 0.8144843328491388,
      "grad_norm": 2.424497940803967,
      "learning_rate": 1.0122903422531588e-06,
      "loss": 0.3424,
      "step": 3925
    },
    {
      "epoch": 0.8155218925088192,
      "grad_norm": 2.4383046197013187,
      "learning_rate": 1.001391871426492e-06,
      "loss": 0.3559,
      "step": 3930
    },
    {
      "epoch": 0.8165594521684997,
      "grad_norm": 2.3911832191085054,
      "learning_rate": 9.90545853472673e-07,
      "loss": 0.3429,
      "step": 3935
    },
    {
      "epoch": 0.8175970118281801,
      "grad_norm": 2.279983678354184,
      "learning_rate": 9.797524306674104e-07,
      "loss": 0.3437,
      "step": 3940
    },
    {
      "epoch": 0.8186345714878606,
      "grad_norm": 2.50140074824066,
      "learning_rate": 9.69011744596477e-07,
      "loss": 0.3402,
      "step": 3945
    },
    {
      "epoch": 0.819672131147541,
      "grad_norm": 2.4365828792120294,
      "learning_rate": 9.583239361538638e-07,
      "loss": 0.3417,
      "step": 3950
    },
    {
      "epoch": 0.8207096908072214,
      "grad_norm": 2.3579736935494955,
      "learning_rate": 9.476891455399168e-07,
      "loss": 0.3346,
      "step": 3955
    },
    {
      "epoch": 0.8217472504669019,
      "grad_norm": 2.5090769048416566,
      "learning_rate": 9.371075122595103e-07,
      "loss": 0.3231,
      "step": 3960
    },
    {
      "epoch": 0.8227848101265823,
      "grad_norm": 2.439152419655965,
      "learning_rate": 9.265791751202113e-07,
      "loss": 0.3437,
      "step": 3965
    },
    {
      "epoch": 0.8238223697862627,
      "grad_norm": 2.4534357310565214,
      "learning_rate": 9.161042722304609e-07,
      "loss": 0.3409,
      "step": 3970
    },
    {
      "epoch": 0.8248599294459431,
      "grad_norm": 2.2889519820080357,
      "learning_rate": 9.056829409977574e-07,
      "loss": 0.3423,
      "step": 3975
    },
    {
      "epoch": 0.8258974891056236,
      "grad_norm": 2.6079615773622544,
      "learning_rate": 8.953153181268609e-07,
      "loss": 0.3412,
      "step": 3980
    },
    {
      "epoch": 0.826935048765304,
      "grad_norm": 2.462659677967327,
      "learning_rate": 8.850015396179962e-07,
      "loss": 0.3399,
      "step": 3985
    },
    {
      "epoch": 0.8279726084249844,
      "grad_norm": 2.453876430295839,
      "learning_rate": 8.747417407650704e-07,
      "loss": 0.3368,
      "step": 3990
    },
    {
      "epoch": 0.8290101680846649,
      "grad_norm": 2.5260638258531722,
      "learning_rate": 8.645360561538935e-07,
      "loss": 0.3506,
      "step": 3995
    },
    {
      "epoch": 0.8300477277443453,
      "grad_norm": 2.392012536793574,
      "learning_rate": 8.543846196604239e-07,
      "loss": 0.3434,
      "step": 4000
    },
    {
      "epoch": 0.8310852874040258,
      "grad_norm": 2.4984903138745587,
      "learning_rate": 8.442875644489962e-07,
      "loss": 0.3351,
      "step": 4005
    },
    {
      "epoch": 0.8321228470637062,
      "grad_norm": 2.464203117108214,
      "learning_rate": 8.342450229705889e-07,
      "loss": 0.3455,
      "step": 4010
    },
    {
      "epoch": 0.8331604067233866,
      "grad_norm": 2.489210802914686,
      "learning_rate": 8.2425712696108e-07,
      "loss": 0.33,
      "step": 4015
    },
    {
      "epoch": 0.834197966383067,
      "grad_norm": 2.306914933986251,
      "learning_rate": 8.143240074395198e-07,
      "loss": 0.3418,
      "step": 4020
    },
    {
      "epoch": 0.8352355260427474,
      "grad_norm": 2.513591795617815,
      "learning_rate": 8.044457947064116e-07,
      "loss": 0.3418,
      "step": 4025
    },
    {
      "epoch": 0.8362730857024279,
      "grad_norm": 2.3537491795821524,
      "learning_rate": 7.946226183420047e-07,
      "loss": 0.3479,
      "step": 4030
    },
    {
      "epoch": 0.8373106453621083,
      "grad_norm": 2.422572559987635,
      "learning_rate": 7.848546072045932e-07,
      "loss": 0.3446,
      "step": 4035
    },
