{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 2415,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00041407867494824016,
      "grad_norm": 23.762958025912543,
      "learning_rate": 4.132231404958678e-08,
      "loss": 1.4082,
      "step": 1
    },
    {
      "epoch": 0.002070393374741201,
      "grad_norm": 25.018778947453203,
      "learning_rate": 2.066115702479339e-07,
      "loss": 1.3951,
      "step": 5
    },
    {
      "epoch": 0.004140786749482402,
      "grad_norm": 16.523853816342054,
      "learning_rate": 4.132231404958678e-07,
      "loss": 1.3708,
      "step": 10
    },
    {
      "epoch": 0.006211180124223602,
      "grad_norm": 8.683918510891152,
      "learning_rate": 6.198347107438018e-07,
      "loss": 1.266,
      "step": 15
    },
    {
      "epoch": 0.008281573498964804,
      "grad_norm": 10.386484958466726,
      "learning_rate": 8.264462809917356e-07,
      "loss": 1.1323,
      "step": 20
    },
    {
      "epoch": 0.010351966873706004,
      "grad_norm": 4.605093604151904,
      "learning_rate": 1.0330578512396695e-06,
      "loss": 1.0154,
      "step": 25
    },
    {
      "epoch": 0.012422360248447204,
      "grad_norm": 3.5661937624270155,
      "learning_rate": 1.2396694214876035e-06,
      "loss": 0.9734,
      "step": 30
    },
    {
      "epoch": 0.014492753623188406,
      "grad_norm": 3.716902571537753,
      "learning_rate": 1.4462809917355372e-06,
      "loss": 0.9487,
      "step": 35
    },
    {
      "epoch": 0.016563146997929608,
      "grad_norm": 3.261636438781403,
      "learning_rate": 1.6528925619834712e-06,
      "loss": 0.92,
      "step": 40
    },
    {
      "epoch": 0.018633540372670808,
      "grad_norm": 2.976830393791539,
      "learning_rate": 1.859504132231405e-06,
      "loss": 0.9264,
      "step": 45
    },
    {
      "epoch": 0.020703933747412008,
      "grad_norm": 2.9783655232995643,
      "learning_rate": 2.066115702479339e-06,
      "loss": 0.9021,
      "step": 50
    },
    {
      "epoch": 0.022774327122153208,
      "grad_norm": 2.962089426499117,
      "learning_rate": 2.2727272727272728e-06,
      "loss": 0.8956,
      "step": 55
    },
    {
      "epoch": 0.024844720496894408,
      "grad_norm": 3.017107188860632,
      "learning_rate": 2.479338842975207e-06,
      "loss": 0.8896,
      "step": 60
    },
    {
      "epoch": 0.026915113871635612,
      "grad_norm": 3.0504869217916166,
      "learning_rate": 2.6859504132231405e-06,
      "loss": 0.8732,
      "step": 65
    },
    {
      "epoch": 0.028985507246376812,
      "grad_norm": 3.182023016740649,
      "learning_rate": 2.8925619834710743e-06,
      "loss": 0.8865,
      "step": 70
    },
    {
      "epoch": 0.031055900621118012,
      "grad_norm": 3.0451954568787736,
      "learning_rate": 3.0991735537190086e-06,
      "loss": 0.8779,
      "step": 75
    },
    {
      "epoch": 0.033126293995859216,
      "grad_norm": 3.266623033949086,
      "learning_rate": 3.3057851239669424e-06,
      "loss": 0.8669,
      "step": 80
    },
    {
      "epoch": 0.035196687370600416,
      "grad_norm": 3.118483445473265,
      "learning_rate": 3.5123966942148763e-06,
      "loss": 0.8636,
      "step": 85
    },
    {
      "epoch": 0.037267080745341616,
      "grad_norm": 3.0271091608802077,
      "learning_rate": 3.71900826446281e-06,
      "loss": 0.8765,
      "step": 90
    },
    {
      "epoch": 0.039337474120082816,
      "grad_norm": 3.0569603826532608,
      "learning_rate": 3.925619834710744e-06,
      "loss": 0.8484,
      "step": 95
    },
    {
      "epoch": 0.041407867494824016,
      "grad_norm": 3.11869623065633,
      "learning_rate": 4.132231404958678e-06,
      "loss": 0.8534,
      "step": 100
    },
    {
      "epoch": 0.043478260869565216,
      "grad_norm": 2.99796733558617,
      "learning_rate": 4.338842975206612e-06,
      "loss": 0.8659,
      "step": 105
    },
    {
      "epoch": 0.045548654244306416,
      "grad_norm": 3.0225524698417283,
      "learning_rate": 4.5454545454545455e-06,
      "loss": 0.8501,
      "step": 110
    },
    {
      "epoch": 0.047619047619047616,
      "grad_norm": 3.273049033434562,
      "learning_rate": 4.75206611570248e-06,
      "loss": 0.8557,
      "step": 115
    },
    {
      "epoch": 0.049689440993788817,
      "grad_norm": 3.2356902496362414,
      "learning_rate": 4.958677685950414e-06,
      "loss": 0.8571,
      "step": 120
    },
    {
      "epoch": 0.051759834368530024,
      "grad_norm": 3.1667154521169176,
      "learning_rate": 5.165289256198347e-06,
      "loss": 0.8458,
      "step": 125
    },
    {
      "epoch": 0.053830227743271224,
      "grad_norm": 3.124334948193542,
      "learning_rate": 5.371900826446281e-06,
      "loss": 0.8393,
      "step": 130
    },
    {
      "epoch": 0.055900621118012424,
      "grad_norm": 3.300947239536423,
      "learning_rate": 5.578512396694216e-06,
      "loss": 0.8231,
      "step": 135
    },
    {
      "epoch": 0.057971014492753624,
      "grad_norm": 3.028821224434978,
      "learning_rate": 5.785123966942149e-06,
      "loss": 0.8189,
      "step": 140
    },
    {
      "epoch": 0.060041407867494824,
      "grad_norm": 3.0713166387468163,
      "learning_rate": 5.991735537190083e-06,
      "loss": 0.8206,
      "step": 145
    },
    {
      "epoch": 0.062111801242236024,
      "grad_norm": 2.9291672578740395,
      "learning_rate": 6.198347107438017e-06,
      "loss": 0.8267,
      "step": 150
    },
    {
      "epoch": 0.06418219461697723,
      "grad_norm": 3.016454488266165,
      "learning_rate": 6.404958677685951e-06,
      "loss": 0.8248,
      "step": 155
    },
    {
      "epoch": 0.06625258799171843,
      "grad_norm": 3.159144146464083,
      "learning_rate": 6.611570247933885e-06,
      "loss": 0.8228,
      "step": 160
    },
    {
      "epoch": 0.06832298136645963,
      "grad_norm": 3.185934793134493,
      "learning_rate": 6.818181818181818e-06,
      "loss": 0.8237,
      "step": 165
    },
    {
      "epoch": 0.07039337474120083,
      "grad_norm": 2.946312652240558,
      "learning_rate": 7.0247933884297525e-06,
      "loss": 0.8185,
      "step": 170
    },
    {
      "epoch": 0.07246376811594203,
      "grad_norm": 3.1484491711371407,
      "learning_rate": 7.231404958677687e-06,
      "loss": 0.828,
      "step": 175
    },
    {
      "epoch": 0.07453416149068323,
      "grad_norm": 3.1865499342066155,
      "learning_rate": 7.43801652892562e-06,
      "loss": 0.8073,
      "step": 180
    },
    {
      "epoch": 0.07660455486542443,
      "grad_norm": 3.1897763608308463,
      "learning_rate": 7.644628099173555e-06,
      "loss": 0.8167,
      "step": 185
    },
    {
      "epoch": 0.07867494824016563,
      "grad_norm": 3.1445729384234538,
      "learning_rate": 7.851239669421489e-06,
      "loss": 0.8136,
      "step": 190
    },
    {
      "epoch": 0.08074534161490683,
      "grad_norm": 2.7998193963531373,
      "learning_rate": 8.057851239669421e-06,
      "loss": 0.8162,
      "step": 195
    },
    {
      "epoch": 0.08281573498964803,
      "grad_norm": 2.905185063170756,
      "learning_rate": 8.264462809917356e-06,
      "loss": 0.8159,
      "step": 200
    },
    {
      "epoch": 0.08488612836438923,
      "grad_norm": 2.959443568495212,
      "learning_rate": 8.47107438016529e-06,
      "loss": 0.821,
      "step": 205
    },
    {
      "epoch": 0.08695652173913043,
      "grad_norm": 3.017913411360177,
      "learning_rate": 8.677685950413224e-06,
      "loss": 0.804,
      "step": 210
    },
    {
      "epoch": 0.08902691511387163,
      "grad_norm": 4.370854357049643,
      "learning_rate": 8.884297520661158e-06,
      "loss": 0.8008,
      "step": 215
    },
    {
      "epoch": 0.09109730848861283,
      "grad_norm": 3.061402038819796,
      "learning_rate": 9.090909090909091e-06,
      "loss": 0.8202,
      "step": 220
    },
    {
      "epoch": 0.09316770186335403,
      "grad_norm": 3.044532011109754,
      "learning_rate": 9.297520661157025e-06,
      "loss": 0.8014,
      "step": 225
    },
    {
      "epoch": 0.09523809523809523,
      "grad_norm": 3.1027112832496235,
      "learning_rate": 9.50413223140496e-06,
      "loss": 0.7926,
      "step": 230
    },
    {
      "epoch": 0.09730848861283643,
      "grad_norm": 3.0500432141664344,
      "learning_rate": 9.710743801652894e-06,
      "loss": 0.7939,
      "step": 235
    },
    {
      "epoch": 0.09937888198757763,
      "grad_norm": 3.133694858260323,
      "learning_rate": 9.917355371900828e-06,
      "loss": 0.8053,
      "step": 240
    },
    {
      "epoch": 0.10144927536231885,
      "grad_norm": 2.9948460396055063,
      "learning_rate": 9.999952971391373e-06,
      "loss": 0.7917,
      "step": 245
    },
    {
      "epoch": 0.10351966873706005,
      "grad_norm": 3.3842529617656556,
      "learning_rate": 9.999665577542406e-06,
      "loss": 0.7917,
      "step": 250
    },
    {
      "epoch": 0.10559006211180125,
      "grad_norm": 3.0944979002960693,
      "learning_rate": 9.999116931848504e-06,
      "loss": 0.795,
      "step": 255
    },
    {
      "epoch": 0.10766045548654245,
      "grad_norm": 2.9068953195451317,
      "learning_rate": 9.99830706297852e-06,
      "loss": 0.8075,
      "step": 260
    },
    {
      "epoch": 0.10973084886128365,
      "grad_norm": 2.8873428503162275,
      "learning_rate": 9.997236013251234e-06,
      "loss": 0.809,
      "step": 265
    },
    {
      "epoch": 0.11180124223602485,
      "grad_norm": 3.039173901234368,
      "learning_rate": 9.995903838633133e-06,
      "loss": 0.7847,
      "step": 270
    },
    {
      "epoch": 0.11387163561076605,
      "grad_norm": 2.8747137326434733,
      "learning_rate": 9.994310608735492e-06,
      "loss": 0.7855,
      "step": 275
    },
    {
      "epoch": 0.11594202898550725,
      "grad_norm": 3.0667840135763678,
      "learning_rate": 9.99245640681072e-06,
      "loss": 0.7942,
      "step": 280
    },
    {
      "epoch": 0.11801242236024845,
      "grad_norm": 2.9327077394564194,
      "learning_rate": 9.99034132974804e-06,
      "loss": 0.7863,
      "step": 285
    },
    {
      "epoch": 0.12008281573498965,
      "grad_norm": 3.123350626594859,
      "learning_rate": 9.987965488068398e-06,
      "loss": 0.8026,
      "step": 290
    },
    {
      "epoch": 0.12215320910973085,
      "grad_norm": 2.9730293010922266,
      "learning_rate": 9.985329005918702e-06,
      "loss": 0.7739,
      "step": 295
    },
    {
      "epoch": 0.12422360248447205,
