{
  "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": 24.082783987617052,
      "learning_rate": 4.132231404958678e-08,
      "loss": 1.3915,
      "step": 1
    },
    {
      "epoch": 0.002070393374741201,
      "grad_norm": 24.927352059558704,
      "learning_rate": 2.066115702479339e-07,
      "loss": 1.4137,
      "step": 5
    },
    {
      "epoch": 0.004140786749482402,
      "grad_norm": 15.956520007454369,
      "learning_rate": 4.132231404958678e-07,
      "loss": 1.3648,
      "step": 10
    },
    {
      "epoch": 0.006211180124223602,
      "grad_norm": 8.550349209845118,
      "learning_rate": 6.198347107438018e-07,
      "loss": 1.2504,
      "step": 15
    },
    {
      "epoch": 0.008281573498964804,
      "grad_norm": 9.55086086758466,
      "learning_rate": 8.264462809917356e-07,
      "loss": 1.1392,
      "step": 20
    },
    {
      "epoch": 0.010351966873706004,
      "grad_norm": 5.927325777610998,
      "learning_rate": 1.0330578512396695e-06,
      "loss": 1.0294,
      "step": 25
    },
    {
      "epoch": 0.012422360248447204,
      "grad_norm": 3.3775223586374876,
      "learning_rate": 1.2396694214876035e-06,
      "loss": 0.9764,
      "step": 30
    },
    {
      "epoch": 0.014492753623188406,
      "grad_norm": 3.366952648284076,
      "learning_rate": 1.4462809917355372e-06,
      "loss": 0.9499,
      "step": 35
    },
    {
      "epoch": 0.016563146997929608,
      "grad_norm": 3.138423239660313,
      "learning_rate": 1.6528925619834712e-06,
      "loss": 0.9269,
      "step": 40
    },
    {
      "epoch": 0.018633540372670808,
      "grad_norm": 2.979310152009545,
      "learning_rate": 1.859504132231405e-06,
      "loss": 0.9134,
      "step": 45
    },
    {
      "epoch": 0.020703933747412008,
      "grad_norm": 3.0241182130041913,
      "learning_rate": 2.066115702479339e-06,
      "loss": 0.9133,
      "step": 50
    },
    {
      "epoch": 0.022774327122153208,
      "grad_norm": 3.1850362063593276,
      "learning_rate": 2.2727272727272728e-06,
      "loss": 0.8945,
      "step": 55
    },
    {
      "epoch": 0.024844720496894408,
      "grad_norm": 3.079066061647991,
      "learning_rate": 2.479338842975207e-06,
      "loss": 0.8983,
      "step": 60
    },
    {
      "epoch": 0.026915113871635612,
      "grad_norm": 3.120502951434461,
      "learning_rate": 2.6859504132231405e-06,
      "loss": 0.8919,
      "step": 65
    },
    {
      "epoch": 0.028985507246376812,
      "grad_norm": 3.054715083150286,
      "learning_rate": 2.8925619834710743e-06,
      "loss": 0.8816,
      "step": 70
    },
    {
      "epoch": 0.031055900621118012,
      "grad_norm": 3.188001221672915,
      "learning_rate": 3.0991735537190086e-06,
      "loss": 0.8766,
      "step": 75
    },
    {
      "epoch": 0.033126293995859216,
      "grad_norm": 3.0101176148714304,
      "learning_rate": 3.3057851239669424e-06,
      "loss": 0.872,
      "step": 80
    },
    {
      "epoch": 0.035196687370600416,
      "grad_norm": 3.149359365247228,
      "learning_rate": 3.5123966942148763e-06,
      "loss": 0.8496,
      "step": 85
    },
    {
      "epoch": 0.037267080745341616,
      "grad_norm": 3.120610784229853,
      "learning_rate": 3.71900826446281e-06,
      "loss": 0.8601,
      "step": 90
    },
    {
      "epoch": 0.039337474120082816,
      "grad_norm": 2.9979744537524375,
      "learning_rate": 3.925619834710744e-06,
      "loss": 0.8664,
      "step": 95
    },
    {
      "epoch": 0.041407867494824016,
      "grad_norm": 3.0423377925355783,
      "learning_rate": 4.132231404958678e-06,
      "loss": 0.8616,
      "step": 100
    },
    {
      "epoch": 0.043478260869565216,
      "grad_norm": 3.090679409633537,
      "learning_rate": 4.338842975206612e-06,
      "loss": 0.8588,
      "step": 105
    },
    {
      "epoch": 0.045548654244306416,
      "grad_norm": 3.142470242173892,
      "learning_rate": 4.5454545454545455e-06,
      "loss": 0.8665,
      "step": 110
    },
    {
      "epoch": 0.047619047619047616,
      "grad_norm": 2.9965627140856133,
      "learning_rate": 4.75206611570248e-06,
      "loss": 0.8471,
      "step": 115
    },
    {
      "epoch": 0.049689440993788817,
      "grad_norm": 2.8450247348143938,
      "learning_rate": 4.958677685950414e-06,
      "loss": 0.8411,
      "step": 120
    },
    {
      "epoch": 0.051759834368530024,
      "grad_norm": 3.1345891420193923,
      "learning_rate": 5.165289256198347e-06,
      "loss": 0.8402,
      "step": 125
    },
    {
      "epoch": 0.053830227743271224,
      "grad_norm": 3.1394429356207096,
      "learning_rate": 5.371900826446281e-06,
      "loss": 0.8435,
      "step": 130
    },
    {
      "epoch": 0.055900621118012424,
      "grad_norm": 5.113603154695779,
      "learning_rate": 5.578512396694216e-06,
      "loss": 0.827,
      "step": 135
    },
    {
      "epoch": 0.057971014492753624,
      "grad_norm": 3.1016506456626254,
      "learning_rate": 5.785123966942149e-06,
      "loss": 0.8402,
      "step": 140
    },
    {
      "epoch": 0.060041407867494824,
      "grad_norm": 2.99784578760738,
      "learning_rate": 5.991735537190083e-06,
      "loss": 0.8265,
      "step": 145
    },
    {
      "epoch": 0.062111801242236024,
      "grad_norm": 3.0494902650762623,
      "learning_rate": 6.198347107438017e-06,
      "loss": 0.8108,
      "step": 150
    },
    {
      "epoch": 0.06418219461697723,
      "grad_norm": 3.1615292574211993,
      "learning_rate": 6.404958677685951e-06,
      "loss": 0.8317,
      "step": 155
    },
    {
      "epoch": 0.06625258799171843,
      "grad_norm": 3.1074859962537067,
      "learning_rate": 6.611570247933885e-06,
      "loss": 0.8112,
      "step": 160
    },
    {
      "epoch": 0.06832298136645963,
      "grad_norm": 3.112241035749247,
      "learning_rate": 6.818181818181818e-06,
      "loss": 0.8197,
      "step": 165
    },
    {
      "epoch": 0.07039337474120083,
      "grad_norm": 3.1217916015425713,
      "learning_rate": 7.0247933884297525e-06,
      "loss": 0.8109,
      "step": 170
    },
    {
      "epoch": 0.07246376811594203,
      "grad_norm": 2.845267862276201,
      "learning_rate": 7.231404958677687e-06,
      "loss": 0.8178,
      "step": 175
    },
    {
      "epoch": 0.07453416149068323,
      "grad_norm": 3.1072753728134916,
      "learning_rate": 7.43801652892562e-06,
      "loss": 0.824,
      "step": 180
    },
    {
      "epoch": 0.07660455486542443,
      "grad_norm": 3.0377955598038473,
      "learning_rate": 7.644628099173555e-06,
      "loss": 0.8055,
      "step": 185
    },
    {
      "epoch": 0.07867494824016563,
      "grad_norm": 2.8402645063559446,
      "learning_rate": 7.851239669421489e-06,
      "loss": 0.8168,
      "step": 190
    },
    {
      "epoch": 0.08074534161490683,
      "grad_norm": 3.0642470118373084,
      "learning_rate": 8.057851239669421e-06,
      "loss": 0.8231,
      "step": 195
    },
    {
      "epoch": 0.08281573498964803,
      "grad_norm": 3.5617224210742764,
      "learning_rate": 8.264462809917356e-06,
      "loss": 0.8134,
      "step": 200
    },
    {
      "epoch": 0.08488612836438923,
      "grad_norm": 2.930705372475165,
      "learning_rate": 8.47107438016529e-06,
      "loss": 0.8081,
      "step": 205
    },
    {
      "epoch": 0.08695652173913043,