    {
      "epoch": 0.8383482050217888,
      "grad_norm": 2.3037996424039084,
      "learning_rate": 7.75141889428826e-07,
      "loss": 0.3257,
      "step": 4040
    },
    {
      "epoch": 0.8393857646814692,
      "grad_norm": 2.3570961713348186,
      "learning_rate": 7.654845924240228e-07,
      "loss": 0.3341,
      "step": 4045
    },
    {
      "epoch": 0.8404233243411496,
      "grad_norm": 2.3236115130562762,
      "learning_rate": 7.558828428725102e-07,
      "loss": 0.3328,
      "step": 4050
    },
    {
      "epoch": 0.8414608840008301,
      "grad_norm": 2.4888053313991603,
      "learning_rate": 7.463367667279515e-07,
      "loss": 0.3429,
      "step": 4055
    },
    {
      "epoch": 0.8424984436605105,
      "grad_norm": 2.4813791916370334,
      "learning_rate": 7.368464892137006e-07,
      "loss": 0.3412,
      "step": 4060
    },
    {
      "epoch": 0.843536003320191,
      "grad_norm": 2.3113527571110737,
      "learning_rate": 7.274121348211582e-07,
      "loss": 0.3475,
      "step": 4065
    },
    {
      "epoch": 0.8445735629798713,
      "grad_norm": 2.587556104836598,
      "learning_rate": 7.180338273081327e-07,
      "loss": 0.3354,
      "step": 4070
    },
    {
      "epoch": 0.8456111226395517,
      "grad_norm": 2.296420055640647,
      "learning_rate": 7.087116896972268e-07,
      "loss": 0.3357,
      "step": 4075
    },
    {
      "epoch": 0.8466486822992322,
      "grad_norm": 2.3266676608899663,
      "learning_rate": 6.994458442742163e-07,
      "loss": 0.3362,
      "step": 4080
    },
    {
      "epoch": 0.8476862419589126,
      "grad_norm": 2.6045589584755984,
      "learning_rate": 6.902364125864496e-07,
      "loss": 0.34,
      "step": 4085
    },
    {
      "epoch": 0.8487238016185931,
      "grad_norm": 2.352378817551258,
      "learning_rate": 6.810835154412487e-07,
      "loss": 0.339,
      "step": 4090
    },
    {
      "epoch": 0.8497613612782735,
      "grad_norm": 2.3547116125315446,
      "learning_rate": 6.719872729043331e-07,
      "loss": 0.3378,
      "step": 4095
    },
    {
      "epoch": 0.850798920937954,
      "grad_norm": 2.321427219175219,
      "learning_rate": 6.629478042982346e-07,
      "loss": 0.3229,
      "step": 4100
    },
    {
      "epoch": 0.8518364805976344,
      "grad_norm": 2.521475742171532,
      "learning_rate": 6.539652282007386e-07,
      "loss": 0.3376,
      "step": 4105
    },
    {
      "epoch": 0.8528740402573148,
      "grad_norm": 2.3926877720494573,
      "learning_rate": 6.450396624433286e-07,
      "loss": 0.3325,
      "step": 4110
    },
    {
      "epoch": 0.8539115999169953,
      "grad_norm": 2.425081699530235,
      "learning_rate": 6.361712241096374e-07,
      "loss": 0.3314,
      "step": 4115
    },
    {
      "epoch": 0.8549491595766756,
      "grad_norm": 2.487314370898736,
      "learning_rate": 6.273600295339111e-07,
      "loss": 0.3352,
      "step": 4120
    },
    {
      "epoch": 0.855986719236356,
      "grad_norm": 2.4556339075670723,
      "learning_rate": 6.186061942994864e-07,
      "loss": 0.3338,
      "step": 4125
    },
    {
      "epoch": 0.8570242788960365,
      "grad_norm": 2.417547682627604,
      "learning_rate": 6.099098332372733e-07,
      "loss": 0.3299,
      "step": 4130
    },
    {
      "epoch": 0.8580618385557169,
      "grad_norm": 2.611592074254707,
      "learning_rate": 6.012710604242478e-07,
      "loss": 0.3331,
      "step": 4135
    },
    {
      "epoch": 0.8590993982153974,
      "grad_norm": 2.448052874786836,
      "learning_rate": 5.926899891819521e-07,
      "loss": 0.3376,
      "step": 4140
    },
    {
      "epoch": 0.8601369578750778,
      "grad_norm": 2.475864292987021,
      "learning_rate": 5.841667320750188e-07,
      "loss": 0.3437,
      "step": 4145
    },
    {
      "epoch": 0.8611745175347583,
      "grad_norm": 2.373719046516931,
      "learning_rate": 5.757014009096801e-07,
      "loss": 0.3357,