      "grad_norm": 2.7829102200794824,
      "learning_rate": 9.982432021065334e-06,
      "loss": 0.7637,
      "step": 300
    },
    {
      "epoch": 0.12629399585921325,
      "grad_norm": 2.9659408452421325,
      "learning_rate": 9.979274684886943e-06,
      "loss": 0.7936,
      "step": 305
    },
    {
      "epoch": 0.12836438923395446,
      "grad_norm": 2.8590342052469615,
      "learning_rate": 9.975857162366547e-06,
      "loss": 0.7734,
      "step": 310
    },
    {
      "epoch": 0.13043478260869565,
      "grad_norm": 2.8910277287275017,
      "learning_rate": 9.972179632082899e-06,
      "loss": 0.779,
      "step": 315
    },
    {
      "epoch": 0.13250517598343686,
      "grad_norm": 2.911402171783133,
      "learning_rate": 9.968242286201171e-06,
      "loss": 0.7706,
      "step": 320
    },
    {
      "epoch": 0.13457556935817805,
      "grad_norm": 2.8078094120564003,
      "learning_rate": 9.964045330462896e-06,
      "loss": 0.7603,
      "step": 325
    },
    {
      "epoch": 0.13664596273291926,
      "grad_norm": 2.9519342478579524,
      "learning_rate": 9.959588984175228e-06,
      "loss": 0.7608,
      "step": 330
    },
    {
      "epoch": 0.13871635610766045,
      "grad_norm": 2.7647092492977823,
      "learning_rate": 9.954873480199483e-06,
      "loss": 0.7599,
      "step": 335
    },
    {
      "epoch": 0.14078674948240166,
      "grad_norm": 2.890665953807783,
      "learning_rate": 9.94989906493896e-06,
      "loss": 0.7644,
      "step": 340
    },
    {
      "epoch": 0.14285714285714285,
      "grad_norm": 3.1328095384526726,
      "learning_rate": 9.94466599832608e-06,
      "loss": 0.749,
      "step": 345
    },
    {
      "epoch": 0.14492753623188406,
      "grad_norm": 3.172879461047842,
      "learning_rate": 9.939174553808793e-06,
      "loss": 0.7817,
      "step": 350
    },
    {
      "epoch": 0.14699792960662525,
      "grad_norm": 2.994989116337477,
      "learning_rate": 9.933425018336292e-06,
      "loss": 0.7473,
      "step": 355
    },
    {
      "epoch": 0.14906832298136646,
      "grad_norm": 2.804791316819872,
      "learning_rate": 9.927417692344025e-06,
      "loss": 0.7539,
      "step": 360
    },
    {
      "epoch": 0.15113871635610765,
      "grad_norm": 2.833426686026014,
      "learning_rate": 9.921152889737985e-06,
      "loss": 0.738,
      "step": 365
    },
    {
      "epoch": 0.15320910973084886,
      "grad_norm": 2.756942183064941,
      "learning_rate": 9.914630937878315e-06,
      "loss": 0.7485,
      "step": 370
    },
    {
      "epoch": 0.15527950310559005,
      "grad_norm": 2.8385059444311116,
      "learning_rate": 9.907852177562201e-06,
      "loss": 0.754,
      "step": 375
    },
    {
      "epoch": 0.15734989648033126,
      "grad_norm": 2.962284781813621,
      "learning_rate": 9.900816963006063e-06,
      "loss": 0.7489,
      "step": 380
    },
    {
      "epoch": 0.15942028985507245,
      "grad_norm": 2.711372880859091,
      "learning_rate": 9.893525661827043e-06,
      "loss": 0.7229,
      "step": 385
    },
    {
      "epoch": 0.16149068322981366,
      "grad_norm": 2.884149082130328,
      "learning_rate": 9.885978655023805e-06,
      "loss": 0.7259,
      "step": 390
    },
    {
      "epoch": 0.16356107660455488,
      "grad_norm": 3.1945304355844457,
      "learning_rate": 9.87817633695661e-06,
      "loss": 0.7218,
      "step": 395
    },
    {
      "epoch": 0.16563146997929606,
      "grad_norm": 2.8827563387322774,
      "learning_rate": 9.87011911532673e-06,
      "loss": 0.7375,
      "step": 400
    },
    {
      "epoch": 0.16770186335403728,
      "grad_norm": 2.95014074683271,
      "learning_rate": 9.861807411155126e-06,
      "loss": 0.7319,
      "step": 405
    },
    {
      "epoch": 0.16977225672877846,
      "grad_norm": 2.7849063486872825,
      "learning_rate": 9.853241658760457e-06,
      "loss": 0.7113,
      "step": 410
    },
    {
      "epoch": 0.17184265010351968,
      "grad_norm": 2.8782594804460464,
      "learning_rate": 9.844422305736383e-06,
      "loss": 0.7264,
      "step": 415
    },
    {
      "epoch": 0.17391304347826086,
      "grad_norm": 2.9263190094804954,
      "learning_rate": 9.835349812928178e-06,
      "loss": 0.7306,
      "step": 420
    },
    {
      "epoch": 0.17598343685300208,
      "grad_norm": 2.760904872765232,
      "learning_rate": 9.826024654408645e-06,
      "loss": 0.7124,
      "step": 425
    },
    {
      "epoch": 0.17805383022774326,
      "grad_norm": 2.8355885732528154,
      "learning_rate": 9.816447317453353e-06,
      "loss": 0.7154,
      "step": 430
    },
    {
      "epoch": 0.18012422360248448,
      "grad_norm": 2.7077368912423396,
      "learning_rate": 9.80661830251516e-06,
      "loss": 0.7206,
      "step": 435
    },
    {
      "epoch": 0.18219461697722567,
      "grad_norm": 2.684217340532534,
      "learning_rate": 9.796538123198077e-06,
      "loss": 0.7219,
      "step": 440
    },
    {
      "epoch": 0.18426501035196688,
      "grad_norm": 2.8088959863950627,
      "learning_rate": 9.786207306230422e-06,
      "loss": 0.7279,
      "step": 445
    },
    {
      "epoch": 0.18633540372670807,
      "grad_norm": 3.0142611729397357,
      "learning_rate": 9.775626391437303e-06,
      "loss": 0.6981,
      "step": 450
    },
    {
      "epoch": 0.18840579710144928,
      "grad_norm": 2.7077703599269687,
      "learning_rate": 9.764795931712396e-06,
      "loss": 0.7198,
      "step": 455
    },
    {
      "epoch": 0.19047619047619047,
      "grad_norm": 2.853217707324979,
      "learning_rate": 9.753716492989076e-06,
      "loss": 0.724,
      "step": 460
    },
    {
      "epoch": 0.19254658385093168,
      "grad_norm": 2.652904990685569,
      "learning_rate": 9.742388654210822e-06,
      "loss": 0.7123,
      "step": 465
    },
    {
      "epoch": 0.19461697722567287,
      "grad_norm": 2.905771516361513,
      "learning_rate": 9.730813007300984e-06,
      "loss": 0.6919,
      "step": 470
    },
    {
      "epoch": 0.19668737060041408,
      "grad_norm": 2.8307439840272672,
      "learning_rate": 9.718990157131841e-06,
      "loss": 0.7139,
      "step": 475
    },
    {
      "epoch": 0.19875776397515527,
      "grad_norm": 2.712449917206168,
      "learning_rate": 9.706920721492995e-06,
      "loss": 0.6772,
      "step": 480
    },
    {
      "epoch": 0.20082815734989648,
      "grad_norm": 2.8773743420198117,
      "learning_rate": 9.694605331059094e-06,
      "loss": 0.7035,
      "step": 485
    },
    {
      "epoch": 0.2028985507246377,
      "grad_norm": 2.9413832058067757,
      "learning_rate": 9.682044629356874e-06,
      "loss": 0.6934,
      "step": 490
    },
    {
      "epoch": 0.20496894409937888,
      "grad_norm": 2.6394745043796024,
      "learning_rate": 9.669239272731538e-06,
      "loss": 0.684,
      "step": 495
    },
    {
      "epoch": 0.2070393374741201,
      "grad_norm": 2.878205778747281,
      "learning_rate": 9.656189930312443e-06,
      "loss": 0.6814,
      "step": 500
    },
    {
      "epoch": 0.20910973084886128,
      "grad_norm": 2.8264154143199516,
      "learning_rate": 9.642897283978157e-06,
      "loss": 0.6692,
      "step": 505
    },
    {
      "epoch": 0.2111801242236025,
      "grad_norm": 2.7679312489741923,
      "learning_rate": 9.629362028320808e-06,
      "loss": 0.6892,
      "step": 510
    },
    {
      "epoch": 0.21325051759834368,
      "grad_norm": 2.722496962704404,
      "learning_rate": 9.615584870609809e-06,
      "loss": 0.6822,
      "step": 515
    },
    {
      "epoch": 0.2153209109730849,
      "grad_norm": 2.7991982869333434,
      "learning_rate": 9.601566530754882e-06,
      "loss": 0.6712,
      "step": 520
    },
    {
      "epoch": 0.21739130434782608,
      "grad_norm": 2.803028692487718,
      "learning_rate": 9.587307741268452e-06,
      "loss": 0.6651,
      "step": 525
    },
    {
      "epoch": 0.2194616977225673,
      "grad_norm": 3.131214470820513,
      "learning_rate": 9.572809247227366e-06,
      "loss": 0.6506,
      "step": 530
    },
    {
      "epoch": 0.22153209109730848,
      "grad_norm": 2.7951670426793465,
      "learning_rate": 9.558071806233955e-06,
      "loss": 0.6742,
      "step": 535
    },
    {
      "epoch": 0.2236024844720497,
      "grad_norm": 2.8950988911898765,
      "learning_rate": 9.54309618837646e-06,
      "loss": 0.6866,
      "step": 540
    },
    {
      "epoch": 0.22567287784679088,
      "grad_norm": 2.718511031914252,
      "learning_rate": 9.52788317618878e-06,
      "loss": 0.6503,
      "step": 545
    },
    {
      "epoch": 0.2277432712215321,
      "grad_norm": 2.7878464040608293,
      "learning_rate": 9.512433564609578e-06,
      "loss": 0.6625,
      "step": 550
    },
    {
      "epoch": 0.22981366459627328,
      "grad_norm": 2.847878590120255,
      "learning_rate": 9.496748160940762e-06,
      "loss": 0.6685,
      "step": 555
    },
    {
      "epoch": 0.2318840579710145,
      "grad_norm": 2.811190550076453,
      "learning_rate": 9.480827784805278e-06,
      "loss": 0.6618,
      "step": 560
    },
    {
      "epoch": 0.23395445134575568,
      "grad_norm": 2.8583977695483576,
      "learning_rate": 9.464673268104299e-06,
      "loss": 0.6611,
      "step": 565
    },
    {
      "epoch": 0.2360248447204969,
      "grad_norm": 2.7286737335894884,
      "learning_rate": 9.448285454973739e-06,
      "loss": 0.6657,
      "step": 570
    },
    {
      "epoch": 0.23809523809523808,
      "grad_norm": 2.820345491066426,
      "learning_rate": 9.431665201740154e-06,
      "loss": 0.6547,
      "step": 575
    },
    {
      "epoch": 0.2401656314699793,
      "grad_norm": 2.5527081352839294,
      "learning_rate": 9.414813376876003e-06,
      "loss": 0.6652,
      "step": 580
    },
    {
      "epoch": 0.2422360248447205,
      "grad_norm": 2.9411785059140865,
      "learning_rate": 9.397730860954242e-06,
      "loss": 0.6516,
      "step": 585
    },
    {
      "epoch": 0.2443064182194617,
      "grad_norm": 2.6327440700946436,
      "learning_rate": 9.38041854660234e-06,
      "loss": 0.6364,
      "step": 590
    },
    {
      "epoch": 0.2463768115942029,
      "grad_norm": 3.043944213064627,
      "learning_rate": 9.362877338455611e-06,
      "loss": 0.6551,
      "step": 595
    },
    {
      "epoch": 0.2484472049689441,
      "grad_norm": 2.848474741399472,
      "learning_rate": 9.345108153109963e-06,
      "loss": 0.6527,
      "step": 600
    },
    {
      "epoch": 0.2505175983436853,