      "grad_norm": 2.982860889121232,
      "learning_rate": 8.677685950413224e-06,
      "loss": 0.809,
      "step": 210
    },
    {
      "epoch": 0.08902691511387163,
      "grad_norm": 2.935587168969496,
      "learning_rate": 8.884297520661158e-06,
      "loss": 0.8169,
      "step": 215
    },
    {
      "epoch": 0.09109730848861283,
      "grad_norm": 2.9901723241847153,
      "learning_rate": 9.090909090909091e-06,
      "loss": 0.8153,
      "step": 220
    },
    {
      "epoch": 0.09316770186335403,
      "grad_norm": 2.921963512037345,
      "learning_rate": 9.297520661157025e-06,
      "loss": 0.7979,
      "step": 225
    },
    {
      "epoch": 0.09523809523809523,
      "grad_norm": 3.080906597029773,
      "learning_rate": 9.50413223140496e-06,
      "loss": 0.7883,
      "step": 230
    },
    {
      "epoch": 0.09730848861283643,
      "grad_norm": 2.9796857583597913,
      "learning_rate": 9.710743801652894e-06,
      "loss": 0.7798,
      "step": 235
    },
    {
      "epoch": 0.09937888198757763,
      "grad_norm": 2.7938646070615105,
      "learning_rate": 9.917355371900828e-06,
      "loss": 0.8149,
      "step": 240
    },
    {
      "epoch": 0.10144927536231885,
      "grad_norm": 3.0770015879691766,
      "learning_rate": 9.999952971391373e-06,
      "loss": 0.8014,
      "step": 245
    },
    {
      "epoch": 0.10351966873706005,
      "grad_norm": 2.964452998837368,
      "learning_rate": 9.999665577542406e-06,
      "loss": 0.8048,
      "step": 250
    },
    {
      "epoch": 0.10559006211180125,
      "grad_norm": 3.022786395730761,
      "learning_rate": 9.999116931848504e-06,
      "loss": 0.7931,
      "step": 255
    },
    {
      "epoch": 0.10766045548654245,
      "grad_norm": 2.7899060573318497,
      "learning_rate": 9.99830706297852e-06,
      "loss": 0.7975,
      "step": 260
    },
    {
      "epoch": 0.10973084886128365,
      "grad_norm": 2.9941483284208803,
      "learning_rate": 9.997236013251234e-06,
      "loss": 0.783,
      "step": 265
    },
    {
      "epoch": 0.11180124223602485,
      "grad_norm": 2.8639530933271597,
      "learning_rate": 9.995903838633133e-06,
      "loss": 0.7767,
      "step": 270
    },
    {
      "epoch": 0.11387163561076605,
      "grad_norm": 2.893987524405616,
      "learning_rate": 9.994310608735492e-06,
      "loss": 0.7975,
      "step": 275
    },
    {
      "epoch": 0.11594202898550725,
      "grad_norm": 3.002021189598603,
      "learning_rate": 9.99245640681072e-06,
      "loss": 0.7962,
      "step": 280
    },
    {
      "epoch": 0.11801242236024845,
      "grad_norm": 2.884029199363375,
      "learning_rate": 9.99034132974804e-06,
      "loss": 0.7815,
      "step": 285
    },
    {
      "epoch": 0.12008281573498965,
      "grad_norm": 2.9227341204938746,
      "learning_rate": 9.987965488068398e-06,
      "loss": 0.7903,
      "step": 290
    },
    {
      "epoch": 0.12215320910973085,
      "grad_norm": 3.09433044004106,
      "learning_rate": 9.985329005918702e-06,
      "loss": 0.7636,
      "step": 295
    },
    {
      "epoch": 0.12422360248447205,
      "grad_norm": 3.14926423201632,
      "learning_rate": 9.982432021065334e-06,
      "loss": 0.7762,
      "step": 300
    },
    {
      "epoch": 0.12629399585921325,
      "grad_norm": 3.000146608974552,
      "learning_rate": 9.979274684886943e-06,
      "loss": 0.7591,
      "step": 305
    },
    {
      "epoch": 0.12836438923395446,
      "grad_norm": 2.9294337201892193,
      "learning_rate": 9.975857162366547e-06,
      "loss": 0.7592,
      "step": 310
    },
    {
      "epoch": 0.13043478260869565,
      "grad_norm": 2.871508509341422,
      "learning_rate": 9.972179632082899e-06,
      "loss": 0.763,
      "step": 315
    },
    {
      "epoch": 0.13250517598343686,
      "grad_norm": 2.9825261500051923,
      "learning_rate": 9.968242286201171e-06,
      "loss": 0.7699,
      "step": 320
    },
    {
      "epoch": 0.13457556935817805,
      "grad_norm": 2.808336705810353,
      "learning_rate": 9.964045330462896e-06,
      "loss": 0.7743,
      "step": 325
    },
    {
      "epoch": 0.13664596273291926,
      "grad_norm": 3.015312838928351,
      "learning_rate": 9.959588984175228e-06,
      "loss": 0.7718,
      "step": 330
    },
    {
      "epoch": 0.13871635610766045,
      "grad_norm": 2.870579878017002,
      "learning_rate": 9.954873480199483e-06,
      "loss": 0.7851,
      "step": 335
    },
    {
      "epoch": 0.14078674948240166,
      "grad_norm": 2.7554815247431983,
      "learning_rate": 9.94989906493896e-06,
      "loss": 0.7545,
      "step": 340
    },
    {
      "epoch": 0.14285714285714285,
      "grad_norm": 2.75690418932418,
      "learning_rate": 9.94466599832608e-06,
      "loss": 0.7563,
      "step": 345
    },
    {
      "epoch": 0.14492753623188406,
      "grad_norm": 2.852997059770447,
      "learning_rate": 9.939174553808793e-06,
      "loss": 0.7651,
      "step": 350
    },
    {
      "epoch": 0.14699792960662525,
      "grad_norm": 2.770789020857082,
      "learning_rate": 9.933425018336292e-06,
      "loss": 0.7594,
      "step": 355
    },
    {
      "epoch": 0.14906832298136646,
      "grad_norm": 3.026030650317298,
      "learning_rate": 9.927417692344025e-06,
      "loss": 0.7355,
      "step": 360
    },
    {
      "epoch": 0.15113871635610765,
      "grad_norm": 2.9786011782652824,
      "learning_rate": 9.921152889737985e-06,
      "loss": 0.7343,
      "step": 365
    },
    {
      "epoch": 0.15320910973084886,
      "grad_norm": 2.8419772278707396,
      "learning_rate": 9.914630937878315e-06,
      "loss": 0.7505,
      "step": 370
    },
    {
      "epoch": 0.15527950310559005,
      "grad_norm": 2.762145018804381,
      "learning_rate": 9.907852177562201e-06,
      "loss": 0.7452,
      "step": 375
    },
    {
      "epoch": 0.15734989648033126,
      "grad_norm": 2.6825337667863023,
      "learning_rate": 9.900816963006063e-06,
      "loss": 0.7423,
      "step": 380
    },
    {
      "epoch": 0.15942028985507245,
      "grad_norm": 2.7080088144526226,
      "learning_rate": 9.893525661827043e-06,
      "loss": 0.7318,
      "step": 385
    },
    {
      "epoch": 0.16149068322981366,
      "grad_norm": 3.030977827239322,
      "learning_rate": 9.885978655023805e-06,
      "loss": 0.7312,
      "step": 390
    },
    {
      "epoch": 0.16356107660455488,
      "grad_norm": 3.040313981795896,
      "learning_rate": 9.87817633695661e-06,
      "loss": 0.7331,
      "step": 395
    },
    {
      "epoch": 0.16563146997929606,
      "grad_norm": 2.8531908645344646,
      "learning_rate": 9.87011911532673e-06,
      "loss": 0.7303,
      "step": 400
    },
    {
      "epoch": 0.16770186335403728,
      "grad_norm": 2.9309768565382885,
      "learning_rate": 9.861807411155126e-06,
      "loss": 0.7289,
      "step": 405
    },
    {
      "epoch": 0.16977225672877846,
      "grad_norm": 2.8699572447082082,
      "learning_rate": 9.853241658760457e-06,
      "loss": 0.7167,
      "step": 410
    },
    {
      "epoch": 0.17184265010351968,
      "grad_norm": 2.7894953022566065,
      "learning_rate": 9.844422305736383e-06,
      "loss": 0.7267,
      "step": 415
    },
    {
      "epoch": 0.17391304347826086,
      "grad_norm": 2.836369561123053,