      "step": 4150
    },
    {
      "epoch": 0.8622120771944387,
      "grad_norm": 2.3162935725557032,
      "learning_rate": 5.672941067323124e-07,
      "loss": 0.3305,
      "step": 4155
    },
    {
      "epoch": 0.8632496368541192,
      "grad_norm": 2.3285428336849736,
      "learning_rate": 5.589449598279762e-07,
      "loss": 0.3278,
      "step": 4160
    },
    {
      "epoch": 0.8642871965137996,
      "grad_norm": 2.6975663812633823,
      "learning_rate": 5.506540697189638e-07,
      "loss": 0.3235,
      "step": 4165
    },
    {
      "epoch": 0.8653247561734799,
      "grad_norm": 2.385391836897359,
      "learning_rate": 5.424215451633719e-07,
      "loss": 0.3366,
      "step": 4170
    },
    {
      "epoch": 0.8663623158331604,
      "grad_norm": 2.509848709680898,
      "learning_rate": 5.342474941536701e-07,
      "loss": 0.3349,
      "step": 4175
    },
    {
      "epoch": 0.8673998754928408,
      "grad_norm": 2.4992507740450063,
      "learning_rate": 5.261320239152851e-07,
      "loss": 0.3248,
      "step": 4180
    },
    {
      "epoch": 0.8684374351525213,
      "grad_norm": 2.4274228991197826,
      "learning_rate": 5.180752409051892e-07,
      "loss": 0.3276,
      "step": 4185
    },
    {
      "epoch": 0.8694749948122017,
      "grad_norm": 2.3127395460474975,
      "learning_rate": 5.100772508105139e-07,
      "loss": 0.318,
      "step": 4190
    },
    {
      "epoch": 0.8705125544718821,
      "grad_norm": 2.468734058758486,
      "learning_rate": 5.021381585471563e-07,
      "loss": 0.3332,
      "step": 4195
    },
    {
      "epoch": 0.8715501141315626,
      "grad_norm": 2.4531926615746564,
      "learning_rate": 4.942580682584041e-07,
      "loss": 0.3286,
      "step": 4200
    },
    {
      "epoch": 0.872587673791243,
      "grad_norm": 2.377223115961875,
      "learning_rate": 4.864370833135673e-07,
      "loss": 0.3342,
      "step": 4205
    },
    {
      "epoch": 0.8736252334509235,
      "grad_norm": 2.449018104688608,
      "learning_rate": 4.786753063066318e-07,
      "loss": 0.3355,
      "step": 4210
    },
    {
      "epoch": 0.8746627931106039,
      "grad_norm": 2.4173614341699685,
      "learning_rate": 4.7097283905489956e-07,
      "loss": 0.3288,
      "step": 4215
    },
    {
      "epoch": 0.8757003527702842,
      "grad_norm": 2.603118901855438,
      "learning_rate": 4.633297825976635e-07,
      "loss": 0.3371,
      "step": 4220
    },
    {
      "epoch": 0.8767379124299647,
      "grad_norm": 2.4773560425700727,
      "learning_rate": 4.5574623719487787e-07,
      "loss": 0.3256,
      "step": 4225
    },
    {
      "epoch": 0.8777754720896451,
      "grad_norm": 2.4564917228085053,
      "learning_rate": 4.482223023258453e-07,
      "loss": 0.337,
      "step": 4230
    },
    {
      "epoch": 0.8788130317493256,
      "grad_norm": 2.4727605694693553,
      "learning_rate": 4.407580766879066e-07,
      "loss": 0.3221,
      "step": 4235
    },
    {
      "epoch": 0.879850591409006,
      "grad_norm": 2.593564549202061,
      "learning_rate": 4.333536581951542e-07,
      "loss": 0.3364,
      "step": 4240
    },
    {
      "epoch": 0.8808881510686865,
      "grad_norm": 2.3065543830266204,
      "learning_rate": 4.2600914397714023e-07,
      "loss": 0.3266,
      "step": 4245
    },
    {
      "epoch": 0.8819257107283669,
      "grad_norm": 2.3918281813023783,
      "learning_rate": 4.1872463037760823e-07,
      "loss": 0.3311,
      "step": 4250
    },
    {
      "epoch": 0.8829632703880473,
      "grad_norm": 2.5113381009120417,
      "learning_rate": 4.1150021295322306e-07,
      "loss": 0.3373,
      "step": 4255
    },
    {
      "epoch": 0.8840008300477278,
      "grad_norm": 2.4230720157484935,
      "learning_rate": 4.043359864723262e-07,
      "loss": 0.3329,
      "step": 4260
    },
    {
      "epoch": 0.8850383897074082,
      "grad_norm": 2.616499204627849,
      "learning_rate": 3.972320449136829e-07,