      "grad_norm": 2.952778244263716,
      "learning_rate": 9.327111919073988e-06,
      "loss": 0.6539,
      "step": 605
    },
    {
      "epoch": 0.2525879917184265,
      "grad_norm": 2.6649622862114786,
      "learning_rate": 9.308889576720453e-06,
      "loss": 0.626,
      "step": 610
    },
    {
      "epoch": 0.2546583850931677,
      "grad_norm": 2.823552162711532,
      "learning_rate": 9.290442078237154e-06,
      "loss": 0.6373,
      "step": 615
    },
    {
      "epoch": 0.2567287784679089,
      "grad_norm": 2.7416237554541594,
      "learning_rate": 9.271770387577168e-06,
      "loss": 0.6473,
      "step": 620
    },
    {
      "epoch": 0.2587991718426501,
      "grad_norm": 2.923293350471095,
      "learning_rate": 9.252875480408479e-06,
      "loss": 0.6688,
      "step": 625
    },
    {
      "epoch": 0.2608695652173913,
      "grad_norm": 2.873447018779356,
      "learning_rate": 9.233758344062996e-06,
      "loss": 0.6358,
      "step": 630
    },
    {
      "epoch": 0.2629399585921325,
      "grad_norm": 2.8056432962640945,
      "learning_rate": 9.21441997748496e-06,
      "loss": 0.6316,
      "step": 635
    },
    {
      "epoch": 0.2650103519668737,
      "grad_norm": 2.730455371079045,
      "learning_rate": 9.194861391178749e-06,
      "loss": 0.6237,
      "step": 640
    },
    {
      "epoch": 0.2670807453416149,
      "grad_norm": 2.802204514104521,
      "learning_rate": 9.175083607156067e-06,
      "loss": 0.619,
      "step": 645
    },
    {
      "epoch": 0.2691511387163561,
      "grad_norm": 2.7509135382081302,
      "learning_rate": 9.155087658882555e-06,
      "loss": 0.6277,
      "step": 650
    },
    {
      "epoch": 0.2712215320910973,
      "grad_norm": 2.705886377125874,
      "learning_rate": 9.134874591223773e-06,
      "loss": 0.6232,
      "step": 655
    },
    {
      "epoch": 0.2732919254658385,
      "grad_norm": 2.679912639223936,
      "learning_rate": 9.114445460390605e-06,
      "loss": 0.6019,
      "step": 660
    },
    {
      "epoch": 0.2753623188405797,
      "grad_norm": 2.794538981694357,
      "learning_rate": 9.093801333884076e-06,
      "loss": 0.6381,
      "step": 665
    },
    {
      "epoch": 0.2774327122153209,
      "grad_norm": 2.6526958858869327,
      "learning_rate": 9.072943290439566e-06,
      "loss": 0.6108,
      "step": 670
    },
    {
      "epoch": 0.2795031055900621,
      "grad_norm": 2.63138640840009,
      "learning_rate": 9.051872419970439e-06,
      "loss": 0.6152,
      "step": 675
    },
    {
      "epoch": 0.2815734989648033,
      "grad_norm": 2.6305618312221637,
      "learning_rate": 9.03058982351109e-06,
      "loss": 0.6244,
      "step": 680
    },
    {
      "epoch": 0.2836438923395445,
      "grad_norm": 2.7172718922096273,
      "learning_rate": 9.009096613159426e-06,
      "loss": 0.6178,
      "step": 685
    },
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 2.838946638052953,
      "learning_rate": 8.987393912018732e-06,
      "loss": 0.6136,
      "step": 690
    },
    {
      "epoch": 0.28778467908902694,
      "grad_norm": 2.964196227945642,
      "learning_rate": 8.965482854139003e-06,
      "loss": 0.6156,
      "step": 695
    },
    {
      "epoch": 0.2898550724637681,
      "grad_norm": 2.6840750910870934,
      "learning_rate": 8.943364584457675e-06,
      "loss": 0.6055,
      "step": 700
    },
    {
      "epoch": 0.2919254658385093,
      "grad_norm": 2.710361815067833,
      "learning_rate": 8.921040258739804e-06,
      "loss": 0.6146,
      "step": 705
    },
    {
      "epoch": 0.2939958592132505,
      "grad_norm": 2.6938443856684566,
      "learning_rate": 8.898511043517668e-06,
      "loss": 0.6023,
      "step": 710
    },
    {
      "epoch": 0.29606625258799174,
      "grad_norm": 3.039550668602838,
      "learning_rate": 8.875778116029816e-06,
      "loss": 0.6099,
      "step": 715
    },
    {
      "epoch": 0.2981366459627329,
      "grad_norm": 2.6851125546932844,
      "learning_rate": 8.85284266415955e-06,
      "loss": 0.5958,
      "step": 720
    },
    {
      "epoch": 0.3002070393374741,
      "grad_norm": 2.769102644505053,
      "learning_rate": 8.829705886372845e-06,
      "loss": 0.5835,
      "step": 725
    },
    {
      "epoch": 0.3022774327122153,
      "grad_norm": 2.7245422741801764,
      "learning_rate": 8.806368991655747e-06,
      "loss": 0.5987,
      "step": 730
    },
    {
      "epoch": 0.30434782608695654,
      "grad_norm": 2.7261962676037617,
      "learning_rate": 8.782833199451177e-06,
      "loss": 0.5809,
      "step": 735
    },
    {
      "epoch": 0.3064182194616977,
      "grad_norm": 2.754432152388888,
      "learning_rate": 8.759099739595215e-06,
      "loss": 0.5949,
      "step": 740
    },
    {
      "epoch": 0.3084886128364389,
      "grad_norm": 2.663972585571618,
      "learning_rate": 8.735169852252848e-06,
      "loss": 0.6007,
      "step": 745
    },
    {
      "epoch": 0.3105590062111801,
      "grad_norm": 2.6375841694420457,
      "learning_rate": 8.71104478785315e-06,
      "loss": 0.5914,
      "step": 750
    },
    {
      "epoch": 0.31262939958592134,
      "grad_norm": 2.8319346788921558,
      "learning_rate": 8.686725807023955e-06,
      "loss": 0.5691,
      "step": 755
    },
    {
      "epoch": 0.3146997929606625,
      "grad_norm": 2.6240714205980353,
      "learning_rate": 8.662214180525982e-06,
      "loss": 0.5793,
      "step": 760
    },
    {
      "epoch": 0.3167701863354037,
      "grad_norm": 2.472338658275942,
      "learning_rate": 8.637511189186425e-06,
      "loss": 0.5731,
      "step": 765
    },
    {
      "epoch": 0.3188405797101449,
      "grad_norm": 2.5924016465268465,
      "learning_rate": 8.612618123832033e-06,
      "loss": 0.5865,
      "step": 770
    },
    {
      "epoch": 0.32091097308488614,
      "grad_norm": 2.6405443390765777,
      "learning_rate": 8.587536285221656e-06,
      "loss": 0.5865,
      "step": 775
    },
    {
      "epoch": 0.32298136645962733,
      "grad_norm": 2.5215271423652963,
      "learning_rate": 8.562266983978278e-06,
      "loss": 0.5777,
      "step": 780
    },
    {
      "epoch": 0.3250517598343685,
      "grad_norm": 2.6637851775195673,
      "learning_rate": 8.536811540520529e-06,
      "loss": 0.5767,
      "step": 785
    },
    {
      "epoch": 0.32712215320910976,
      "grad_norm": 2.6904258127362106,
      "learning_rate": 8.511171284993686e-06,
      "loss": 0.5787,
      "step": 790
    },
    {
      "epoch": 0.32919254658385094,
      "grad_norm": 2.687778300078121,
      "learning_rate": 8.485347557200177e-06,
      "loss": 0.5698,
      "step": 795
    },
    {
      "epoch": 0.33126293995859213,
      "grad_norm": 2.6651484601638313,
      "learning_rate": 8.459341706529557e-06,
      "loss": 0.5741,
      "step": 800
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 2.6748542832312774,
      "learning_rate": 8.43315509188801e-06,
      "loss": 0.5697,
      "step": 805
    },
    {
      "epoch": 0.33540372670807456,
      "grad_norm": 2.86863347937226,
      "learning_rate": 8.406789081627337e-06,
      "loss": 0.5711,
      "step": 810
    },
    {
      "epoch": 0.33747412008281574,
      "grad_norm": 2.611903241997596,
      "learning_rate": 8.380245053473452e-06,
      "loss": 0.5655,
      "step": 815
    },
    {
      "epoch": 0.33954451345755693,
      "grad_norm": 2.4615158195004025,
      "learning_rate": 8.353524394454388e-06,
      "loss": 0.5809,
      "step": 820
    },
    {
      "epoch": 0.3416149068322981,
      "grad_norm": 2.7624083954477836,
      "learning_rate": 8.326628500827826e-06,
      "loss": 0.5632,
      "step": 825
    },
    {
      "epoch": 0.34368530020703936,
      "grad_norm": 2.6351139394976495,
      "learning_rate": 8.299558778008137e-06,
      "loss": 0.5525,
      "step": 830
    },
    {
      "epoch": 0.34575569358178054,
      "grad_norm": 2.5043190162644153,
      "learning_rate": 8.272316640492932e-06,
      "loss": 0.5392,
      "step": 835
    },
    {
      "epoch": 0.34782608695652173,
      "grad_norm": 2.867606809151843,
      "learning_rate": 8.244903511789158e-06,
      "loss": 0.557,
      "step": 840
    },
    {
      "epoch": 0.3498964803312629,
      "grad_norm": 2.684801681192226,
      "learning_rate": 8.217320824338713e-06,
      "loss": 0.5745,
      "step": 845
    },
    {
      "epoch": 0.35196687370600416,
      "grad_norm": 2.6586432255444055,
      "learning_rate": 8.189570019443597e-06,
      "loss": 0.5656,
      "step": 850
    },
    {
      "epoch": 0.35403726708074534,
      "grad_norm": 2.6258843569228634,
      "learning_rate": 8.161652547190593e-06,
      "loss": 0.5486,
      "step": 855
    },
    {
      "epoch": 0.35610766045548653,
      "grad_norm": 2.4933425053393807,
      "learning_rate": 8.133569866375497e-06,
      "loss": 0.5508,
      "step": 860
    },
    {
      "epoch": 0.3581780538302277,
      "grad_norm": 2.564550608540871,
      "learning_rate": 8.105323444426891e-06,
      "loss": 0.5541,
      "step": 865
    },
    {
      "epoch": 0.36024844720496896,
      "grad_norm": 2.5540667156660355,
      "learning_rate": 8.076914757329467e-06,
      "loss": 0.5417,
      "step": 870
    },
    {
      "epoch": 0.36231884057971014,
      "grad_norm": 2.582689022943949,
      "learning_rate": 8.048345289546895e-06,
      "loss": 0.5657,
      "step": 875
    },
    {
      "epoch": 0.36438923395445133,
      "grad_norm": 2.7609773998784397,
      "learning_rate": 8.01961653394426e-06,
      "loss": 0.5454,
      "step": 880
    },
    {
      "epoch": 0.36645962732919257,
      "grad_norm": 2.5771925893254872,
      "learning_rate": 7.990729991710046e-06,
      "loss": 0.5695,
      "step": 885
    },
    {
      "epoch": 0.36853002070393376,
      "grad_norm": 2.7074639836593204,
      "learning_rate": 7.9616871722777e-06,
      "loss": 0.5435,
      "step": 890
    },
    {
      "epoch": 0.37060041407867494,
      "grad_norm": 2.5488339841796446,
      "learning_rate": 7.932489593246764e-06,
      "loss": 0.5435,
      "step": 895
    },
    {
      "epoch": 0.37267080745341613,
      "grad_norm": 2.7167581061080246,
      "learning_rate": 7.903138780303556e-06,
      "loss": 0.5345,
      "step": 900
    },
    {
      "epoch": 0.3747412008281574,
      "grad_norm": 2.531463814204388,
      "learning_rate": 7.873636267141463e-06,
      "loss": 0.5279,
      "step": 905
    },
    {
      "epoch": 0.37681159420289856,
      "grad_norm": 2.6017286242409394,