      "learning_rate": 9.835349812928178e-06,
      "loss": 0.7182,
      "step": 420
    },
    {
      "epoch": 0.17598343685300208,
      "grad_norm": 2.8544078673515867,
      "learning_rate": 9.826024654408645e-06,
      "loss": 0.737,
      "step": 425
    },
    {
      "epoch": 0.17805383022774326,
      "grad_norm": 2.711311382577373,
      "learning_rate": 9.816447317453353e-06,
      "loss": 0.7037,
      "step": 430
    },
    {
      "epoch": 0.18012422360248448,
      "grad_norm": 2.810433161088231,
      "learning_rate": 9.80661830251516e-06,
      "loss": 0.6939,
      "step": 435
    },
    {
      "epoch": 0.18219461697722567,
      "grad_norm": 3.289800860905077,
      "learning_rate": 9.796538123198077e-06,
      "loss": 0.7174,
      "step": 440
    },
    {
      "epoch": 0.18426501035196688,
      "grad_norm": 2.587129399365275,
      "learning_rate": 9.786207306230422e-06,
      "loss": 0.7105,
      "step": 445
    },
    {
      "epoch": 0.18633540372670807,
      "grad_norm": 2.907975372472839,
      "learning_rate": 9.775626391437303e-06,
      "loss": 0.7078,
      "step": 450
    },
    {
      "epoch": 0.18840579710144928,
      "grad_norm": 2.725655072081525,
      "learning_rate": 9.764795931712396e-06,
      "loss": 0.7172,
      "step": 455
    },
    {
      "epoch": 0.19047619047619047,
      "grad_norm": 2.685932024684609,
      "learning_rate": 9.753716492989076e-06,
      "loss": 0.6874,
      "step": 460
    },
    {
      "epoch": 0.19254658385093168,
      "grad_norm": 2.933612272185732,
      "learning_rate": 9.742388654210822e-06,
      "loss": 0.6839,
      "step": 465
    },
    {
      "epoch": 0.19461697722567287,
      "grad_norm": 2.8194507459788625,
      "learning_rate": 9.730813007300984e-06,
      "loss": 0.6978,
      "step": 470
    },
    {
      "epoch": 0.19668737060041408,
      "grad_norm": 2.987546380858413,
      "learning_rate": 9.718990157131841e-06,
      "loss": 0.7011,
      "step": 475
    },
    {
      "epoch": 0.19875776397515527,
      "grad_norm": 2.7372414926444817,
      "learning_rate": 9.706920721492995e-06,
      "loss": 0.6903,
      "step": 480
    },
    {
      "epoch": 0.20082815734989648,
      "grad_norm": 2.846181638635384,
      "learning_rate": 9.694605331059094e-06,
      "loss": 0.7003,
      "step": 485
    },
    {
      "epoch": 0.2028985507246377,
      "grad_norm": 2.9193024154798195,
      "learning_rate": 9.682044629356874e-06,
      "loss": 0.6906,
      "step": 490
    },
    {
      "epoch": 0.20496894409937888,
      "grad_norm": 2.964752286342872,
      "learning_rate": 9.669239272731538e-06,
      "loss": 0.6787,
      "step": 495
    },
    {
      "epoch": 0.2070393374741201,
      "grad_norm": 2.967854934387599,
      "learning_rate": 9.656189930312443e-06,
      "loss": 0.678,
      "step": 500
    },
    {
      "epoch": 0.20910973084886128,
      "grad_norm": 3.0935315145303206,
      "learning_rate": 9.642897283978157e-06,
      "loss": 0.6699,
      "step": 505
    },
    {
      "epoch": 0.2111801242236025,
      "grad_norm": 2.8949142000184396,
      "learning_rate": 9.629362028320808e-06,
      "loss": 0.6883,
      "step": 510
    },
    {
      "epoch": 0.21325051759834368,
      "grad_norm": 2.7523262750257524,
      "learning_rate": 9.615584870609809e-06,
      "loss": 0.6906,
      "step": 515
    },
    {
      "epoch": 0.2153209109730849,
      "grad_norm": 2.8266613600542474,
      "learning_rate": 9.601566530754882e-06,
      "loss": 0.6714,
      "step": 520
    },
    {
      "epoch": 0.21739130434782608,
      "grad_norm": 2.7407909229675607,
      "learning_rate": 9.587307741268452e-06,
      "loss": 0.6741,
      "step": 525
    },
    {
      "epoch": 0.2194616977225673,
      "grad_norm": 2.85692797255821,
      "learning_rate": 9.572809247227366e-06,
      "loss": 0.6869,
      "step": 530
    },
    {
      "epoch": 0.22153209109730848,
      "grad_norm": 3.0795533589960327,
      "learning_rate": 9.558071806233955e-06,
      "loss": 0.6523,
      "step": 535
    },
    {
      "epoch": 0.2236024844720497,
      "grad_norm": 3.0611844069540703,
      "learning_rate": 9.54309618837646e-06,
      "loss": 0.68,
      "step": 540
    },
    {
      "epoch": 0.22567287784679088,
      "grad_norm": 2.835988286170059,
      "learning_rate": 9.52788317618878e-06,
      "loss": 0.6853,
      "step": 545
    },
    {
      "epoch": 0.2277432712215321,
      "grad_norm": 2.7835609496718714,
      "learning_rate": 9.512433564609578e-06,
      "loss": 0.6555,
      "step": 550
    },
    {
      "epoch": 0.22981366459627328,
      "grad_norm": 2.808940901509538,
      "learning_rate": 9.496748160940762e-06,
      "loss": 0.668,
      "step": 555
    },
    {
      "epoch": 0.2318840579710145,
      "grad_norm": 2.714688284772169,
      "learning_rate": 9.480827784805278e-06,
      "loss": 0.6623,
      "step": 560
    },
    {
      "epoch": 0.23395445134575568,
      "grad_norm": 2.8864412901700343,
      "learning_rate": 9.464673268104299e-06,
      "loss": 0.6474,
      "step": 565
    },
    {
      "epoch": 0.2360248447204969,
      "grad_norm": 2.701727867973264,
      "learning_rate": 9.448285454973739e-06,
      "loss": 0.6619,
      "step": 570
    },
    {
      "epoch": 0.23809523809523808,
      "grad_norm": 2.9859632562398715,
      "learning_rate": 9.431665201740154e-06,
      "loss": 0.6673,
      "step": 575
    },
    {
      "epoch": 0.2401656314699793,
      "grad_norm": 2.702445930286614,
      "learning_rate": 9.414813376876003e-06,
      "loss": 0.64,
      "step": 580
    },
    {
      "epoch": 0.2422360248447205,
      "grad_norm": 2.84928084374811,
      "learning_rate": 9.397730860954242e-06,
      "loss": 0.6611,
      "step": 585
    },
    {
      "epoch": 0.2443064182194617,
      "grad_norm": 2.7099926341587484,
      "learning_rate": 9.38041854660234e-06,
      "loss": 0.6648,
      "step": 590
    },
    {
      "epoch": 0.2463768115942029,
      "grad_norm": 2.60007554596025,
      "learning_rate": 9.362877338455611e-06,
      "loss": 0.6183,
      "step": 595
    },
    {
      "epoch": 0.2484472049689441,
      "grad_norm": 2.846852340669187,
      "learning_rate": 9.345108153109963e-06,
      "loss": 0.6322,
      "step": 600
    },
    {
      "epoch": 0.2505175983436853,
      "grad_norm": 2.6294782931340404,
      "learning_rate": 9.327111919073988e-06,
      "loss": 0.644,
      "step": 605
    },
    {
      "epoch": 0.2525879917184265,
      "grad_norm": 3.0332486107737653,
      "learning_rate": 9.308889576720453e-06,
      "loss": 0.6454,
      "step": 610
    },
    {
      "epoch": 0.2546583850931677,
      "grad_norm": 2.7542953723768098,
      "learning_rate": 9.290442078237154e-06,
      "loss": 0.6122,
      "step": 615
    },
    {
      "epoch": 0.2567287784679089,
      "grad_norm": 2.7123354633572694,
      "learning_rate": 9.271770387577168e-06,
      "loss": 0.6313,
      "step": 620
    },
    {
      "epoch": 0.2587991718426501,
      "grad_norm": 2.8410949470006424,
      "learning_rate": 9.252875480408479e-06,
      "loss": 0.6434,
      "step": 625
    },
    {
      "epoch": 0.2608695652173913,
      "grad_norm": 2.8320923536012095,
      "learning_rate": 9.233758344062996e-06,
      "loss": 0.6339,
      "step": 630
    },