      "loss": 0.3295,
      "step": 4265
    },
    {
      "epoch": 0.8860759493670886,
      "grad_norm": 2.5866660581654677,
      "learning_rate": 3.90188481465254e-07,
      "loss": 0.3276,
      "step": 4270
    },
    {
      "epoch": 0.887113509026769,
      "grad_norm": 2.4795883896583364,
      "learning_rate": 3.8320538852297694e-07,
      "loss": 0.3339,
      "step": 4275
    },
    {
      "epoch": 0.8881510686864494,
      "grad_norm": 2.3898148517858244,
      "learning_rate": 3.762828576895472e-07,
      "loss": 0.3373,
      "step": 4280
    },
    {
      "epoch": 0.8891886283461299,
      "grad_norm": 2.4124000504118044,
      "learning_rate": 3.694209797732201e-07,
      "loss": 0.332,
      "step": 4285
    },
    {
      "epoch": 0.8902261880058103,
      "grad_norm": 2.724223147572049,
      "learning_rate": 3.6261984478662025e-07,
      "loss": 0.3417,
      "step": 4290
    },
    {
      "epoch": 0.8912637476654908,
      "grad_norm": 2.4453049065554278,
      "learning_rate": 3.558795419455596e-07,
      "loss": 0.3198,
      "step": 4295
    },
    {
      "epoch": 0.8923013073251712,
      "grad_norm": 2.3658613859590165,
      "learning_rate": 3.492001596678651e-07,
      "loss": 0.3248,
      "step": 4300
    },
    {
      "epoch": 0.8933388669848517,
      "grad_norm": 2.415791137686327,
      "learning_rate": 3.4258178557222354e-07,
      "loss": 0.3238,
      "step": 4305
    },
    {
      "epoch": 0.8943764266445321,
      "grad_norm": 2.2801295759850566,
      "learning_rate": 3.3602450647702847e-07,
      "loss": 0.3317,
      "step": 4310
    },
    {
      "epoch": 0.8954139863042125,
      "grad_norm": 2.532814151310614,
      "learning_rate": 3.295284083992434e-07,
      "loss": 0.3217,
      "step": 4315
    },
    {
      "epoch": 0.8964515459638929,
      "grad_norm": 2.5019562185543993,
      "learning_rate": 3.2309357655326945e-07,
      "loss": 0.333,
      "step": 4320
    },
    {
      "epoch": 0.8974891056235733,
      "grad_norm": 2.2943698094322955,
      "learning_rate": 3.167200953498367e-07,
      "loss": 0.3181,
      "step": 4325
    },
    {
      "epoch": 0.8985266652832538,
      "grad_norm": 2.405841184133046,
      "learning_rate": 3.1040804839488406e-07,
      "loss": 0.3339,
      "step": 4330
    },
    {
      "epoch": 0.8995642249429342,
      "grad_norm": 2.386316529188201,
      "learning_rate": 3.041575184884732e-07,
      "loss": 0.3341,
      "step": 4335
    },
    {
      "epoch": 0.9006017846026146,
      "grad_norm": 2.517011507851569,
      "learning_rate": 2.979685876236982e-07,
      "loss": 0.3346,
      "step": 4340
    },
    {
      "epoch": 0.9016393442622951,
      "grad_norm": 2.5508437593952094,
      "learning_rate": 2.918413369856105e-07,
      "loss": 0.3258,
      "step": 4345
    },
    {
      "epoch": 0.9026769039219755,
      "grad_norm": 2.457005114059522,
      "learning_rate": 2.857758469501509e-07,
      "loss": 0.3197,
      "step": 4350
    },
    {
      "epoch": 0.903714463581656,
      "grad_norm": 2.3964711097696836,
      "learning_rate": 2.7977219708310134e-07,
      "loss": 0.3288,
      "step": 4355
    },
    {
      "epoch": 0.9047520232413364,
      "grad_norm": 2.4092985201196084,
      "learning_rate": 2.7383046613903676e-07,
      "loss": 0.325,
      "step": 4360
    },
    {
      "epoch": 0.9057895829010169,
      "grad_norm": 2.519420182065836,
      "learning_rate": 2.679507320602931e-07,
      "loss": 0.325,
      "step": 4365
    },
    {
      "epoch": 0.9068271425606972,
      "grad_norm": 2.4612555403219214,
      "learning_rate": 2.6213307197594353e-07,
      "loss": 0.3261,
      "step": 4370
    },
    {
      "epoch": 0.9078647022203776,
      "grad_norm": 2.568118436273184,
      "learning_rate": 2.5637756220079135e-07,
      "loss": 0.3246,
      "step": 4375
    },
    {
      "epoch": 0.9089022618800581,
      "grad_norm": 2.5397440534691373,