      "learning_rate": 7.843983595380793e-06,
      "loss": 0.5291,
      "step": 910
    },
    {
      "epoch": 0.37888198757763975,
      "grad_norm": 2.5985329952689638,
      "learning_rate": 7.814182314488225e-06,
      "loss": 0.5494,
      "step": 915
    },
    {
      "epoch": 0.38095238095238093,
      "grad_norm": 2.5237542203948333,
      "learning_rate": 7.784233981695835e-06,
      "loss": 0.5415,
      "step": 920
    },
    {
      "epoch": 0.3830227743271222,
      "grad_norm": 2.529651558899943,
      "learning_rate": 7.754140161919732e-06,
      "loss": 0.5106,
      "step": 925
    },
    {
      "epoch": 0.38509316770186336,
      "grad_norm": 2.6750645042360364,
      "learning_rate": 7.72390242767828e-06,
      "loss": 0.538,
      "step": 930
    },
    {
      "epoch": 0.38716356107660455,
      "grad_norm": 2.580657267143272,
      "learning_rate": 7.693522359009931e-06,
      "loss": 0.5451,
      "step": 935
    },
    {
      "epoch": 0.38923395445134573,
      "grad_norm": 2.615969551634785,
      "learning_rate": 7.663001543390657e-06,
      "loss": 0.5277,
      "step": 940
    },
    {
      "epoch": 0.391304347826087,
      "grad_norm": 2.592791677872246,
      "learning_rate": 7.63234157565101e-06,
      "loss": 0.5323,
      "step": 945
    },
    {
      "epoch": 0.39337474120082816,
      "grad_norm": 2.487384749510932,
      "learning_rate": 7.601544057892769e-06,
      "loss": 0.51,
      "step": 950
    },
    {
      "epoch": 0.39544513457556935,
      "grad_norm": 2.593418655541288,
      "learning_rate": 7.570610599405242e-06,
      "loss": 0.5343,
      "step": 955
    },
    {
      "epoch": 0.39751552795031053,
      "grad_norm": 2.8999221466042915,
      "learning_rate": 7.539542816581157e-06,
      "loss": 0.524,
      "step": 960
    },
    {
      "epoch": 0.3995859213250518,
      "grad_norm": 2.650408033983142,
      "learning_rate": 7.508342332832213e-06,
      "loss": 0.5217,
      "step": 965
    },
    {
      "epoch": 0.40165631469979296,
      "grad_norm": 2.9552337045137183,
      "learning_rate": 7.477010778504241e-06,
      "loss": 0.5222,
      "step": 970
    },
    {
      "epoch": 0.40372670807453415,
      "grad_norm": 2.520049180680822,
      "learning_rate": 7.445549790792021e-06,
      "loss": 0.5252,
      "step": 975
    },
    {
      "epoch": 0.4057971014492754,
      "grad_norm": 2.696990346347782,
      "learning_rate": 7.413961013653725e-06,
      "loss": 0.5091,
      "step": 980
    },
    {
      "epoch": 0.4078674948240166,
      "grad_norm": 2.6730517410678925,
      "learning_rate": 7.3822460977250145e-06,
      "loss": 0.5269,
      "step": 985
    },
    {
      "epoch": 0.40993788819875776,
      "grad_norm": 2.5599496905533448,
      "learning_rate": 7.350406700232794e-06,
      "loss": 0.5119,
      "step": 990
    },
    {
      "epoch": 0.41200828157349895,
      "grad_norm": 2.6712648459988446,
      "learning_rate": 7.318444484908606e-06,
      "loss": 0.5056,
      "step": 995
    },
    {
      "epoch": 0.4140786749482402,
      "grad_norm": 2.6461277264747958,
      "learning_rate": 7.286361121901706e-06,
      "loss": 0.4928,
      "step": 1000
    },
    {
      "epoch": 0.4161490683229814,
      "grad_norm": 2.6077956361678645,
      "learning_rate": 7.254158287691775e-06,
      "loss": 0.5173,
      "step": 1005
    },
    {
      "epoch": 0.41821946169772256,
      "grad_norm": 2.7386442940777487,
      "learning_rate": 7.221837665001335e-06,
      "loss": 0.5097,
      "step": 1010
    },
    {
      "epoch": 0.42028985507246375,
      "grad_norm": 2.652197927020302,
      "learning_rate": 7.189400942707804e-06,
      "loss": 0.5028,
      "step": 1015
    },
    {
      "epoch": 0.422360248447205,
      "grad_norm": 2.563491443097895,
      "learning_rate": 7.1568498157552576e-06,
      "loss": 0.5181,
      "step": 1020
    },
    {
      "epoch": 0.4244306418219462,
      "grad_norm": 2.67515213627178,
      "learning_rate": 7.124185985065856e-06,
      "loss": 0.4922,
      "step": 1025
    },
    {
      "epoch": 0.42650103519668736,
      "grad_norm": 2.466636516441553,
      "learning_rate": 7.091411157450965e-06,
      "loss": 0.5051,
      "step": 1030
    },
    {
      "epoch": 0.42857142857142855,
      "grad_norm": 2.5736481344369153,
      "learning_rate": 7.0585270455219654e-06,
      "loss": 0.498,
      "step": 1035
    },
    {
      "epoch": 0.4306418219461698,
      "grad_norm": 2.4892776880439733,
      "learning_rate": 7.025535367600771e-06,
      "loss": 0.487,
      "step": 1040
    },
    {
      "epoch": 0.432712215320911,
      "grad_norm": 2.5603591701407558,
      "learning_rate": 6.992437847630031e-06,
      "loss": 0.4948,
      "step": 1045
    },
    {
      "epoch": 0.43478260869565216,
      "grad_norm": 2.6583351951818424,
      "learning_rate": 6.95923621508305e-06,
      "loss": 0.5025,
      "step": 1050
    },
    {
      "epoch": 0.43685300207039335,
      "grad_norm": 2.683322824382553,
      "learning_rate": 6.92593220487342e-06,
      "loss": 0.4992,
      "step": 1055
    },
    {
      "epoch": 0.4389233954451346,
      "grad_norm": 2.6381061202717313,
      "learning_rate": 6.892527557264358e-06,
      "loss": 0.4973,
      "step": 1060
    },
    {
      "epoch": 0.4409937888198758,
      "grad_norm": 2.517843029266686,
      "learning_rate": 6.859024017777779e-06,
      "loss": 0.4964,
      "step": 1065
    },
    {
      "epoch": 0.44306418219461696,
      "grad_norm": 2.4754926250370146,
      "learning_rate": 6.825423337103074e-06,
      "loss": 0.4949,
      "step": 1070
    },
    {
      "epoch": 0.4451345755693582,
      "grad_norm": 2.635485649219349,
      "learning_rate": 6.791727271005642e-06,
      "loss": 0.5003,
      "step": 1075
    },
    {
      "epoch": 0.4472049689440994,
      "grad_norm": 2.6120492040696726,
      "learning_rate": 6.757937580235138e-06,
      "loss": 0.4914,
      "step": 1080
    },
    {
      "epoch": 0.4492753623188406,
      "grad_norm": 2.5774198115084817,
      "learning_rate": 6.724056030433464e-06,
      "loss": 0.4949,
      "step": 1085
    },
    {
      "epoch": 0.45134575569358176,
      "grad_norm": 2.482721848182028,
      "learning_rate": 6.690084392042514e-06,
      "loss": 0.496,
      "step": 1090
    },
    {
      "epoch": 0.453416149068323,
      "grad_norm": 2.6018747677626886,
      "learning_rate": 6.656024440211662e-06,
      "loss": 0.5045,
      "step": 1095
    },
    {
      "epoch": 0.4554865424430642,
      "grad_norm": 2.4671473100342687,
      "learning_rate": 6.621877954704996e-06,
      "loss": 0.4734,
      "step": 1100
    },
    {
      "epoch": 0.4575569358178054,
      "grad_norm": 2.4572774152094996,
      "learning_rate": 6.5876467198083235e-06,
      "loss": 0.4751,
      "step": 1105
    },
    {
      "epoch": 0.45962732919254656,
      "grad_norm": 2.474193371330068,
      "learning_rate": 6.553332524235937e-06,
      "loss": 0.4873,
      "step": 1110
    },
    {
      "epoch": 0.4616977225672878,
      "grad_norm": 2.6481958552698863,
      "learning_rate": 6.518937161037144e-06,
      "loss": 0.4814,
      "step": 1115
    },
    {
      "epoch": 0.463768115942029,
      "grad_norm": 2.6220723699684676,
      "learning_rate": 6.484462427502572e-06,
      "loss": 0.4888,
      "step": 1120
    },
    {
      "epoch": 0.4658385093167702,
      "grad_norm": 2.3888509125191435,
      "learning_rate": 6.44991012507026e-06,
      "loss": 0.4886,
      "step": 1125
    },
    {
      "epoch": 0.46790890269151136,
      "grad_norm": 2.447502949451342,
      "learning_rate": 6.415282059231518e-06,
      "loss": 0.4765,
      "step": 1130
    },
    {
      "epoch": 0.4699792960662526,
      "grad_norm": 2.457351061389078,
      "learning_rate": 6.380580039436586e-06,
      "loss": 0.4766,
      "step": 1135
    },
    {
      "epoch": 0.4720496894409938,
      "grad_norm": 2.651861961957335,
      "learning_rate": 6.345805879000087e-06,
      "loss": 0.4799,
      "step": 1140
    },
    {
      "epoch": 0.474120082815735,
      "grad_norm": 2.5297244979998585,
      "learning_rate": 6.31096139500627e-06,
      "loss": 0.4663,
      "step": 1145
    },
    {
      "epoch": 0.47619047619047616,
      "grad_norm": 2.582774310852352,
      "learning_rate": 6.2760484082140604e-06,
      "loss": 0.4804,
      "step": 1150
    },
    {
      "epoch": 0.4782608695652174,
      "grad_norm": 2.7841763944290387,
      "learning_rate": 6.24106874296192e-06,
      "loss": 0.475,
      "step": 1155
    },
    {
      "epoch": 0.4803312629399586,
      "grad_norm": 2.598926437772734,
      "learning_rate": 6.20602422707252e-06,
      "loss": 0.4828,
      "step": 1160
    },
    {
      "epoch": 0.4824016563146998,
      "grad_norm": 2.570108784715483,
      "learning_rate": 6.1709166917572264e-06,
      "loss": 0.4741,
      "step": 1165
    },
    {
      "epoch": 0.484472049689441,
      "grad_norm": 2.4732626483650777,
      "learning_rate": 6.135747971520412e-06,
      "loss": 0.4617,
      "step": 1170
    },
    {
      "epoch": 0.4865424430641822,
      "grad_norm": 2.500231147008793,
      "learning_rate": 6.100519904063597e-06,
      "loss": 0.4741,
      "step": 1175
    },
    {
      "epoch": 0.4886128364389234,
      "grad_norm": 2.451179307556641,
      "learning_rate": 6.0652343301894345e-06,
      "loss": 0.464,
      "step": 1180
    },
    {
      "epoch": 0.4906832298136646,
      "grad_norm": 2.5319245325648203,
      "learning_rate": 6.029893093705492e-06,
      "loss": 0.4588,
      "step": 1185
    },
    {
      "epoch": 0.4927536231884058,
      "grad_norm": 2.643929350614943,
      "learning_rate": 5.99449804132794e-06,
      "loss": 0.4754,
      "step": 1190
    },
    {
      "epoch": 0.494824016563147,
      "grad_norm": 2.8193165892727845,
      "learning_rate": 5.959051022585025e-06,
      "loss": 0.4496,
      "step": 1195
    },
    {
      "epoch": 0.4968944099378882,
      "grad_norm": 2.4984753809584928,
      "learning_rate": 5.923553889720447e-06,
      "loss": 0.4446,
      "step": 1200
    },
    {
      "epoch": 0.4989648033126294,
      "grad_norm": 2.6670817354067564,
      "learning_rate": 5.888008497596553e-06,
      "loss": 0.4426,
      "step": 1205
    },
    {
      "epoch": 0.5010351966873706,
      "grad_norm": 2.399729029047019,
      "learning_rate": 5.852416703597431e-06,
      "loss": 0.4577,
      "step": 1210
    },
    {
      "epoch": 0.5031055900621118,
      "grad_norm": 2.784756711467886,