    {
      "epoch": 0.2629399585921325,
      "grad_norm": 3.059074128048717,
      "learning_rate": 9.21441997748496e-06,
      "loss": 0.6227,
      "step": 635
    },
    {
      "epoch": 0.2650103519668737,
      "grad_norm": 2.7546107516694343,
      "learning_rate": 9.194861391178749e-06,
      "loss": 0.6313,
      "step": 640
    },
    {
      "epoch": 0.2670807453416149,
      "grad_norm": 2.785158104538987,
      "learning_rate": 9.175083607156067e-06,
      "loss": 0.6427,
      "step": 645
    },
    {
      "epoch": 0.2691511387163561,
      "grad_norm": 2.519237303716458,
      "learning_rate": 9.155087658882555e-06,
      "loss": 0.634,
      "step": 650
    },
    {
      "epoch": 0.2712215320910973,
      "grad_norm": 2.62906922660549,
      "learning_rate": 9.134874591223773e-06,
      "loss": 0.6202,
      "step": 655
    },
    {
      "epoch": 0.2732919254658385,
      "grad_norm": 8.776521328405046,
      "learning_rate": 9.114445460390605e-06,
      "loss": 0.6359,
      "step": 660
    },
    {
      "epoch": 0.2753623188405797,
      "grad_norm": 2.67193668011001,
      "learning_rate": 9.093801333884076e-06,
      "loss": 0.6279,
      "step": 665
    },
    {
      "epoch": 0.2774327122153209,
      "grad_norm": 2.7070968852937134,
      "learning_rate": 9.072943290439566e-06,
      "loss": 0.6364,
      "step": 670
    },
    {
      "epoch": 0.2795031055900621,
      "grad_norm": 2.613414279867121,
      "learning_rate": 9.051872419970439e-06,
      "loss": 0.6196,
      "step": 675
    },
    {
      "epoch": 0.2815734989648033,
      "grad_norm": 2.8664991598537877,
      "learning_rate": 9.03058982351109e-06,
      "loss": 0.6066,
      "step": 680
    },
    {
      "epoch": 0.2836438923395445,
      "grad_norm": 2.641252430207258,
      "learning_rate": 9.009096613159426e-06,
      "loss": 0.6093,
      "step": 685
    },
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 2.93725140870653,
      "learning_rate": 8.987393912018732e-06,
      "loss": 0.6183,
      "step": 690
    },
    {
      "epoch": 0.28778467908902694,
      "grad_norm": 2.7035731768373314,
      "learning_rate": 8.965482854139003e-06,
      "loss": 0.5946,
      "step": 695
    },
    {
      "epoch": 0.2898550724637681,
      "grad_norm": 2.7946358641447238,
      "learning_rate": 8.943364584457675e-06,
      "loss": 0.596,
      "step": 700
    },
    {
      "epoch": 0.2919254658385093,
      "grad_norm": 2.79245272627438,
      "learning_rate": 8.921040258739804e-06,
      "loss": 0.6038,
      "step": 705
    },
    {
      "epoch": 0.2939958592132505,
      "grad_norm": 2.807065308334584,
      "learning_rate": 8.898511043517668e-06,
      "loss": 0.6047,
      "step": 710
    },
    {
      "epoch": 0.29606625258799174,
      "grad_norm": 2.9178175401076354,
      "learning_rate": 8.875778116029816e-06,
      "loss": 0.5961,
      "step": 715
    },
    {
      "epoch": 0.2981366459627329,
      "grad_norm": 2.795866621933458,
      "learning_rate": 8.85284266415955e-06,
      "loss": 0.5999,
      "step": 720
    },
    {
      "epoch": 0.3002070393374741,
      "grad_norm": 2.661912145911008,
      "learning_rate": 8.829705886372845e-06,
      "loss": 0.6016,
      "step": 725
    },
    {
      "epoch": 0.3022774327122153,
      "grad_norm": 2.55928416799956,
      "learning_rate": 8.806368991655747e-06,
      "loss": 0.603,
      "step": 730
    },
    {
      "epoch": 0.30434782608695654,
      "grad_norm": 2.742096866025067,
      "learning_rate": 8.782833199451177e-06,
      "loss": 0.593,
      "step": 735
    },
    {
      "epoch": 0.3064182194616977,
      "grad_norm": 2.5488046049420694,
      "learning_rate": 8.759099739595215e-06,
      "loss": 0.5985,
      "step": 740
    },
    {
      "epoch": 0.3084886128364389,
      "grad_norm": 2.6637124391121567,
      "learning_rate": 8.735169852252848e-06,
      "loss": 0.5912,
      "step": 745
    },
    {
      "epoch": 0.3105590062111801,
      "grad_norm": 2.8576203296823017,
      "learning_rate": 8.71104478785315e-06,
      "loss": 0.5922,
      "step": 750
    },
    {
      "epoch": 0.31262939958592134,
      "grad_norm": 2.6109792985439486,
      "learning_rate": 8.686725807023955e-06,
      "loss": 0.5821,
      "step": 755
    },
    {
      "epoch": 0.3146997929606625,
      "grad_norm": 2.5680758470796254,
      "learning_rate": 8.662214180525982e-06,
      "loss": 0.5901,
      "step": 760
    },
    {
      "epoch": 0.3167701863354037,
      "grad_norm": 2.773059019392135,
      "learning_rate": 8.637511189186425e-06,
      "loss": 0.584,
      "step": 765
    },
    {
      "epoch": 0.3188405797101449,
      "grad_norm": 2.723807950487225,
      "learning_rate": 8.612618123832033e-06,
      "loss": 0.5637,
      "step": 770
    },
    {
      "epoch": 0.32091097308488614,
      "grad_norm": 2.719604774085743,
      "learning_rate": 8.587536285221656e-06,
      "loss": 0.593,
      "step": 775
    },
    {
      "epoch": 0.32298136645962733,
      "grad_norm": 2.6940868411607855,
      "learning_rate": 8.562266983978278e-06,
      "loss": 0.5808,
      "step": 780
    },
    {
      "epoch": 0.3250517598343685,
      "grad_norm": 2.71247751226207,
      "learning_rate": 8.536811540520529e-06,
      "loss": 0.5719,
      "step": 785
    },
    {
      "epoch": 0.32712215320910976,
      "grad_norm": 2.8315337603387327,
      "learning_rate": 8.511171284993686e-06,
      "loss": 0.5792,
      "step": 790
    },
    {
      "epoch": 0.32919254658385094,
      "grad_norm": 2.7344602885064035,
      "learning_rate": 8.485347557200177e-06,
      "loss": 0.5745,
      "step": 795
    },
    {
      "epoch": 0.33126293995859213,
      "grad_norm": 2.5600120348098225,
      "learning_rate": 8.459341706529557e-06,
      "loss": 0.5692,
      "step": 800
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 2.624179495038465,
      "learning_rate": 8.43315509188801e-06,
      "loss": 0.5931,
      "step": 805
    },
    {
      "epoch": 0.33540372670807456,
      "grad_norm": 2.680547730395255,
      "learning_rate": 8.406789081627337e-06,
      "loss": 0.5558,
      "step": 810
    },
    {
      "epoch": 0.33747412008281574,
      "grad_norm": 2.5218096483801746,
      "learning_rate": 8.380245053473452e-06,
      "loss": 0.5458,
      "step": 815
    },
    {
      "epoch": 0.33954451345755693,
      "grad_norm": 2.6618542869475323,
      "learning_rate": 8.353524394454388e-06,
      "loss": 0.5714,
      "step": 820
    },
    {
      "epoch": 0.3416149068322981,
      "grad_norm": 2.682552924118767,
      "learning_rate": 8.326628500827826e-06,
      "loss": 0.574,
      "step": 825
    },
    {
      "epoch": 0.34368530020703936,
      "grad_norm": 2.7546015850988432,
      "learning_rate": 8.299558778008137e-06,
      "loss": 0.5614,
      "step": 830
    },
    {
      "epoch": 0.34575569358178054,
      "grad_norm": 2.693315288346116,
      "learning_rate": 8.272316640492932e-06,
      "loss": 0.5606,
      "step": 835
    },
    {
      "epoch": 0.34782608695652173,
      "grad_norm": 2.6093176240781846,
      "learning_rate": 8.244903511789158e-06,
      "loss": 0.5674,
      "step": 840
    },
    {
      "epoch": 0.3498964803312629,
      "grad_norm": 3.0660262244457415,
      "learning_rate": 8.217320824338713e-06,