      "learning_rate": 2.506842782343627e-07,
      "loss": 0.3162,
      "step": 4380
    },
    {
      "epoch": 0.9099398215397385,
      "grad_norm": 2.4925975703816245,
      "learning_rate": 2.4505329475991823e-07,
      "loss": 0.331,
      "step": 4385
    },
    {
      "epoch": 0.910977381199419,
      "grad_norm": 3.3523863369544986,
      "learning_rate": 2.3948468564347904e-07,
      "loss": 0.3277,
      "step": 4390
    },
    {
      "epoch": 0.9120149408590994,
      "grad_norm": 2.575554252160631,
      "learning_rate": 2.3397852393284792e-07,
      "loss": 0.3404,
      "step": 4395
    },
    {
      "epoch": 0.9130525005187798,
      "grad_norm": 2.481671747693274,
      "learning_rate": 2.2853488185665796e-07,
      "loss": 0.3206,
      "step": 4400
    },
    {
      "epoch": 0.9140900601784603,
      "grad_norm": 2.4222594768146415,
      "learning_rate": 2.231538308234249e-07,
      "loss": 0.3216,
      "step": 4405
    },
    {
      "epoch": 0.9151276198381407,
      "grad_norm": 2.579114792911807,
      "learning_rate": 2.178354414206063e-07,
      "loss": 0.3267,
      "step": 4410
    },
    {
      "epoch": 0.9161651794978212,
      "grad_norm": 2.472569881694958,
      "learning_rate": 2.125797834136789e-07,
      "loss": 0.3212,
      "step": 4415
    },
    {
      "epoch": 0.9172027391575015,
      "grad_norm": 2.4091565145640605,
      "learning_rate": 2.0738692574522324e-07,
      "loss": 0.3246,
      "step": 4420
    },
    {
      "epoch": 0.9182402988171819,
      "grad_norm": 2.4603025192987644,
      "learning_rate": 2.0225693653401824e-07,
      "loss": 0.3221,
      "step": 4425
    },
    {
      "epoch": 0.9192778584768624,
      "grad_norm": 2.536120994058227,
      "learning_rate": 1.9718988307414866e-07,
      "loss": 0.3289,
      "step": 4430
    },
    {
      "epoch": 0.9203154181365428,
      "grad_norm": 2.423050703239737,
      "learning_rate": 1.921858318341191e-07,
      "loss": 0.3258,
      "step": 4435
    },
    {
      "epoch": 0.9213529777962233,
      "grad_norm": 2.429414838855423,
      "learning_rate": 1.8724484845598855e-07,
      "loss": 0.3196,
      "step": 4440
    },
    {
      "epoch": 0.9223905374559037,
      "grad_norm": 2.7662680215892004,
      "learning_rate": 1.8236699775450338e-07,
      "loss": 0.3258,
      "step": 4445
    },
    {
      "epoch": 0.9234280971155842,
      "grad_norm": 2.45993148237687,
      "learning_rate": 1.7755234371624908e-07,
      "loss": 0.318,
      "step": 4450
    },
    {
      "epoch": 0.9244656567752646,
      "grad_norm": 2.4765798027549644,
      "learning_rate": 1.7280094949881144e-07,
      "loss": 0.3287,
      "step": 4455
    },
    {
      "epoch": 0.925503216434945,
      "grad_norm": 2.5990110307214027,
      "learning_rate": 1.6811287742994897e-07,
      "loss": 0.3203,
      "step": 4460
    },
    {
      "epoch": 0.9265407760946255,
      "grad_norm": 2.7248202388018563,
      "learning_rate": 1.6348818900677077e-07,
      "loss": 0.3196,
      "step": 4465
    },
    {
      "epoch": 0.9275783357543059,
      "grad_norm": 2.360215127100717,
      "learning_rate": 1.5892694489493598e-07,
      "loss": 0.3266,
      "step": 4470
    },
    {
      "epoch": 0.9286158954139863,
      "grad_norm": 2.3701662902928415,
      "learning_rate": 1.5442920492785396e-07,
      "loss": 0.3217,
      "step": 4475
    },
    {
      "epoch": 0.9296534550736667,
      "grad_norm": 2.5408968036091593,
      "learning_rate": 1.4999502810590094e-07,
      "loss": 0.3147,
      "step": 4480
    },
    {
      "epoch": 0.9306910147333471,
      "grad_norm": 2.3872752954425276,
      "learning_rate": 1.456244725956446e-07,
      "loss": 0.3314,
      "step": 4485
    },
    {
      "epoch": 0.9317285743930276,
      "grad_norm": 2.510401039927896,
      "learning_rate": 1.4131759572908354e-07,
      "loss": 0.3233,
      "step": 4490
    },
    {
      "epoch": 0.932766134052708,