      "learning_rate": 5.816780367531841e-06,
      "loss": 0.4516,
      "step": 1215
    },
    {
      "epoch": 0.505175983436853,
      "grad_norm": 2.4161773663343005,
      "learning_rate": 5.781101351536041e-06,
      "loss": 0.4485,
      "step": 1220
    },
    {
      "epoch": 0.5072463768115942,
      "grad_norm": 2.4753805663487527,
      "learning_rate": 5.745381519976477e-06,
      "loss": 0.4461,
      "step": 1225
    },
    {
      "epoch": 0.5093167701863354,
      "grad_norm": 2.590134004742379,
      "learning_rate": 5.7096227393523716e-06,
      "loss": 0.4424,
      "step": 1230
    },
    {
      "epoch": 0.5113871635610766,
      "grad_norm": 2.4167105774866178,
      "learning_rate": 5.673826878198181e-06,
      "loss": 0.4582,
      "step": 1235
    },
    {
      "epoch": 0.5134575569358178,
      "grad_norm": 2.575817535734993,
      "learning_rate": 5.637995806985961e-06,
      "loss": 0.4555,
      "step": 1240
    },
    {
      "epoch": 0.515527950310559,
      "grad_norm": 2.726215777213599,
      "learning_rate": 5.602131398027637e-06,
      "loss": 0.4533,
      "step": 1245
    },
    {
      "epoch": 0.5175983436853002,
      "grad_norm": 2.434999896670774,
      "learning_rate": 5.566235525377155e-06,
      "loss": 0.4545,
      "step": 1250
    },
    {
      "epoch": 0.5196687370600414,
      "grad_norm": 2.572821476079958,
      "learning_rate": 5.530310064732559e-06,
      "loss": 0.4638,
      "step": 1255
    },
    {
      "epoch": 0.5217391304347826,
      "grad_norm": 2.681764613329783,
      "learning_rate": 5.494356893337985e-06,
      "loss": 0.444,
      "step": 1260
    },
    {
      "epoch": 0.5238095238095238,
      "grad_norm": 2.3842952421074353,
      "learning_rate": 5.4583778898855576e-06,
      "loss": 0.4391,
      "step": 1265
    },
    {
      "epoch": 0.525879917184265,
      "grad_norm": 2.475955241043757,
      "learning_rate": 5.422374934417228e-06,
      "loss": 0.4377,
      "step": 1270
    },
    {
      "epoch": 0.5279503105590062,
      "grad_norm": 2.3691643158148845,
      "learning_rate": 5.386349908226538e-06,
      "loss": 0.4443,
      "step": 1275
    },
    {
      "epoch": 0.5300207039337475,
      "grad_norm": 2.5636774466815044,
      "learning_rate": 5.350304693760301e-06,
      "loss": 0.4381,
      "step": 1280
    },
    {
      "epoch": 0.5320910973084886,
      "grad_norm": 2.4902170343931385,
      "learning_rate": 5.314241174520251e-06,
      "loss": 0.4497,
      "step": 1285
    },
    {
      "epoch": 0.5341614906832298,
      "grad_norm": 2.4675491824566853,
      "learning_rate": 5.2781612349646175e-06,
      "loss": 0.452,
      "step": 1290
    },
    {
      "epoch": 0.5362318840579711,
      "grad_norm": 2.6063673150297006,
      "learning_rate": 5.242066760409653e-06,
      "loss": 0.4418,
      "step": 1295
    },
    {
      "epoch": 0.5383022774327122,
      "grad_norm": 2.5335407332398927,
      "learning_rate": 5.205959636931121e-06,
      "loss": 0.4403,
      "step": 1300
    },
    {
      "epoch": 0.5403726708074534,
      "grad_norm": 2.4091180065963975,
      "learning_rate": 5.169841751265738e-06,
      "loss": 0.4387,
      "step": 1305
    },
    {
      "epoch": 0.5424430641821946,
      "grad_norm": 2.424987007890647,
      "learning_rate": 5.13371499071259e-06,
      "loss": 0.4479,
      "step": 1310
    },
    {
      "epoch": 0.5445134575569358,
      "grad_norm": 2.51062913767981,
      "learning_rate": 5.09758124303451e-06,
      "loss": 0.4433,
      "step": 1315
    },
    {
      "epoch": 0.546583850931677,
      "grad_norm": 2.4874780691845557,
      "learning_rate": 5.0614423963594295e-06,
      "loss": 0.4584,
      "step": 1320
    },
    {
      "epoch": 0.5486542443064182,
      "grad_norm": 2.5756703742769553,
      "learning_rate": 5.0253003390817264e-06,
      "loss": 0.4303,
      "step": 1325
    },
    {
      "epoch": 0.5507246376811594,
      "grad_norm": 2.377595858135612,
      "learning_rate": 4.989156959763551e-06,
      "loss": 0.4413,
      "step": 1330
    },
    {
      "epoch": 0.5527950310559007,
      "grad_norm": 2.6615519130034984,
      "learning_rate": 4.953014147036121e-06,
      "loss": 0.4364,
      "step": 1335
    },
    {
      "epoch": 0.5548654244306418,
      "grad_norm": 2.389728145124797,
      "learning_rate": 4.9168737895010615e-06,
      "loss": 0.4178,
      "step": 1340
    },
    {
      "epoch": 0.556935817805383,
      "grad_norm": 2.4448080484811063,
      "learning_rate": 4.880737775631698e-06,
      "loss": 0.4408,
      "step": 1345
    },
    {
      "epoch": 0.5590062111801242,
      "grad_norm": 2.538046447227198,
      "learning_rate": 4.844607993674382e-06,
      "loss": 0.419,
      "step": 1350
    },
    {
      "epoch": 0.5610766045548654,
      "grad_norm": 2.5715273742005826,
      "learning_rate": 4.808486331549824e-06,
      "loss": 0.4184,
      "step": 1355
    },
    {
      "epoch": 0.5631469979296067,
      "grad_norm": 2.401834413284635,
      "learning_rate": 4.772374676754444e-06,
      "loss": 0.4296,
      "step": 1360
    },
    {
      "epoch": 0.5652173913043478,
      "grad_norm": 2.49827808323981,
      "learning_rate": 4.736274916261741e-06,
      "loss": 0.4199,
      "step": 1365
    },
    {
      "epoch": 0.567287784679089,
      "grad_norm": 2.3378556121176906,
      "learning_rate": 4.700188936423683e-06,
      "loss": 0.4366,
      "step": 1370
    },
    {
      "epoch": 0.5693581780538303,
      "grad_norm": 2.7241212266546575,
      "learning_rate": 4.664118622872157e-06,
      "loss": 0.4249,
      "step": 1375
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 2.5314627132108427,
      "learning_rate": 4.628065860420417e-06,
      "loss": 0.4253,
      "step": 1380
    },
    {
      "epoch": 0.5734989648033126,
      "grad_norm": 2.3329708690084234,
      "learning_rate": 4.592032532964611e-06,
      "loss": 0.4218,
      "step": 1385
    },
    {
      "epoch": 0.5755693581780539,
      "grad_norm": 2.4531049657154966,
      "learning_rate": 4.556020523385326e-06,
      "loss": 0.4153,
      "step": 1390
    },
    {
      "epoch": 0.577639751552795,
      "grad_norm": 2.4101911561459546,
      "learning_rate": 4.520031713449215e-06,
      "loss": 0.4146,
      "step": 1395
    },
    {
      "epoch": 0.5797101449275363,
      "grad_norm": 2.4926174054292805,
      "learning_rate": 4.484067983710653e-06,
      "loss": 0.4133,
      "step": 1400
    },
    {
      "epoch": 0.5817805383022774,
      "grad_norm": 2.498578807282515,
      "learning_rate": 4.448131213413485e-06,
      "loss": 0.4162,
      "step": 1405
    },
    {
      "epoch": 0.5838509316770186,
      "grad_norm": 2.664768704875309,
      "learning_rate": 4.4122232803928145e-06,
      "loss": 0.4105,
      "step": 1410
    },
    {
      "epoch": 0.5859213250517599,
      "grad_norm": 2.2692861748470508,
      "learning_rate": 4.376346060976888e-06,
      "loss": 0.4059,
      "step": 1415
    },
    {
      "epoch": 0.587991718426501,
      "grad_norm": 2.4319188428005623,
      "learning_rate": 4.340501429889053e-06,
      "loss": 0.4076,
      "step": 1420
    },
    {
      "epoch": 0.5900621118012422,
      "grad_norm": 2.502953068392033,
      "learning_rate": 4.30469126014978e-06,
      "loss": 0.4158,
      "step": 1425
    },
    {
      "epoch": 0.5921325051759835,
      "grad_norm": 2.453793897664311,
      "learning_rate": 4.268917422978811e-06,
      "loss": 0.4168,
      "step": 1430
    },
    {
      "epoch": 0.5942028985507246,
      "grad_norm": 2.3774700360836607,
      "learning_rate": 4.233181787697364e-06,
      "loss": 0.4212,
      "step": 1435
    },
    {
      "epoch": 0.5962732919254659,
      "grad_norm": 2.3933353889641737,
      "learning_rate": 4.197486221630467e-06,
      "loss": 0.4118,
      "step": 1440
    },
    {
      "epoch": 0.598343685300207,
      "grad_norm": 2.598159871976961,
      "learning_rate": 4.161832590009371e-06,
      "loss": 0.4071,
      "step": 1445
    },
    {
      "epoch": 0.6004140786749482,
      "grad_norm": 2.390460104457502,
      "learning_rate": 4.1262227558740915e-06,
      "loss": 0.4076,
      "step": 1450
    },
    {
      "epoch": 0.6024844720496895,
      "grad_norm": 2.532902902648854,
      "learning_rate": 4.090658579976058e-06,
      "loss": 0.4112,
      "step": 1455
    },
    {
      "epoch": 0.6045548654244306,
      "grad_norm": 2.3357432837994945,
      "learning_rate": 4.055141920680883e-06,
      "loss": 0.4078,
      "step": 1460
    },
    {
      "epoch": 0.6066252587991718,
      "grad_norm": 2.4941110251854615,
      "learning_rate": 4.019674633871246e-06,
      "loss": 0.4076,
      "step": 1465
    },
    {
      "epoch": 0.6086956521739131,
      "grad_norm": 2.336216844284834,
      "learning_rate": 3.984258572849926e-06,
      "loss": 0.4041,
      "step": 1470
    },
    {
      "epoch": 0.6107660455486542,
      "grad_norm": 2.3720339996655047,
      "learning_rate": 3.9488955882429605e-06,
      "loss": 0.3938,
      "step": 1475
    },
    {
      "epoch": 0.6128364389233955,
      "grad_norm": 2.442624850251411,
      "learning_rate": 3.9135875279029366e-06,
      "loss": 0.4127,
      "step": 1480
    },
    {
      "epoch": 0.6149068322981367,
      "grad_norm": 2.443161565789866,
      "learning_rate": 3.878336236812432e-06,
      "loss": 0.4087,
      "step": 1485
    },
    {
      "epoch": 0.6169772256728778,
      "grad_norm": 2.529104103341601,
      "learning_rate": 3.84314355698762e-06,
      "loss": 0.4096,
      "step": 1490
    },
    {
      "epoch": 0.6190476190476191,
      "grad_norm": 2.5082983553904974,
      "learning_rate": 3.8080113273820024e-06,
      "loss": 0.3852,
      "step": 1495
    },
    {
      "epoch": 0.6211180124223602,
      "grad_norm": 2.5252781128593114,
      "learning_rate": 3.7729413837903285e-06,
      "loss": 0.393,
      "step": 1500
    },
    {
      "epoch": 0.6231884057971014,
      "grad_norm": 2.4163618496590544,
      "learning_rate": 3.7379355587526543e-06,
      "loss": 0.403,
      "step": 1505
    },
    {
      "epoch": 0.6252587991718427,
      "grad_norm": 2.4485095808931074,
      "learning_rate": 3.702995681458605e-06,
      "loss": 0.407,
      "step": 1510
    },
    {
      "epoch": 0.6273291925465838,
      "grad_norm": 2.2696712829578622,
      "learning_rate": 3.668123577651773e-06,
      "loss": 0.4,
      "step": 1515
    },
    {
      "epoch": 0.629399585921325,
      "grad_norm": 2.3377005487919127,