      "loss": 0.5678,
      "step": 845
    },
    {
      "epoch": 0.35196687370600416,
      "grad_norm": 2.601542138482762,
      "learning_rate": 8.189570019443597e-06,
      "loss": 0.5545,
      "step": 850
    },
    {
      "epoch": 0.35403726708074534,
      "grad_norm": 2.680496513220756,
      "learning_rate": 8.161652547190593e-06,
      "loss": 0.5505,
      "step": 855
    },
    {
      "epoch": 0.35610766045548653,
      "grad_norm": 2.71559115559276,
      "learning_rate": 8.133569866375497e-06,
      "loss": 0.5507,
      "step": 860
    },
    {
      "epoch": 0.3581780538302277,
      "grad_norm": 2.5491866007518174,
      "learning_rate": 8.105323444426891e-06,
      "loss": 0.5454,
      "step": 865
    },
    {
      "epoch": 0.36024844720496896,
      "grad_norm": 2.5929968996001964,
      "learning_rate": 8.076914757329467e-06,
      "loss": 0.5744,
      "step": 870
    },
    {
      "epoch": 0.36231884057971014,
      "grad_norm": 2.8545194694179257,
      "learning_rate": 8.048345289546895e-06,
      "loss": 0.5452,
      "step": 875
    },
    {
      "epoch": 0.36438923395445133,
      "grad_norm": 2.6374637614435263,
      "learning_rate": 8.01961653394426e-06,
      "loss": 0.5527,
      "step": 880
    },
    {
      "epoch": 0.36645962732919257,
      "grad_norm": 2.805538262703338,
      "learning_rate": 7.990729991710046e-06,
      "loss": 0.5275,
      "step": 885
    },
    {
      "epoch": 0.36853002070393376,
      "grad_norm": 2.600470832967854,
      "learning_rate": 7.9616871722777e-06,
      "loss": 0.5304,
      "step": 890
    },
    {
      "epoch": 0.37060041407867494,
      "grad_norm": 2.5695628696449133,
      "learning_rate": 7.932489593246764e-06,
      "loss": 0.5332,
      "step": 895
    },
    {
      "epoch": 0.37267080745341613,
      "grad_norm": 2.707737452763526,
      "learning_rate": 7.903138780303556e-06,
      "loss": 0.5663,
      "step": 900
    },
    {
      "epoch": 0.3747412008281574,
      "grad_norm": 2.6330755439053575,
      "learning_rate": 7.873636267141463e-06,
      "loss": 0.5398,
      "step": 905
    },
    {
      "epoch": 0.37681159420289856,
      "grad_norm": 2.629184440328501,
      "learning_rate": 7.843983595380793e-06,
      "loss": 0.5407,
      "step": 910
    },
    {
      "epoch": 0.37888198757763975,
      "grad_norm": 2.523570743421866,
      "learning_rate": 7.814182314488225e-06,
      "loss": 0.5339,
      "step": 915
    },
    {
      "epoch": 0.38095238095238093,
      "grad_norm": 2.5767883541234293,
      "learning_rate": 7.784233981695835e-06,
      "loss": 0.5333,
      "step": 920
    },
    {
      "epoch": 0.3830227743271222,
      "grad_norm": 2.4631684405639267,
      "learning_rate": 7.754140161919732e-06,
      "loss": 0.5253,
      "step": 925
    },
    {
      "epoch": 0.38509316770186336,
      "grad_norm": 2.501412257400789,
      "learning_rate": 7.72390242767828e-06,
      "loss": 0.5247,
      "step": 930
    },
    {
      "epoch": 0.38716356107660455,
      "grad_norm": 2.7412684154036233,
      "learning_rate": 7.693522359009931e-06,
      "loss": 0.5304,
      "step": 935
    },
    {
      "epoch": 0.38923395445134573,
      "grad_norm": 2.519374343070241,
      "learning_rate": 7.663001543390657e-06,
      "loss": 0.5178,
      "step": 940
    },
    {
      "epoch": 0.391304347826087,
      "grad_norm": 2.659503227676441,
      "learning_rate": 7.63234157565101e-06,
      "loss": 0.504,
      "step": 945
    },
    {
      "epoch": 0.39337474120082816,
      "grad_norm": 2.6028733871149337,
      "learning_rate": 7.601544057892769e-06,
      "loss": 0.5281,
      "step": 950
    },
    {
      "epoch": 0.39544513457556935,
      "grad_norm": 2.6684608680601425,
      "learning_rate": 7.570610599405242e-06,
      "loss": 0.5289,
      "step": 955
    },
    {
      "epoch": 0.39751552795031053,
      "grad_norm": 2.580954402164553,
      "learning_rate": 7.539542816581157e-06,
      "loss": 0.5243,
      "step": 960
    },
    {
      "epoch": 0.3995859213250518,
      "grad_norm": 2.656105021997607,
      "learning_rate": 7.508342332832213e-06,
      "loss": 0.5337,
      "step": 965
    },
    {
      "epoch": 0.40165631469979296,
      "grad_norm": 2.5689808051046885,
      "learning_rate": 7.477010778504241e-06,
      "loss": 0.5244,
      "step": 970
    },
    {
      "epoch": 0.40372670807453415,
      "grad_norm": 2.660859162384531,
      "learning_rate": 7.445549790792021e-06,
      "loss": 0.5249,
      "step": 975
    },
    {
      "epoch": 0.4057971014492754,
      "grad_norm": 2.5110342843182645,
      "learning_rate": 7.413961013653725e-06,
      "loss": 0.529,
      "step": 980
    },
    {
      "epoch": 0.4078674948240166,
      "grad_norm": 2.6278595424042757,
      "learning_rate": 7.3822460977250145e-06,
      "loss": 0.508,
      "step": 985
    },
    {
      "epoch": 0.40993788819875776,
      "grad_norm": 2.570820560507206,
      "learning_rate": 7.350406700232794e-06,
      "loss": 0.5135,
      "step": 990
    },
    {
      "epoch": 0.41200828157349895,
      "grad_norm": 2.4853904047248876,
      "learning_rate": 7.318444484908606e-06,
      "loss": 0.515,
      "step": 995
    },
    {
      "epoch": 0.4140786749482402,
      "grad_norm": 2.544205959165577,
      "learning_rate": 7.286361121901706e-06,
      "loss": 0.5141,
      "step": 1000
    },
    {
      "epoch": 0.4161490683229814,
      "grad_norm": 2.750413997685881,
      "learning_rate": 7.254158287691775e-06,
      "loss": 0.5058,
      "step": 1005
    },
    {
      "epoch": 0.41821946169772256,
      "grad_norm": 2.4471491822679905,
      "learning_rate": 7.221837665001335e-06,
      "loss": 0.508,
      "step": 1010
    },
    {
      "epoch": 0.42028985507246375,
      "grad_norm": 2.569088939282542,
      "learning_rate": 7.189400942707804e-06,
      "loss": 0.5008,
      "step": 1015
    },
    {
      "epoch": 0.422360248447205,
      "grad_norm": 2.58098669010246,
      "learning_rate": 7.1568498157552576e-06,
      "loss": 0.5122,
      "step": 1020
    },
    {
      "epoch": 0.4244306418219462,
      "grad_norm": 2.5507810903416748,
      "learning_rate": 7.124185985065856e-06,
      "loss": 0.4997,
      "step": 1025
    },
    {
      "epoch": 0.42650103519668736,
      "grad_norm": 2.640859603851859,
      "learning_rate": 7.091411157450965e-06,
      "loss": 0.5063,
      "step": 1030
    },
    {
      "epoch": 0.42857142857142855,
      "grad_norm": 2.5916583011234753,
      "learning_rate": 7.0585270455219654e-06,
      "loss": 0.5004,
      "step": 1035
    },
    {
      "epoch": 0.4306418219461698,
      "grad_norm": 2.5837013370937907,
      "learning_rate": 7.025535367600771e-06,
      "loss": 0.5131,
      "step": 1040
    },
    {
      "epoch": 0.432712215320911,
      "grad_norm": 2.4630480113570243,
      "learning_rate": 6.992437847630031e-06,
      "loss": 0.4833,
      "step": 1045
    },
    {
      "epoch": 0.43478260869565216,
      "grad_norm": 2.6336472266054725,
      "learning_rate": 6.95923621508305e-06,
      "loss": 0.4843,
      "step": 1050
    },
    {
      "epoch": 0.43685300207039335,
      "grad_norm": 2.5354926775279996,
      "learning_rate": 6.92593220487342e-06,
      "loss": 0.4896,
      "step": 1055