      "grad_norm": 2.452160220911644,
      "learning_rate": 1.370744540028929e-07,
      "loss": 0.3216,
      "step": 4495
    },
    {
      "epoch": 0.9338036937123885,
      "grad_norm": 2.5517139832044826,
      "learning_rate": 1.328951030776826e-07,
      "loss": 0.322,
      "step": 4500
    },
    {
      "epoch": 0.9348412533720689,
      "grad_norm": 2.600821629157991,
      "learning_rate": 1.2877959777727212e-07,
      "loss": 0.328,
      "step": 4505
    },
    {
      "epoch": 0.9358788130317494,
      "grad_norm": 2.4402893254633207,
      "learning_rate": 1.2472799208796517e-07,
      "loss": 0.3352,
      "step": 4510
    },
    {
      "epoch": 0.9369163726914298,
      "grad_norm": 2.4890223515899748,
      "learning_rate": 1.2074033915784543e-07,
      "loss": 0.3273,
      "step": 4515
    },
    {
      "epoch": 0.9379539323511102,
      "grad_norm": 2.541162954952577,
      "learning_rate": 1.168166912960772e-07,
      "loss": 0.3198,
      "step": 4520
    },
    {
      "epoch": 0.9389914920107906,
      "grad_norm": 2.510264694405438,
      "learning_rate": 1.1295709997222182e-07,
      "loss": 0.3278,
      "step": 4525
    },
    {
      "epoch": 0.940029051670471,
      "grad_norm": 2.4830766706425895,
      "learning_rate": 1.0916161581555895e-07,
      "loss": 0.3153,
      "step": 4530
    },
    {
      "epoch": 0.9410666113301515,
      "grad_norm": 2.3227486456210222,
      "learning_rate": 1.0543028861442539e-07,
      "loss": 0.3177,
      "step": 4535
    },
    {
      "epoch": 0.9421041709898319,
      "grad_norm": 2.5419821606370174,
      "learning_rate": 1.0176316731556112e-07,
      "loss": 0.3322,
      "step": 4540
    },
    {
      "epoch": 0.9431417306495123,
      "grad_norm": 2.6673497102822536,
      "learning_rate": 9.816030002346766e-08,
      "loss": 0.3319,
      "step": 4545
    },
    {
      "epoch": 0.9441792903091928,
      "grad_norm": 2.3937271331548904,
      "learning_rate": 9.462173399977348e-08,
      "loss": 0.3267,
      "step": 4550
    },
    {
      "epoch": 0.9452168499688732,
      "grad_norm": 2.3407051661833123,
      "learning_rate": 9.11475156626207e-08,
      "loss": 0.3103,
      "step": 4555
    },
    {
      "epoch": 0.9462544096285537,
      "grad_norm": 2.5377283482688755,
      "learning_rate": 8.773769058605053e-08,
      "loss": 0.3356,
      "step": 4560
    },
    {
      "epoch": 0.9472919692882341,
      "grad_norm": 2.3498925102259904,
      "learning_rate": 8.439230349940708e-08,
      "loss": 0.3254,
      "step": 4565
    },
    {
      "epoch": 0.9483295289479146,
      "grad_norm": 2.3416477533349145,
      "learning_rate": 8.111139828675175e-08,
      "loss": 0.3293,
      "step": 4570
    },
    {
      "epoch": 0.9493670886075949,
      "grad_norm": 2.559009235835814,
      "learning_rate": 7.78950179862864e-08,
      "loss": 0.3185,
      "step": 4575
    },
    {
      "epoch": 0.9504046482672753,
      "grad_norm": 2.586458213825969,
      "learning_rate": 7.474320478978946e-08,
      "loss": 0.3246,
      "step": 4580
    },
    {
      "epoch": 0.9514422079269558,
      "grad_norm": 2.3485714776360407,
      "learning_rate": 7.16560000420613e-08,
      "loss": 0.3238,
      "step": 4585
    },
    {
      "epoch": 0.9524797675866362,
      "grad_norm": 2.369694918601893,
      "learning_rate": 6.863344424038354e-08,
      "loss": 0.3103,
      "step": 4590
    },
    {
      "epoch": 0.9535173272463167,
      "grad_norm": 2.4252802076691538,
      "learning_rate": 6.567557703398675e-08,
      "loss": 0.3273,
      "step": 4595
    },
    {
      "epoch": 0.9545548869059971,
      "grad_norm": 2.552882664412225,
      "learning_rate": 6.278243722352973e-08,
      "loss": 0.3182,
      "step": 4600
    },
    {
      "epoch": 0.9555924465656775,
      "grad_norm": 2.5353013413397507,
      "learning_rate": 5.995406276059267e-08,
      "loss": 0.3335,
      "step": 4605
    },