      "learning_rate": 3.633321069534325e-06,
      "loss": 0.4025,
      "step": 1520
    },
    {
      "epoch": 0.6314699792960663,
      "grad_norm": 2.4763791943488807,
      "learning_rate": 3.5985899756717872e-06,
      "loss": 0.4079,
      "step": 1525
    },
    {
      "epoch": 0.6335403726708074,
      "grad_norm": 2.4427448992246954,
      "learning_rate": 3.563932110898015e-06,
      "loss": 0.3965,
      "step": 1530
    },
    {
      "epoch": 0.6356107660455487,
      "grad_norm": 2.4062512467642483,
      "learning_rate": 3.5293492862203595e-06,
      "loss": 0.3867,
      "step": 1535
    },
    {
      "epoch": 0.6376811594202898,
      "grad_norm": 2.320844833466471,
      "learning_rate": 3.494843308725032e-06,
      "loss": 0.3919,
      "step": 1540
    },
    {
      "epoch": 0.639751552795031,
      "grad_norm": 2.388654503556393,
      "learning_rate": 3.460415981482691e-06,
      "loss": 0.3917,
      "step": 1545
    },
    {
      "epoch": 0.6418219461697723,
      "grad_norm": 2.477382899154321,
      "learning_rate": 3.426069103454209e-06,
      "loss": 0.3912,
      "step": 1550
    },
    {
      "epoch": 0.6438923395445134,
      "grad_norm": 2.532588346477511,
      "learning_rate": 3.3918044693966802e-06,
      "loss": 0.3895,
      "step": 1555
    },
    {
      "epoch": 0.6459627329192547,
      "grad_norm": 2.432564497678655,
      "learning_rate": 3.357623869769628e-06,
      "loss": 0.3941,
      "step": 1560
    },
    {
      "epoch": 0.6480331262939959,
      "grad_norm": 2.4869615373260765,
      "learning_rate": 3.3235290906414575e-06,
      "loss": 0.3929,
      "step": 1565
    },
    {
      "epoch": 0.650103519668737,
      "grad_norm": 2.53125677103149,
      "learning_rate": 3.289521913596117e-06,
      "loss": 0.3888,
      "step": 1570
    },
    {
      "epoch": 0.6521739130434783,
      "grad_norm": 2.3998855509485812,
      "learning_rate": 3.255604115640014e-06,
      "loss": 0.3778,
      "step": 1575
    },
    {
      "epoch": 0.6542443064182195,
      "grad_norm": 2.5025809525184277,
      "learning_rate": 3.2217774691091437e-06,
      "loss": 0.3856,
      "step": 1580
    },
    {
      "epoch": 0.6563146997929606,
      "grad_norm": 2.4192258044044976,
      "learning_rate": 3.1880437415764936e-06,
      "loss": 0.3825,
      "step": 1585
    },
    {
      "epoch": 0.6583850931677019,
      "grad_norm": 2.517380764270827,
      "learning_rate": 3.1544046957596746e-06,
      "loss": 0.3857,
      "step": 1590
    },
    {
      "epoch": 0.660455486542443,
      "grad_norm": 2.3289865674318526,
      "learning_rate": 3.1208620894288105e-06,
      "loss": 0.3754,
      "step": 1595
    },
    {
      "epoch": 0.6625258799171843,
      "grad_norm": 2.411312312267412,
      "learning_rate": 3.087417675314687e-06,
      "loss": 0.3866,
      "step": 1600
    },
    {
      "epoch": 0.6645962732919255,
      "grad_norm": 2.5129201240721004,
      "learning_rate": 3.0540732010171704e-06,
      "loss": 0.394,
      "step": 1605
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 2.418471297283698,
      "learning_rate": 3.0208304089138873e-06,
      "loss": 0.3935,
      "step": 1610
    },
    {
      "epoch": 0.6687370600414079,
      "grad_norm": 2.4403467679412985,
      "learning_rate": 2.9876910360691707e-06,
      "loss": 0.3923,
      "step": 1615
    },
    {
      "epoch": 0.6708074534161491,
      "grad_norm": 2.4851198009596343,
      "learning_rate": 2.9546568141433007e-06,
      "loss": 0.3755,
      "step": 1620
    },
    {
      "epoch": 0.6728778467908902,
      "grad_norm": 2.3242099004881585,
      "learning_rate": 2.921729469302018e-06,
      "loss": 0.3845,
      "step": 1625
    },
    {
      "epoch": 0.6749482401656315,
      "grad_norm": 2.478904355258984,
      "learning_rate": 2.8889107221263203e-06,
      "loss": 0.3796,
      "step": 1630
    },
    {
      "epoch": 0.6770186335403726,
      "grad_norm": 2.42867764849453,
      "learning_rate": 2.856202287522556e-06,
      "loss": 0.3804,
      "step": 1635
    },
    {
      "epoch": 0.6790890269151139,
      "grad_norm": 2.307301703407307,
      "learning_rate": 2.8236058746328203e-06,
      "loss": 0.37,
      "step": 1640
    },
    {
      "epoch": 0.6811594202898551,
      "grad_norm": 2.5744127565219586,
      "learning_rate": 2.7911231867456345e-06,
      "loss": 0.3862,
      "step": 1645
    },
    {
      "epoch": 0.6832298136645962,
      "grad_norm": 2.3479644350997466,
      "learning_rate": 2.7587559212069554e-06,
      "loss": 0.3889,
      "step": 1650
    },
    {
      "epoch": 0.6853002070393375,
      "grad_norm": 2.5447600622383835,
      "learning_rate": 2.72650576933147e-06,
      "loss": 0.3724,
      "step": 1655
    },
    {
      "epoch": 0.6873706004140787,
      "grad_norm": 2.3578779353515453,
      "learning_rate": 2.6943744163142292e-06,
      "loss": 0.3718,
      "step": 1660
    },
    {
      "epoch": 0.6894409937888198,
      "grad_norm": 2.263601488644489,
      "learning_rate": 2.662363541142585e-06,
      "loss": 0.3684,
      "step": 1665
    },
    {
      "epoch": 0.6915113871635611,
      "grad_norm": 2.330454074745477,
      "learning_rate": 2.6304748165084503e-06,
      "loss": 0.364,
      "step": 1670
    },
    {
      "epoch": 0.6935817805383023,
      "grad_norm": 2.2702455387553493,
      "learning_rate": 2.5987099087209035e-06,
      "loss": 0.3632,
      "step": 1675
    },
    {
      "epoch": 0.6956521739130435,
      "grad_norm": 2.2711128268754726,
      "learning_rate": 2.5670704776191158e-06,
      "loss": 0.3724,
      "step": 1680
    },
    {
      "epoch": 0.6977225672877847,
      "grad_norm": 2.3743576993945164,
      "learning_rate": 2.5355581764856186e-06,
      "loss": 0.364,
      "step": 1685
    },
    {
      "epoch": 0.6997929606625258,
      "grad_norm": 2.321783997397547,
      "learning_rate": 2.5041746519599036e-06,
      "loss": 0.3746,
      "step": 1690
    },
    {
      "epoch": 0.7018633540372671,
      "grad_norm": 2.3385931927779495,
      "learning_rate": 2.4729215439523945e-06,
      "loss": 0.3668,
      "step": 1695
    },
    {
      "epoch": 0.7039337474120083,
      "grad_norm": 2.454951833218205,
      "learning_rate": 2.441800485558748e-06,
      "loss": 0.3642,
      "step": 1700
    },
    {
      "epoch": 0.7060041407867494,
      "grad_norm": 2.4177454976507997,
      "learning_rate": 2.410813102974512e-06,
      "loss": 0.3693,
      "step": 1705
    },
    {
      "epoch": 0.7080745341614907,
      "grad_norm": 2.343959459554647,
      "learning_rate": 2.3799610154101603e-06,
      "loss": 0.3603,
      "step": 1710
    },
    {
      "epoch": 0.7101449275362319,
      "grad_norm": 2.4564454853030746,
      "learning_rate": 2.3492458350064805e-06,
      "loss": 0.3683,
      "step": 1715
    },
    {
      "epoch": 0.7122153209109731,
      "grad_norm": 2.294885451688505,
      "learning_rate": 2.3186691667503315e-06,
      "loss": 0.3755,
      "step": 1720
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 2.5144966859161983,
      "learning_rate": 2.288232608390778e-06,
      "loss": 0.3704,
      "step": 1725
    },
    {
      "epoch": 0.7163561076604554,
      "grad_norm": 2.3221968894149287,
      "learning_rate": 2.2579377503555967e-06,
      "loss": 0.3649,
      "step": 1730
    },
    {
      "epoch": 0.7184265010351967,
      "grad_norm": 2.2316884265787555,
      "learning_rate": 2.227786175668181e-06,
      "loss": 0.362,
      "step": 1735
    },
    {
      "epoch": 0.7204968944099379,
      "grad_norm": 2.2783681292524096,
      "learning_rate": 2.1977794598648106e-06,
      "loss": 0.3719,
      "step": 1740
    },
    {
      "epoch": 0.722567287784679,
      "grad_norm": 2.2718482289334814,
      "learning_rate": 2.167919170912333e-06,
      "loss": 0.3722,
      "step": 1745
    },
    {
      "epoch": 0.7246376811594203,
      "grad_norm": 2.3280044977557366,
      "learning_rate": 2.138206869126222e-06,
      "loss": 0.3456,
      "step": 1750
    },
    {
      "epoch": 0.7267080745341615,
      "grad_norm": 2.429072655182024,
      "learning_rate": 2.108644107089055e-06,
      "loss": 0.365,
      "step": 1755
    },
    {
      "epoch": 0.7287784679089027,
      "grad_norm": 2.288093361852132,
      "learning_rate": 2.0792324295693782e-06,
      "loss": 0.3527,
      "step": 1760
    },
    {
      "epoch": 0.7308488612836439,
      "grad_norm": 2.403717546765947,
      "learning_rate": 2.0499733734409884e-06,
      "loss": 0.3704,
      "step": 1765
    },
    {
      "epoch": 0.7329192546583851,
      "grad_norm": 2.358248905812587,
      "learning_rate": 2.020868467602623e-06,
      "loss": 0.3617,
      "step": 1770
    },
    {
      "epoch": 0.7349896480331263,
      "grad_norm": 2.257158135310934,
      "learning_rate": 1.9919192328980703e-06,
      "loss": 0.3619,
      "step": 1775
    },
    {
      "epoch": 0.7370600414078675,
      "grad_norm": 2.3892307889716218,
      "learning_rate": 1.9631271820367058e-06,
      "loss": 0.3594,
      "step": 1780
    },
    {
      "epoch": 0.7391304347826086,
      "grad_norm": 2.3381779133079545,
      "learning_rate": 1.9344938195144353e-06,
      "loss": 0.3669,
      "step": 1785
    },
    {
      "epoch": 0.7412008281573499,
      "grad_norm": 2.404083527843501,
      "learning_rate": 1.9060206415350896e-06,
      "loss": 0.3529,
      "step": 1790
    },
    {
      "epoch": 0.7432712215320911,
      "grad_norm": 2.294713286248942,
      "learning_rate": 1.8777091359322386e-06,
      "loss": 0.3461,
      "step": 1795
    },
    {
      "epoch": 0.7453416149068323,
      "grad_norm": 2.541463332479839,
      "learning_rate": 1.8495607820914451e-06,
      "loss": 0.3523,
      "step": 1800
    },
    {
      "epoch": 0.7474120082815735,
      "grad_norm": 2.3026780271234992,
      "learning_rate": 1.8215770508729602e-06,
      "loss": 0.3371,
      "step": 1805
    },
    {
      "epoch": 0.7494824016563147,
      "grad_norm": 2.287962750587562,
      "learning_rate": 1.7937594045348666e-06,
      "loss": 0.3455,
      "step": 1810
    },
    {
      "epoch": 0.7515527950310559,
      "grad_norm": 2.2990081929334885,
      "learning_rate": 1.7661092966566728e-06,
      "loss": 0.3515,
      "step": 1815
    },
    {
      "epoch": 0.7536231884057971,
      "grad_norm": 2.4526981280198656,
      "learning_rate": 1.7386281720633569e-06,
      "loss": 0.349,
      "step": 1820
    },
    {
      "epoch": 0.7556935817805382,
      "grad_norm": 2.3254078483771754,