    },
    {
      "epoch": 0.4389233954451346,
      "grad_norm": 2.571259522555512,
      "learning_rate": 6.892527557264358e-06,
      "loss": 0.5064,
      "step": 1060
    },
    {
      "epoch": 0.4409937888198758,
      "grad_norm": 2.645183661888645,
      "learning_rate": 6.859024017777779e-06,
      "loss": 0.5047,
      "step": 1065
    },
    {
      "epoch": 0.44306418219461696,
      "grad_norm": 2.469286466558817,
      "learning_rate": 6.825423337103074e-06,
      "loss": 0.4955,
      "step": 1070
    },
    {
      "epoch": 0.4451345755693582,
      "grad_norm": 2.5492415232980457,
      "learning_rate": 6.791727271005642e-06,
      "loss": 0.4947,
      "step": 1075
    },
    {
      "epoch": 0.4472049689440994,
      "grad_norm": 2.6201976455165155,
      "learning_rate": 6.757937580235138e-06,
      "loss": 0.4938,
      "step": 1080
    },
    {
      "epoch": 0.4492753623188406,
      "grad_norm": 2.6040132677270167,
      "learning_rate": 6.724056030433464e-06,
      "loss": 0.495,
      "step": 1085
    },
    {
      "epoch": 0.45134575569358176,
      "grad_norm": 2.4219926979901576,
      "learning_rate": 6.690084392042514e-06,
      "loss": 0.4821,
      "step": 1090
    },
    {
      "epoch": 0.453416149068323,
      "grad_norm": 2.629031701175065,
      "learning_rate": 6.656024440211662e-06,
      "loss": 0.4842,
      "step": 1095
    },
    {
      "epoch": 0.4554865424430642,
      "grad_norm": 2.4715473741738427,
      "learning_rate": 6.621877954704996e-06,
      "loss": 0.4877,
      "step": 1100
    },
    {
      "epoch": 0.4575569358178054,
      "grad_norm": 2.6497148727345903,
      "learning_rate": 6.5876467198083235e-06,
      "loss": 0.4904,
      "step": 1105
    },
    {
      "epoch": 0.45962732919254656,
      "grad_norm": 2.529203852840936,
      "learning_rate": 6.553332524235937e-06,
      "loss": 0.4707,
      "step": 1110
    },
    {
      "epoch": 0.4616977225672878,
      "grad_norm": 2.565462399734998,
      "learning_rate": 6.518937161037144e-06,
      "loss": 0.4908,
      "step": 1115
    },
    {
      "epoch": 0.463768115942029,
      "grad_norm": 2.5253689137313704,
      "learning_rate": 6.484462427502572e-06,
      "loss": 0.482,
      "step": 1120
    },
    {
      "epoch": 0.4658385093167702,
      "grad_norm": 2.7298927082155497,
      "learning_rate": 6.44991012507026e-06,
      "loss": 0.4635,
      "step": 1125
    },
    {
      "epoch": 0.46790890269151136,
      "grad_norm": 2.4968057887779542,
      "learning_rate": 6.415282059231518e-06,
      "loss": 0.4734,
      "step": 1130
    },
    {
      "epoch": 0.4699792960662526,
      "grad_norm": 2.558749462242469,
      "learning_rate": 6.380580039436586e-06,
      "loss": 0.4724,
      "step": 1135
    },
    {
      "epoch": 0.4720496894409938,
      "grad_norm": 2.6054561174991475,
      "learning_rate": 6.345805879000087e-06,
      "loss": 0.4804,
      "step": 1140
    },
    {
      "epoch": 0.474120082815735,
      "grad_norm": 2.4181929103768782,
      "learning_rate": 6.31096139500627e-06,
      "loss": 0.4653,
      "step": 1145
    },
    {
      "epoch": 0.47619047619047616,
      "grad_norm": 2.6283847623760046,
      "learning_rate": 6.2760484082140604e-06,
      "loss": 0.4777,
      "step": 1150
    },
    {
      "epoch": 0.4782608695652174,
      "grad_norm": 2.4346152969867987,
      "learning_rate": 6.24106874296192e-06,
      "loss": 0.4504,
      "step": 1155
    },
    {
      "epoch": 0.4803312629399586,
      "grad_norm": 2.526644880284406,
      "learning_rate": 6.20602422707252e-06,
      "loss": 0.4736,
      "step": 1160
    },
    {
      "epoch": 0.4824016563146998,
      "grad_norm": 2.53799587618545,
      "learning_rate": 6.1709166917572264e-06,
      "loss": 0.4795,
      "step": 1165
    },
    {
      "epoch": 0.484472049689441,
      "grad_norm": 2.3638682374352173,
      "learning_rate": 6.135747971520412e-06,
      "loss": 0.4698,
      "step": 1170
    },
    {
      "epoch": 0.4865424430641822,
      "grad_norm": 2.460720323189321,
      "learning_rate": 6.100519904063597e-06,
      "loss": 0.465,
      "step": 1175
    },
    {
      "epoch": 0.4886128364389234,
      "grad_norm": 2.5958597363625584,
      "learning_rate": 6.0652343301894345e-06,
      "loss": 0.4804,
      "step": 1180
    },
    {
      "epoch": 0.4906832298136646,
      "grad_norm": 2.66548966977513,
      "learning_rate": 6.029893093705492e-06,
      "loss": 0.4529,
      "step": 1185
    },
    {
      "epoch": 0.4927536231884058,
      "grad_norm": 2.702340679280387,
      "learning_rate": 5.99449804132794e-06,
      "loss": 0.4577,
      "step": 1190
    },
    {
      "epoch": 0.494824016563147,
      "grad_norm": 2.45914300253678,
      "learning_rate": 5.959051022585025e-06,
      "loss": 0.4577,
      "step": 1195
    },
    {
      "epoch": 0.4968944099378882,
      "grad_norm": 2.630887103863114,
      "learning_rate": 5.923553889720447e-06,
      "loss": 0.4776,
      "step": 1200
    },
    {
      "epoch": 0.4989648033126294,
      "grad_norm": 2.5737488710043213,
      "learning_rate": 5.888008497596553e-06,
      "loss": 0.4704,
      "step": 1205
    },
    {
      "epoch": 0.5010351966873706,
      "grad_norm": 2.642459617063942,
      "learning_rate": 5.852416703597431e-06,
      "loss": 0.4723,
      "step": 1210
    },
    {
      "epoch": 0.5031055900621118,
      "grad_norm": 2.4855448696474767,
      "learning_rate": 5.816780367531841e-06,
      "loss": 0.4566,
      "step": 1215
    },
    {
      "epoch": 0.505175983436853,
      "grad_norm": 2.6308877301818083,
      "learning_rate": 5.781101351536041e-06,
      "loss": 0.4642,
      "step": 1220
    },
    {
      "epoch": 0.5072463768115942,
      "grad_norm": 2.3851591122343687,
      "learning_rate": 5.745381519976477e-06,
      "loss": 0.4683,
      "step": 1225
    },
    {
      "epoch": 0.5093167701863354,
      "grad_norm": 2.707845768280749,
      "learning_rate": 5.7096227393523716e-06,
      "loss": 0.4617,
      "step": 1230
    },
    {
      "epoch": 0.5113871635610766,
      "grad_norm": 2.519861478183832,
      "learning_rate": 5.673826878198181e-06,
      "loss": 0.4417,
      "step": 1235
    },
    {
      "epoch": 0.5134575569358178,
      "grad_norm": 2.4813956531185153,
      "learning_rate": 5.637995806985961e-06,
      "loss": 0.4547,
      "step": 1240
    },
    {
      "epoch": 0.515527950310559,
      "grad_norm": 2.5303176264976357,
      "learning_rate": 5.602131398027637e-06,
      "loss": 0.4457,
      "step": 1245
    },
    {
      "epoch": 0.5175983436853002,
      "grad_norm": 2.578815009552858,
      "learning_rate": 5.566235525377155e-06,
      "loss": 0.4676,
      "step": 1250
    },
    {
      "epoch": 0.5196687370600414,
      "grad_norm": 2.436584983948615,
      "learning_rate": 5.530310064732559e-06,
      "loss": 0.4462,
      "step": 1255
    },
    {
      "epoch": 0.5217391304347826,
      "grad_norm": 2.6753470832060193,
      "learning_rate": 5.494356893337985e-06,
      "loss": 0.4458,
      "step": 1260
    },
    {
      "epoch": 0.5238095238095238,
      "grad_norm": 2.46988695395891,
      "learning_rate": 5.4583778898855576e-06,
      "loss": 0.4412,
      "step": 1265
    },
    {
      "epoch": 0.525879917184265,