    {
      "epoch": 0.956630006225358,
      "grad_norm": 2.4812443315722934,
      "learning_rate": 5.719049074717764e-08,
      "loss": 0.3191,
      "step": 4610
    },
    {
      "epoch": 0.9576675658850384,
      "grad_norm": 2.427622171598608,
      "learning_rate": 5.4491757435220505e-08,
      "loss": 0.3236,
      "step": 4615
    },
    {
      "epoch": 0.9587051255447189,
      "grad_norm": 2.601332407638946,
      "learning_rate": 5.185789822612086e-08,
      "loss": 0.3335,
      "step": 4620
    },
    {
      "epoch": 0.9597426852043992,
      "grad_norm": 2.4380419721098976,
      "learning_rate": 4.9288947670270146e-08,
      "loss": 0.3193,
      "step": 4625
    },
    {
      "epoch": 0.9607802448640796,
      "grad_norm": 2.36715515753985,
      "learning_rate": 4.678493946660423e-08,
      "loss": 0.3228,
      "step": 4630
    },
    {
      "epoch": 0.9618178045237601,
      "grad_norm": 2.4303066509314677,
      "learning_rate": 4.434590646215819e-08,
      "loss": 0.3266,
      "step": 4635
    },
    {
      "epoch": 0.9628553641834405,
      "grad_norm": 2.865610610806469,
      "learning_rate": 4.1971880651638376e-08,
      "loss": 0.3247,
      "step": 4640
    },
    {
      "epoch": 0.963892923843121,
      "grad_norm": 2.5407706997414268,
      "learning_rate": 3.966289317699878e-08,
      "loss": 0.3288,
      "step": 4645
    },
    {
      "epoch": 0.9649304835028014,
      "grad_norm": 2.4204826042021437,
      "learning_rate": 3.74189743270359e-08,
      "loss": 0.3145,
      "step": 4650
    },
    {
      "epoch": 0.9659680431624819,
      "grad_norm": 2.2543367363310716,
      "learning_rate": 3.5240153536988954e-08,
      "loss": 0.3307,
      "step": 4655
    },
    {
      "epoch": 0.9670056028221623,
      "grad_norm": 2.3642879401404677,
      "learning_rate": 3.312645938815695e-08,
      "loss": 0.3111,
      "step": 4660
    },
    {
      "epoch": 0.9680431624818427,
      "grad_norm": 2.551087158396502,
      "learning_rate": 3.107791960752005e-08,
      "loss": 0.3202,
      "step": 4665
    },
    {
      "epoch": 0.9690807221415232,
      "grad_norm": 2.431842119612971,
      "learning_rate": 2.909456106737818e-08,
      "loss": 0.3271,
      "step": 4670
    },
    {
      "epoch": 0.9701182818012035,
      "grad_norm": 2.469650411851549,
      "learning_rate": 2.7176409784998027e-08,
      "loss": 0.3225,
      "step": 4675
    },
    {
      "epoch": 0.971155841460884,
      "grad_norm": 2.6570472493964337,
      "learning_rate": 2.5323490922271044e-08,
      "loss": 0.3233,
      "step": 4680
    },
    {
      "epoch": 0.9721934011205644,
      "grad_norm": 2.792428588323876,
      "learning_rate": 2.3535828785384296e-08,
      "loss": 0.3263,
      "step": 4685
    },
    {
      "epoch": 0.9732309607802448,
      "grad_norm": 2.3558732345665456,
      "learning_rate": 2.1813446824502372e-08,
      "loss": 0.3136,
      "step": 4690
    },
    {
      "epoch": 0.9742685204399253,
      "grad_norm": 2.356877231683961,
      "learning_rate": 2.0156367633455965e-08,
      "loss": 0.3239,
      "step": 4695
    },
    {
      "epoch": 0.9753060800996057,
      "grad_norm": 2.4529847227301818,
      "learning_rate": 1.8564612949451555e-08,
      "loss": 0.3168,
      "step": 4700
    },
    {
      "epoch": 0.9763436397592862,
      "grad_norm": 2.382662217743029,
      "learning_rate": 1.7038203652781083e-08,
      "loss": 0.3234,
      "step": 4705
    },
    {
      "epoch": 0.9773811994189666,
      "grad_norm": 2.333476690943818,
      "learning_rate": 1.5577159766548832e-08,
      "loss": 0.3274,
      "step": 4710
    },
    {
      "epoch": 0.9784187590786471,
      "grad_norm": 2.402141955867801,
      "learning_rate": 1.4181500456412755e-08,
      "loss": 0.3215,
      "step": 4715
    },
    {
      "epoch": 0.9794563187383275,
      "grad_norm": 2.501308170912781,
      "learning_rate": 1.2851244030328004e-08,