      "learning_rate": 1.7113174667498612e-06,
      "loss": 0.3514,
      "step": 1825
    },
    {
      "epoch": 0.7577639751552795,
      "grad_norm": 2.29073556555259,
      "learning_rate": 1.6841786078060674e-06,
      "loss": 0.3441,
      "step": 1830
    },
    {
      "epoch": 0.7598343685300207,
      "grad_norm": 2.4712240180219234,
      "learning_rate": 1.6572130133422204e-06,
      "loss": 0.3498,
      "step": 1835
    },
    {
      "epoch": 0.7619047619047619,
      "grad_norm": 2.414467739041196,
      "learning_rate": 1.630422092414823e-06,
      "loss": 0.3489,
      "step": 1840
    },
    {
      "epoch": 0.7639751552795031,
      "grad_norm": 2.302312716480218,
      "learning_rate": 1.6038072449530118e-06,
      "loss": 0.3501,
      "step": 1845
    },
    {
      "epoch": 0.7660455486542443,
      "grad_norm": 2.450748622606659,
      "learning_rate": 1.5773698616854065e-06,
      "loss": 0.3532,
      "step": 1850
    },
    {
      "epoch": 0.7681159420289855,
      "grad_norm": 2.449347937864923,
      "learning_rate": 1.5511113240674364e-06,
      "loss": 0.3466,
      "step": 1855
    },
    {
      "epoch": 0.7701863354037267,
      "grad_norm": 2.291726324349712,
      "learning_rate": 1.5250330042091543e-06,
      "loss": 0.3463,
      "step": 1860
    },
    {
      "epoch": 0.772256728778468,
      "grad_norm": 2.3236010903958464,
      "learning_rate": 1.4991362648035368e-06,
      "loss": 0.3464,
      "step": 1865
    },
    {
      "epoch": 0.7743271221532091,
      "grad_norm": 2.419100172109552,
      "learning_rate": 1.4734224590552844e-06,
      "loss": 0.3394,
      "step": 1870
    },
    {
      "epoch": 0.7763975155279503,
      "grad_norm": 2.5063677477126145,
      "learning_rate": 1.4478929306101025e-06,
      "loss": 0.354,
      "step": 1875
    },
    {
      "epoch": 0.7784679089026915,
      "grad_norm": 2.359552483635086,
      "learning_rate": 1.4225490134844967e-06,
      "loss": 0.3389,
      "step": 1880
    },
    {
      "epoch": 0.7805383022774327,
      "grad_norm": 2.389057465344352,
      "learning_rate": 1.3973920319960654e-06,
      "loss": 0.3464,
      "step": 1885
    },
    {
      "epoch": 0.782608695652174,
      "grad_norm": 2.56591176495094,
      "learning_rate": 1.3724233006942972e-06,
      "loss": 0.3401,
      "step": 1890
    },
    {
      "epoch": 0.7846790890269151,
      "grad_norm": 2.353473827366181,
      "learning_rate": 1.347644124291883e-06,
      "loss": 0.3349,
      "step": 1895
    },
    {
      "epoch": 0.7867494824016563,
      "grad_norm": 2.3922132511834775,
      "learning_rate": 1.323055797596532e-06,
      "loss": 0.3443,
      "step": 1900
    },
    {
      "epoch": 0.7888198757763976,
      "grad_norm": 2.321235848929393,
      "learning_rate": 1.2986596054433255e-06,
      "loss": 0.3396,
      "step": 1905
    },
    {
      "epoch": 0.7908902691511387,
      "grad_norm": 2.33355008688893,
      "learning_rate": 1.2744568226275662e-06,
      "loss": 0.3373,
      "step": 1910
    },
    {
      "epoch": 0.7929606625258799,
      "grad_norm": 2.375277589026336,
      "learning_rate": 1.2504487138381782e-06,
      "loss": 0.3397,
      "step": 1915
    },
    {
      "epoch": 0.7950310559006211,
      "grad_norm": 2.3311427058270313,
      "learning_rate": 1.2266365335916086e-06,
      "loss": 0.3328,
      "step": 1920
    },
    {
      "epoch": 0.7971014492753623,
      "grad_norm": 2.303375666326696,
      "learning_rate": 1.203021526166287e-06,
      "loss": 0.3413,
      "step": 1925
    },
    {
      "epoch": 0.7991718426501035,
      "grad_norm": 2.363927557987927,
      "learning_rate": 1.179604925537598e-06,
      "loss": 0.3345,
      "step": 1930
    },
    {
      "epoch": 0.8012422360248447,
      "grad_norm": 2.375976453459839,
      "learning_rate": 1.1563879553134072e-06,
      "loss": 0.3325,
      "step": 1935
    },
    {
      "epoch": 0.8033126293995859,
      "grad_norm": 2.3880275309939365,
      "learning_rate": 1.1333718286701162e-06,
      "loss": 0.3467,
      "step": 1940
    },
    {
      "epoch": 0.8053830227743272,
      "grad_norm": 2.2832719512904323,
      "learning_rate": 1.110557748289275e-06,
      "loss": 0.3288,
      "step": 1945
    },
    {
      "epoch": 0.8074534161490683,
      "grad_norm": 2.650603625119056,
      "learning_rate": 1.0879469062947369e-06,
      "loss": 0.3445,
      "step": 1950
    },
    {
      "epoch": 0.8095238095238095,
      "grad_norm": 2.3557690102944124,
      "learning_rate": 1.0655404841903648e-06,
      "loss": 0.3378,
      "step": 1955
    },
    {
      "epoch": 0.8115942028985508,
      "grad_norm": 2.3935304302890525,
      "learning_rate": 1.0433396527982893e-06,
      "loss": 0.3413,
      "step": 1960
    },
    {
      "epoch": 0.8136645962732919,
      "grad_norm": 2.2158689019496167,
      "learning_rate": 1.0213455721977356e-06,
      "loss": 0.3353,
      "step": 1965
    },
    {
      "epoch": 0.8157349896480331,
      "grad_norm": 2.4076687892725923,
      "learning_rate": 9.995593916643981e-07,
      "loss": 0.3408,
      "step": 1970
    },
    {
      "epoch": 0.8178053830227743,
      "grad_norm": 2.34671395844793,
      "learning_rate": 9.779822496103902e-07,
      "loss": 0.3398,
      "step": 1975
    },
    {
      "epoch": 0.8198757763975155,
      "grad_norm": 2.333807147075891,
      "learning_rate": 9.566152735247576e-07,
      "loss": 0.3201,
      "step": 1980
    },
    {
      "epoch": 0.8219461697722568,
      "grad_norm": 2.48104141477904,
      "learning_rate": 9.354595799145627e-07,
      "loss": 0.329,
      "step": 1985
    },
    {
      "epoch": 0.8240165631469979,
      "grad_norm": 2.286948319254527,
      "learning_rate": 9.145162742465424e-07,
      "loss": 0.3215,
      "step": 1990
    },
    {
      "epoch": 0.8260869565217391,
      "grad_norm": 2.294781106101465,
      "learning_rate": 8.937864508893396e-07,
      "loss": 0.3451,
      "step": 1995
    },
    {
      "epoch": 0.8281573498964804,
      "grad_norm": 2.3419090436451317,
      "learning_rate": 8.732711930563259e-07,
      "loss": 0.3341,
      "step": 2000
    },
    {
      "epoch": 0.8302277432712215,
      "grad_norm": 2.3213935472448313,
      "learning_rate": 8.529715727489912e-07,
      "loss": 0.3223,
      "step": 2005
    },
    {
      "epoch": 0.8322981366459627,
      "grad_norm": 2.292958375067994,
      "learning_rate": 8.32888650700936e-07,
      "loss": 0.3309,
      "step": 2010
    },
    {
      "epoch": 0.8343685300207039,
      "grad_norm": 2.2875579360410376,
      "learning_rate": 8.130234763224342e-07,
      "loss": 0.3277,
      "step": 2015
    },
    {
      "epoch": 0.8364389233954451,
      "grad_norm": 2.5084450712201,
      "learning_rate": 7.93377087645607e-07,
      "loss": 0.3212,
      "step": 2020
    },
    {
      "epoch": 0.8385093167701864,
      "grad_norm": 2.3607289166265164,
      "learning_rate": 7.739505112701751e-07,
      "loss": 0.3211,
      "step": 2025
    },
    {
      "epoch": 0.8405797101449275,
      "grad_norm": 2.2928319848853276,
      "learning_rate": 7.547447623098191e-07,
      "loss": 0.3307,
      "step": 2030
    },
    {
      "epoch": 0.8426501035196687,
      "grad_norm": 2.3048158294478696,
      "learning_rate": 7.357608443391312e-07,
      "loss": 0.3328,
      "step": 2035
    },
    {
      "epoch": 0.84472049689441,
      "grad_norm": 2.4088323317177793,
      "learning_rate": 7.169997493411774e-07,
      "loss": 0.3315,
      "step": 2040
    },
    {
      "epoch": 0.8467908902691511,
      "grad_norm": 2.144393386573033,
      "learning_rate": 6.984624576556647e-07,
      "loss": 0.3288,
      "step": 2045
    },
    {
      "epoch": 0.8488612836438924,
      "grad_norm": 2.5243055903104574,
      "learning_rate": 6.801499379277115e-07,
      "loss": 0.3197,
      "step": 2050
    },
    {
      "epoch": 0.8509316770186336,
      "grad_norm": 2.4162988014293436,
      "learning_rate": 6.620631470572314e-07,
      "loss": 0.3229,
      "step": 2055
    },
    {
      "epoch": 0.8530020703933747,
      "grad_norm": 2.3234003177788307,
      "learning_rate": 6.442030301489338e-07,
      "loss": 0.3283,
      "step": 2060
    },
    {
      "epoch": 0.855072463768116,
      "grad_norm": 2.412583394523917,
      "learning_rate": 6.265705204629402e-07,
      "loss": 0.3155,
      "step": 2065
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 2.1849577042614774,
      "learning_rate": 6.091665393660112e-07,
      "loss": 0.3199,
      "step": 2070
    },
    {
      "epoch": 0.8592132505175983,
      "grad_norm": 2.4623418090724414,
      "learning_rate": 5.919919962834069e-07,
      "loss": 0.3262,
      "step": 2075
    },
    {
      "epoch": 0.8612836438923396,
      "grad_norm": 2.3034907256570927,
      "learning_rate": 5.750477886513656e-07,
      "loss": 0.3381,
      "step": 2080
    },
    {
      "epoch": 0.8633540372670807,
      "grad_norm": 2.370216659385342,
      "learning_rate": 5.58334801870209e-07,
      "loss": 0.3261,
      "step": 2085
    },
    {
      "epoch": 0.865424430641822,
      "grad_norm": 2.3509467079102415,
      "learning_rate": 5.418539092580727e-07,
      "loss": 0.3167,
      "step": 2090
    },
    {
      "epoch": 0.8674948240165632,
      "grad_norm": 2.532187145684702,
      "learning_rate": 5.256059720052787e-07,
      "loss": 0.3335,
      "step": 2095
    },
    {
      "epoch": 0.8695652173913043,
      "grad_norm": 2.2570775053057757,
      "learning_rate": 5.095918391293303e-07,
      "loss": 0.3266,
      "step": 2100
    },
    {
      "epoch": 0.8716356107660456,
      "grad_norm": 2.6011146727795453,
      "learning_rate": 4.938123474305473e-07,
      "loss": 0.314,
      "step": 2105
    },
    {
      "epoch": 0.8737060041407867,
      "grad_norm": 2.3499523752616045,
      "learning_rate": 4.782683214483413e-07,
      "loss": 0.3199,
      "step": 2110
    },
    {
      "epoch": 0.8757763975155279,
      "grad_norm": 2.2824953520044167,
      "learning_rate": 4.629605734181331e-07,
      "loss": 0.3243,
      "step": 2115
    },
    {
      "epoch": 0.8778467908902692,
      "grad_norm": 2.395624713850349,
      "learning_rate": 4.478899032289047e-07,
      "loss": 0.3303,
      "step": 2120
    },
    {
      "epoch": 0.8799171842650103,
      "grad_norm": 2.3740512007356913,
      "learning_rate": 4.3305709838140755e-07,
      "loss": 0.324,
      "step": 2125
    },
    {
      "epoch": 0.8819875776397516,