      "grad_norm": 2.4718187832098875,
      "learning_rate": 5.422374934417228e-06,
      "loss": 0.4459,
      "step": 1270
    },
    {
      "epoch": 0.5279503105590062,
      "grad_norm": 2.488115607824269,
      "learning_rate": 5.386349908226538e-06,
      "loss": 0.4402,
      "step": 1275
    },
    {
      "epoch": 0.5300207039337475,
      "grad_norm": 2.6040039995460855,
      "learning_rate": 5.350304693760301e-06,
      "loss": 0.4527,
      "step": 1280
    },
    {
      "epoch": 0.5320910973084886,
      "grad_norm": 2.564562451871434,
      "learning_rate": 5.314241174520251e-06,
      "loss": 0.4444,
      "step": 1285
    },
    {
      "epoch": 0.5341614906832298,
      "grad_norm": 2.540617138455307,
      "learning_rate": 5.2781612349646175e-06,
      "loss": 0.432,
      "step": 1290
    },
    {
      "epoch": 0.5362318840579711,
      "grad_norm": 2.5068443392119675,
      "learning_rate": 5.242066760409653e-06,
      "loss": 0.4508,
      "step": 1295
    },
    {
      "epoch": 0.5383022774327122,
      "grad_norm": 2.6199384954342255,
      "learning_rate": 5.205959636931121e-06,
      "loss": 0.45,
      "step": 1300
    },
    {
      "epoch": 0.5403726708074534,
      "grad_norm": 2.3921563239272468,
      "learning_rate": 5.169841751265738e-06,
      "loss": 0.4225,
      "step": 1305
    },
    {
      "epoch": 0.5424430641821946,
      "grad_norm": 2.5822600397309516,
      "learning_rate": 5.13371499071259e-06,
      "loss": 0.4359,
      "step": 1310
    },
    {
      "epoch": 0.5445134575569358,
      "grad_norm": 2.4411351270519375,
      "learning_rate": 5.09758124303451e-06,
      "loss": 0.4386,
      "step": 1315
    },
    {
      "epoch": 0.546583850931677,
      "grad_norm": 2.3525010597630076,
      "learning_rate": 5.0614423963594295e-06,
      "loss": 0.4259,
      "step": 1320
    },
    {
      "epoch": 0.5486542443064182,
      "grad_norm": 2.597248272707148,
      "learning_rate": 5.0253003390817264e-06,
      "loss": 0.4368,
      "step": 1325
    },
    {
      "epoch": 0.5507246376811594,
      "grad_norm": 2.7667455072332667,
      "learning_rate": 4.989156959763551e-06,
      "loss": 0.4282,
      "step": 1330
    },
    {
      "epoch": 0.5527950310559007,
      "grad_norm": 2.5638911790116996,
      "learning_rate": 4.953014147036121e-06,
      "loss": 0.4311,
      "step": 1335
    },
    {
      "epoch": 0.5548654244306418,
      "grad_norm": 2.438552616930726,
      "learning_rate": 4.9168737895010615e-06,
      "loss": 0.4331,
      "step": 1340
    },
    {
      "epoch": 0.556935817805383,
      "grad_norm": 2.4621791786870073,
      "learning_rate": 4.880737775631698e-06,
      "loss": 0.4266,
      "step": 1345
    },
    {
      "epoch": 0.5590062111801242,
      "grad_norm": 2.4126817538417376,
      "learning_rate": 4.844607993674382e-06,
      "loss": 0.427,
      "step": 1350
    },
    {
      "epoch": 0.5610766045548654,
      "grad_norm": 2.3616856968719557,
      "learning_rate": 4.808486331549824e-06,
      "loss": 0.4296,
      "step": 1355
    },
    {
      "epoch": 0.5631469979296067,
      "grad_norm": 2.460116527100817,
      "learning_rate": 4.772374676754444e-06,
      "loss": 0.4264,
      "step": 1360
    },
    {
      "epoch": 0.5652173913043478,
      "grad_norm": 2.461765468947944,
      "learning_rate": 4.736274916261741e-06,
      "loss": 0.431,
      "step": 1365
    },
    {
      "epoch": 0.567287784679089,
      "grad_norm": 2.3783642292234384,
      "learning_rate": 4.700188936423683e-06,
      "loss": 0.4234,
      "step": 1370
    },
    {
      "epoch": 0.5693581780538303,
      "grad_norm": 2.355313906431993,
      "learning_rate": 4.664118622872157e-06,
      "loss": 0.437,
      "step": 1375
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 2.475468685556764,
      "learning_rate": 4.628065860420417e-06,
      "loss": 0.4242,
      "step": 1380
    },
    {
      "epoch": 0.5734989648033126,
      "grad_norm": 2.4672709194920945,
      "learning_rate": 4.592032532964611e-06,
      "loss": 0.438,
      "step": 1385
    },
    {
      "epoch": 0.5755693581780539,
      "grad_norm": 2.400293343212344,
      "learning_rate": 4.556020523385326e-06,
      "loss": 0.42,
      "step": 1390
    },
    {
      "epoch": 0.577639751552795,
      "grad_norm": 2.449404289410476,
      "learning_rate": 4.520031713449215e-06,
      "loss": 0.4063,
      "step": 1395
    },
    {
      "epoch": 0.5797101449275363,
      "grad_norm": 2.324896626756316,
      "learning_rate": 4.484067983710653e-06,
      "loss": 0.4187,
      "step": 1400
    },
    {
      "epoch": 0.5817805383022774,
      "grad_norm": 2.5021116418161387,
      "learning_rate": 4.448131213413485e-06,
      "loss": 0.4247,
      "step": 1405
    },
    {
      "epoch": 0.5838509316770186,
      "grad_norm": 2.5924308849874205,
      "learning_rate": 4.4122232803928145e-06,
      "loss": 0.4193,
      "step": 1410
    },
    {
      "epoch": 0.5859213250517599,
      "grad_norm": 2.3873380861073437,
      "learning_rate": 4.376346060976888e-06,
      "loss": 0.4052,
      "step": 1415
    },
    {
      "epoch": 0.587991718426501,
      "grad_norm": 2.463768787838596,
      "learning_rate": 4.340501429889053e-06,
      "loss": 0.4245,
      "step": 1420
    },
    {
      "epoch": 0.5900621118012422,
      "grad_norm": 2.2860972698262496,
      "learning_rate": 4.30469126014978e-06,
      "loss": 0.4153,
      "step": 1425
    },
    {
      "epoch": 0.5921325051759835,
      "grad_norm": 2.4871075210675597,
      "learning_rate": 4.268917422978811e-06,
      "loss": 0.4077,
      "step": 1430
    },
    {
      "epoch": 0.5942028985507246,
      "grad_norm": 2.4904076368286896,
      "learning_rate": 4.233181787697364e-06,
      "loss": 0.4256,
      "step": 1435
    },
    {
      "epoch": 0.5962732919254659,
      "grad_norm": 2.2742029323016193,
      "learning_rate": 4.197486221630467e-06,
      "loss": 0.4219,
      "step": 1440
    },
    {
      "epoch": 0.598343685300207,
      "grad_norm": 2.3406912575365197,
      "learning_rate": 4.161832590009371e-06,
      "loss": 0.4143,
      "step": 1445
    },
    {
      "epoch": 0.6004140786749482,
      "grad_norm": 2.4141594569450597,
      "learning_rate": 4.1262227558740915e-06,
      "loss": 0.4141,
      "step": 1450
    },
    {
      "epoch": 0.6024844720496895,
      "grad_norm": 2.3987457790948126,
      "learning_rate": 4.090658579976058e-06,
      "loss": 0.4007,
      "step": 1455
    },
    {
      "epoch": 0.6045548654244306,
      "grad_norm": 2.369365793426951,
      "learning_rate": 4.055141920680883e-06,
      "loss": 0.4004,
      "step": 1460
    },
    {
      "epoch": 0.6066252587991718,
      "grad_norm": 2.4192836777887874,
      "learning_rate": 4.019674633871246e-06,
      "loss": 0.4019,
      "step": 1465
    },
    {
      "epoch": 0.6086956521739131,
      "grad_norm": 2.422293912807025,
      "learning_rate": 3.984258572849926e-06,
      "loss": 0.411,
      "step": 1470
    },
    {
      "epoch": 0.6107660455486542,
      "grad_norm": 2.3117305367025756,
      "learning_rate": 3.9488955882429605e-06,
      "loss": 0.3967,
      "step": 1475
    },
    {
      "epoch": 0.6128364389233955,
      "grad_norm": 2.578467608284442,
      "learning_rate": 3.9135875279029366e-06,