      "loss": 0.3212,
      "step": 4720
    },
    {
      "epoch": 0.9804938783980078,
      "grad_norm": 2.6177200025512626,
      "learning_rate": 1.1586407938308785e-08,
      "loss": 0.3136,
      "step": 4725
    },
    {
      "epoch": 0.9815314380576883,
      "grad_norm": 2.548386853689889,
      "learning_rate": 1.0387008772200779e-08,
      "loss": 0.342,
      "step": 4730
    },
    {
      "epoch": 0.9825689977173687,
      "grad_norm": 2.552595538532663,
      "learning_rate": 9.253062265461855e-09,
      "loss": 0.3234,
      "step": 4735
    },
    {
      "epoch": 0.9836065573770492,
      "grad_norm": 2.489168583450402,
      "learning_rate": 8.184583292955572e-09,
      "loss": 0.318,
      "step": 4740
    },
    {
      "epoch": 0.9846441170367296,
      "grad_norm": 2.380606466941288,
      "learning_rate": 7.181585870757457e-09,
      "loss": 0.326,
      "step": 4745
    },
    {
      "epoch": 0.98568167669641,
      "grad_norm": 2.4173383787142297,
      "learning_rate": 6.2440831559690275e-09,
      "loss": 0.3185,
      "step": 4750
    },
    {
      "epoch": 0.9867192363560905,
      "grad_norm": 2.542878596032387,
      "learning_rate": 5.372087446547935e-09,
      "loss": 0.3252,
      "step": 4755
    },
    {
      "epoch": 0.9877567960157709,
      "grad_norm": 2.4745977048628918,
      "learning_rate": 4.565610181144209e-09,
      "loss": 0.3304,
      "step": 4760
    },
    {
      "epoch": 0.9887943556754514,
      "grad_norm": 2.4052051630616895,
      "learning_rate": 3.824661938951479e-09,
      "loss": 0.3243,
      "step": 4765
    },
    {
      "epoch": 0.9898319153351318,
      "grad_norm": 2.6094698060669264,
      "learning_rate": 3.1492524395682065e-09,
      "loss": 0.3189,
      "step": 4770
    },
    {
      "epoch": 0.9908694749948121,
      "grad_norm": 2.3476458011761134,
      "learning_rate": 2.5393905428688913e-09,
      "loss": 0.3231,
      "step": 4775
    },
    {
      "epoch": 0.9919070346544926,
      "grad_norm": 2.3445329353887456,
      "learning_rate": 1.9950842488891674e-09,
      "loss": 0.3355,
      "step": 4780
    },
    {
      "epoch": 0.992944594314173,
      "grad_norm": 2.4693805125337094,
      "learning_rate": 1.5163406977219963e-09,
      "loss": 0.3256,
      "step": 4785
    },
    {
      "epoch": 0.9939821539738535,
      "grad_norm": 2.563629879915604,
      "learning_rate": 1.103166169420522e-09,
      "loss": 0.3254,
      "step": 4790
    },
    {
      "epoch": 0.9950197136335339,
      "grad_norm": 2.494757369109791,
      "learning_rate": 7.555660839181356e-10,
      "loss": 0.3088,
      "step": 4795
    },
    {
      "epoch": 0.9960572732932144,
      "grad_norm": 2.373070173051403,
      "learning_rate": 4.735450009579756e-10,
      "loss": 0.323,
      "step": 4800
    },
    {
      "epoch": 0.9970948329528948,
      "grad_norm": 2.397230200383968,
      "learning_rate": 2.571066200307559e-10,
      "loss": 0.3303,
      "step": 4805
    },
    {
      "epoch": 0.9981323926125752,
      "grad_norm": 2.6318567902913985,
      "learning_rate": 1.0625378032813604e-10,
      "loss": 0.3156,
      "step": 4810
    },
    {
      "epoch": 0.9991699522722557,
      "grad_norm": 2.3100958526551696,
      "learning_rate": 2.0988460705528846e-11,
      "loss": 0.3192,
      "step": 4815
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.2939795255661011,
      "eval_runtime": 0.9809,
      "eval_samples_per_second": 3.058,
      "eval_steps_per_second": 1.019,
      "step": 4819
    },
    {
      "epoch": 1.0,
      "step": 4819,
      "total_flos": 504500280360960.0,
      "train_loss": 0.5211530197307415,
      "train_runtime": 55738.8815,
      "train_samples_per_second": 1.383,
      "train_steps_per_second": 0.086
    }
  ],
  "logging_steps": 5,
  "max_steps": 4819,
  "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": 504500280360960.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}