      "grad_norm": 2.34060899939904,
      "learning_rate": 4.1846293394700645e-07,
      "loss": 0.3146,
      "step": 2130
    },
    {
      "epoch": 0.8840579710144928,
      "grad_norm": 2.528957480954012,
      "learning_rate": 4.041081725271856e-07,
      "loss": 0.3271,
      "step": 2135
    },
    {
      "epoch": 0.8861283643892339,
      "grad_norm": 2.367140360672427,
      "learning_rate": 3.8999356421369426e-07,
      "loss": 0.3195,
      "step": 2140
    },
    {
      "epoch": 0.8881987577639752,
      "grad_norm": 2.3097614396359707,
      "learning_rate": 3.7611984654935707e-07,
      "loss": 0.3239,
      "step": 2145
    },
    {
      "epoch": 0.8902691511387164,
      "grad_norm": 2.5183764481920923,
      "learning_rate": 3.6248774448952695e-07,
      "loss": 0.3153,
      "step": 2150
    },
    {
      "epoch": 0.8923395445134575,
      "grad_norm": 2.586353838926294,
      "learning_rate": 3.490979703642117e-07,
      "loss": 0.3382,
      "step": 2155
    },
    {
      "epoch": 0.8944099378881988,
      "grad_norm": 2.3896229838176786,
      "learning_rate": 3.35951223840848e-07,
      "loss": 0.3178,
      "step": 2160
    },
    {
      "epoch": 0.8964803312629399,
      "grad_norm": 2.487395540010883,
      "learning_rate": 3.2304819188773815e-07,
      "loss": 0.3276,
      "step": 2165
    },
    {
      "epoch": 0.8985507246376812,
      "grad_norm": 2.3790497808037374,
      "learning_rate": 3.1038954873816027e-07,
      "loss": 0.31,
      "step": 2170
    },
    {
      "epoch": 0.9006211180124224,
      "grad_norm": 2.2942628007276435,
      "learning_rate": 2.9797595585512986e-07,
      "loss": 0.3214,
      "step": 2175
    },
    {
      "epoch": 0.9026915113871635,
      "grad_norm": 2.540059203143415,
      "learning_rate": 2.858080618968423e-07,
      "loss": 0.3197,
      "step": 2180
    },
    {
      "epoch": 0.9047619047619048,
      "grad_norm": 2.452337835828254,
      "learning_rate": 2.738865026827714e-07,
      "loss": 0.3273,
      "step": 2185
    },
    {
      "epoch": 0.906832298136646,
      "grad_norm": 2.26680030771922,
      "learning_rate": 2.6221190116045126e-07,
      "loss": 0.3082,
      "step": 2190
    },
    {
      "epoch": 0.9089026915113871,
      "grad_norm": 2.3996079968188635,
      "learning_rate": 2.507848673729224e-07,
      "loss": 0.3265,
      "step": 2195
    },
    {
      "epoch": 0.9109730848861284,
      "grad_norm": 2.458176828695482,
      "learning_rate": 2.3960599842685394e-07,
      "loss": 0.3224,
      "step": 2200
    },
    {
      "epoch": 0.9130434782608695,
      "grad_norm": 2.3483468753530685,
      "learning_rate": 2.2867587846134188e-07,
      "loss": 0.3099,
      "step": 2205
    },
    {
      "epoch": 0.9151138716356108,
      "grad_norm": 2.1887293353292803,
      "learning_rate": 2.179950786173879e-07,
      "loss": 0.3077,
      "step": 2210
    },
    {
      "epoch": 0.917184265010352,
      "grad_norm": 2.3872010256944503,
      "learning_rate": 2.07564157008055e-07,
      "loss": 0.3205,
      "step": 2215
    },
    {
      "epoch": 0.9192546583850931,
      "grad_norm": 2.374962965449597,
      "learning_rate": 1.9738365868930188e-07,
      "loss": 0.3179,
      "step": 2220
    },
    {
      "epoch": 0.9213250517598344,
      "grad_norm": 2.34765412276636,
      "learning_rate": 1.8745411563150128e-07,
      "loss": 0.3118,
      "step": 2225
    },
    {
      "epoch": 0.9233954451345756,
      "grad_norm": 2.384272001437427,
      "learning_rate": 1.7777604669164727e-07,
      "loss": 0.3191,
      "step": 2230
    },
    {
      "epoch": 0.9254658385093167,
      "grad_norm": 2.56561694406812,
      "learning_rate": 1.6834995758623685e-07,
      "loss": 0.3241,
      "step": 2235
    },
    {
      "epoch": 0.927536231884058,
      "grad_norm": 2.300312987212365,
      "learning_rate": 1.5917634086484778e-07,
      "loss": 0.3149,
      "step": 2240
    },
    {
      "epoch": 0.9296066252587992,
      "grad_norm": 2.35456052793434,
      "learning_rate": 1.5025567588439925e-07,
      "loss": 0.3198,
      "step": 2245
    },
    {
      "epoch": 0.9316770186335404,
      "grad_norm": 2.220250392327445,
      "learning_rate": 1.415884287841074e-07,
      "loss": 0.3184,
      "step": 2250
    },
    {
      "epoch": 0.9337474120082816,
      "grad_norm": 2.3466072873242214,
      "learning_rate": 1.331750524611225e-07,
      "loss": 0.3186,
      "step": 2255
    },
    {
      "epoch": 0.9358178053830227,
      "grad_norm": 2.2289914595941562,
      "learning_rate": 1.2501598654686532e-07,
      "loss": 0.3129,
      "step": 2260
    },
    {
      "epoch": 0.937888198757764,
      "grad_norm": 2.3299165520939,
      "learning_rate": 1.1711165738405749e-07,
      "loss": 0.3215,
      "step": 2265
    },
    {
      "epoch": 0.9399585921325052,
      "grad_norm": 2.358226508179777,
      "learning_rate": 1.0946247800443899e-07,
      "loss": 0.3317,
      "step": 2270
    },
    {
      "epoch": 0.9420289855072463,
      "grad_norm": 2.506547423370588,
      "learning_rate": 1.0206884810718964e-07,
      "loss": 0.3324,
      "step": 2275
    },
    {
      "epoch": 0.9440993788819876,
      "grad_norm": 2.4345957750296097,
      "learning_rate": 9.493115403803876e-08,
      "loss": 0.3141,
      "step": 2280
    },
    {
      "epoch": 0.9461697722567288,
      "grad_norm": 2.4966050771543555,
      "learning_rate": 8.804976876908178e-08,
      "loss": 0.3193,
      "step": 2285
    },
    {
      "epoch": 0.94824016563147,
      "grad_norm": 2.3773068455485227,
      "learning_rate": 8.142505187928861e-08,
      "loss": 0.3286,
      "step": 2290
    },
    {
      "epoch": 0.9503105590062112,
      "grad_norm": 2.36270388623886,
      "learning_rate": 7.505734953571486e-08,
      "loss": 0.3168,
      "step": 2295
    },
    {
      "epoch": 0.9523809523809523,
      "grad_norm": 2.1753275617985537,
      "learning_rate": 6.894699447541231e-08,
      "loss": 0.3166,
      "step": 2300
    },
    {
      "epoch": 0.9544513457556936,
      "grad_norm": 2.5830239699043713,
      "learning_rate": 6.309430598804234e-08,
      "loss": 0.3208,
      "step": 2305
    },
    {
      "epoch": 0.9565217391304348,
      "grad_norm": 2.3442757362694153,
      "learning_rate": 5.7499589899193134e-08,
      "loss": 0.3179,
      "step": 2310
    },
    {
      "epoch": 0.9585921325051759,
      "grad_norm": 2.575291888361898,
      "learning_rate": 5.216313855439858e-08,
      "loss": 0.3181,
      "step": 2315
    },
    {
      "epoch": 0.9606625258799172,
      "grad_norm": 2.372429648001184,
      "learning_rate": 4.7085230803860513e-08,
      "loss": 0.3125,
      "step": 2320
    },
    {
      "epoch": 0.9627329192546584,
      "grad_norm": 2.3903157282944507,
      "learning_rate": 4.2266131987880876e-08,
      "loss": 0.3199,
      "step": 2325
    },
    {
      "epoch": 0.9648033126293996,
      "grad_norm": 2.629984381467656,
      "learning_rate": 3.770609392299285e-08,
      "loss": 0.3273,
      "step": 2330
    },
    {
      "epoch": 0.9668737060041408,
      "grad_norm": 2.2856042279572666,
      "learning_rate": 3.340535488880525e-08,
      "loss": 0.3109,
      "step": 2335
    },
    {
      "epoch": 0.968944099378882,
      "grad_norm": 2.399447582602549,
      "learning_rate": 2.9364139615550846e-08,
      "loss": 0.3174,
      "step": 2340
    },
    {
      "epoch": 0.9710144927536232,
      "grad_norm": 2.3704897505981566,
      "learning_rate": 2.5582659272341827e-08,
      "loss": 0.3152,
      "step": 2345
    },
    {
      "epoch": 0.9730848861283644,
      "grad_norm": 2.447121307306242,
      "learning_rate": 2.206111145613865e-08,
      "loss": 0.3049,
      "step": 2350
    },
    {
      "epoch": 0.9751552795031055,
      "grad_norm": 2.604309573240356,
      "learning_rate": 1.8799680181421087e-08,
      "loss": 0.3284,
      "step": 2355
    },
    {
      "epoch": 0.9772256728778468,
      "grad_norm": 2.3870370176370326,
      "learning_rate": 1.5798535870575893e-08,
      "loss": 0.3099,
      "step": 2360
    },
    {
      "epoch": 0.979296066252588,
      "grad_norm": 2.510927340719228,
      "learning_rate": 1.30578353449895e-08,
      "loss": 0.3164,
      "step": 2365
    },
    {
      "epoch": 0.9813664596273292,
      "grad_norm": 2.4135910096530284,
      "learning_rate": 1.0577721816854015e-08,
      "loss": 0.3147,
      "step": 2370
    },
    {
      "epoch": 0.9834368530020704,
      "grad_norm": 2.48975469369635,
      "learning_rate": 8.35832488168431e-09,
      "loss": 0.3178,
      "step": 2375
    },
    {
      "epoch": 0.9855072463768116,
      "grad_norm": 2.471436091156265,
      "learning_rate": 6.399760511546227e-09,
      "loss": 0.3244,
      "step": 2380
    },
    {
      "epoch": 0.9875776397515528,
      "grad_norm": 2.2547600080916315,
      "learning_rate": 4.7021310489953065e-09,
      "loss": 0.3155,
      "step": 2385
    },
    {
      "epoch": 0.989648033126294,
      "grad_norm": 2.3332031612865873,
      "learning_rate": 3.265525201731623e-09,
      "loss": 0.3049,
      "step": 2390
    },
    {
      "epoch": 0.9917184265010351,
      "grad_norm": 2.3293110535007386,
      "learning_rate": 2.090018037960717e-09,
      "loss": 0.3333,
      "step": 2395
    },
    {
      "epoch": 0.9937888198757764,
      "grad_norm": 2.4045952132798174,
      "learning_rate": 1.1756709824750633e-09,
      "loss": 0.3175,
      "step": 2400
    },
    {
      "epoch": 0.9958592132505176,
      "grad_norm": 2.495730153719708,
      "learning_rate": 5.225318134399749e-10,
      "loss": 0.3136,
      "step": 2405
    },
    {
      "epoch": 0.9979296066252588,
      "grad_norm": 2.4326500702237412,
      "learning_rate": 1.306346599011521e-10,
      "loss": 0.3058,
      "step": 2410
    },
    {
      "epoch": 1.0,
      "grad_norm": 2.385291628188146,
      "learning_rate": 0.0,
      "loss": 0.313,
      "step": 2415
    },
    {
      "epoch": 1.0,
      "eval_runtime": 3.4442,
      "eval_samples_per_second": 2.903,
      "eval_steps_per_second": 0.871,
      "step": 2415
    },
    {
      "epoch": 1.0,
      "step": 2415,
      "total_flos": 252825934233600.0,
      "train_loss": 0.5146494265175261,
      "train_runtime": 21469.3859,
      "train_samples_per_second": 1.799,
      "train_steps_per_second": 0.112
    }
  ],
  "logging_steps": 5,
  "max_steps": 2415,
  "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": 252825934233600.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}