      "loss": 0.4143,
      "step": 1480
    },
    {
      "epoch": 0.6149068322981367,
      "grad_norm": 2.455925307040979,
      "learning_rate": 3.878336236812432e-06,
      "loss": 0.4036,
      "step": 1485
    },
    {
      "epoch": 0.6169772256728778,
      "grad_norm": 2.4003915989485805,
      "learning_rate": 3.84314355698762e-06,
      "loss": 0.3966,
      "step": 1490
    },
    {
      "epoch": 0.6190476190476191,
      "grad_norm": 2.34779098734652,
      "learning_rate": 3.8080113273820024e-06,
      "loss": 0.3924,
      "step": 1495
    },
    {
      "epoch": 0.6211180124223602,
      "grad_norm": 2.4786298276436365,
      "learning_rate": 3.7729413837903285e-06,
      "loss": 0.3975,
      "step": 1500
    },
    {
      "epoch": 0.6231884057971014,
      "grad_norm": 2.4104863082788452,
      "learning_rate": 3.7379355587526543e-06,
      "loss": 0.398,
      "step": 1505
    },
    {
      "epoch": 0.6252587991718427,
      "grad_norm": 2.4123648264663,
      "learning_rate": 3.702995681458605e-06,
      "loss": 0.3882,
      "step": 1510
    },
    {
      "epoch": 0.6273291925465838,
      "grad_norm": 2.588148178298002,
      "learning_rate": 3.668123577651773e-06,
      "loss": 0.395,
      "step": 1515
    },
    {
      "epoch": 0.629399585921325,
      "grad_norm": 2.353179982037503,
      "learning_rate": 3.633321069534325e-06,
      "loss": 0.391,
      "step": 1520
    },
    {
      "epoch": 0.6314699792960663,
      "grad_norm": 2.3563480016131004,
      "learning_rate": 3.5985899756717872e-06,
      "loss": 0.3923,
      "step": 1525
    },
    {
      "epoch": 0.6335403726708074,
      "grad_norm": 2.352308073230629,
      "learning_rate": 3.563932110898015e-06,
      "loss": 0.401,
      "step": 1530
    },
    {
      "epoch": 0.6356107660455487,
      "grad_norm": 2.5237027304693482,
      "learning_rate": 3.5293492862203595e-06,
      "loss": 0.4018,
      "step": 1535
    },
    {
      "epoch": 0.6376811594202898,
      "grad_norm": 2.44013503338041,
      "learning_rate": 3.494843308725032e-06,
      "loss": 0.395,
      "step": 1540
    },
    {
      "epoch": 0.639751552795031,
      "grad_norm": 2.382891312283338,
      "learning_rate": 3.460415981482691e-06,
      "loss": 0.3926,
      "step": 1545
    },
    {
      "epoch": 0.6418219461697723,
      "grad_norm": 2.28195542777783,
      "learning_rate": 3.426069103454209e-06,
      "loss": 0.3817,
      "step": 1550
    },
    {
      "epoch": 0.6438923395445134,
      "grad_norm": 2.3945835017219985,
      "learning_rate": 3.3918044693966802e-06,
      "loss": 0.3933,
      "step": 1555
    },
    {
      "epoch": 0.6459627329192547,
      "grad_norm": 2.415327395237391,
      "learning_rate": 3.357623869769628e-06,
      "loss": 0.4023,
      "step": 1560
    },
    {
      "epoch": 0.6480331262939959,
      "grad_norm": 2.4479280468812648,
      "learning_rate": 3.3235290906414575e-06,
      "loss": 0.3828,
      "step": 1565
    },
    {
      "epoch": 0.650103519668737,
      "grad_norm": 2.3114942420521554,
      "learning_rate": 3.289521913596117e-06,
      "loss": 0.3926,
      "step": 1570
    },
    {
      "epoch": 0.6521739130434783,
      "grad_norm": 2.372619591954581,
      "learning_rate": 3.255604115640014e-06,
      "loss": 0.3752,
      "step": 1575
    },
    {
      "epoch": 0.6542443064182195,
      "grad_norm": 2.4017704692611095,
      "learning_rate": 3.2217774691091437e-06,
      "loss": 0.3944,
      "step": 1580
    },
    {
      "epoch": 0.6563146997929606,
      "grad_norm": 2.2470810475680585,
      "learning_rate": 3.1880437415764936e-06,
      "loss": 0.3815,
      "step": 1585
    },
    {
      "epoch": 0.6583850931677019,
      "grad_norm": 2.423539483725975,
      "learning_rate": 3.1544046957596746e-06,
      "loss": 0.3812,
      "step": 1590
    },
    {
      "epoch": 0.660455486542443,
      "grad_norm": 2.4639445576402466,
      "learning_rate": 3.1208620894288105e-06,
      "loss": 0.3856,
      "step": 1595
    },
    {
      "epoch": 0.6625258799171843,
      "grad_norm": 2.5119740354112294,
      "learning_rate": 3.087417675314687e-06,
      "loss": 0.3875,
      "step": 1600
    },
    {
      "epoch": 0.6645962732919255,
      "grad_norm": 2.338985888997888,
      "learning_rate": 3.0540732010171704e-06,
      "loss": 0.3829,
      "step": 1605
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 2.4458177835617088,
      "learning_rate": 3.0208304089138873e-06,
      "loss": 0.374,
      "step": 1610
    },
    {
      "epoch": 0.6687370600414079,
      "grad_norm": 2.488509970043159,
      "learning_rate": 2.9876910360691707e-06,
      "loss": 0.381,
      "step": 1615
    },
    {
      "epoch": 0.6708074534161491,
      "grad_norm": 2.4903330967689423,
      "learning_rate": 2.9546568141433007e-06,
      "loss": 0.3905,
      "step": 1620
    },
    {
      "epoch": 0.6728778467908902,
      "grad_norm": 2.4193684443863557,
      "learning_rate": 2.921729469302018e-06,
      "loss": 0.3701,
      "step": 1625
    },
    {
      "epoch": 0.6749482401656315,
      "grad_norm": 2.509686800128473,
      "learning_rate": 2.8889107221263203e-06,
      "loss": 0.3847,
      "step": 1630
    },
    {
      "epoch": 0.6770186335403726,
      "grad_norm": 2.311241975406654,
      "learning_rate": 2.856202287522556e-06,
      "loss": 0.3819,
      "step": 1635
    },
    {
      "epoch": 0.6790890269151139,
      "grad_norm": 2.29165835256008,
      "learning_rate": 2.8236058746328203e-06,
      "loss": 0.3711,
      "step": 1640
    },
    {
      "epoch": 0.6811594202898551,
      "grad_norm": 2.448953347522542,
      "learning_rate": 2.7911231867456345e-06,
      "loss": 0.3694,
      "step": 1645
    },
    {
      "epoch": 0.6832298136645962,
      "grad_norm": 2.377258062615133,
      "learning_rate": 2.7587559212069554e-06,
      "loss": 0.3824,
      "step": 1650
    },
    {
      "epoch": 0.6853002070393375,
      "grad_norm": 2.3413695990375145,
      "learning_rate": 2.72650576933147e-06,
      "loss": 0.3772,
      "step": 1655
    },
    {
      "epoch": 0.6873706004140787,
      "grad_norm": 2.5551246519692072,
      "learning_rate": 2.6943744163142292e-06,
      "loss": 0.3728,
      "step": 1660
    },
    {
      "epoch": 0.6894409937888198,
      "grad_norm": 2.3533635279053247,
      "learning_rate": 2.662363541142585e-06,
      "loss": 0.3711,
      "step": 1665
    },
    {
      "epoch": 0.6915113871635611,
      "grad_norm": 2.3353654763094163,
      "learning_rate": 2.6304748165084503e-06,
      "loss": 0.3608,
      "step": 1670
    },
    {
      "epoch": 0.6935817805383023,
      "grad_norm": 2.274006785717291,
      "learning_rate": 2.5987099087209035e-06,
      "loss": 0.3716,
      "step": 1675
    },
    {
      "epoch": 0.6956521739130435,
      "grad_norm": 2.295142139581534,
      "learning_rate": 2.5670704776191158e-06,
      "loss": 0.3757,
      "step": 1680
    },
    {
      "epoch": 0.6977225672877847,
      "grad_norm": 2.483766482734329,
      "learning_rate": 2.5355581764856186e-06,
      "loss": 0.3696,
      "step": 1685
    },
    {
      "epoch": 0.6997929606625258,
      "grad_norm": 2.3582111413078835,
      "learning_rate": 2.5041746519599036e-06,
      "loss": 0.3676,
      "step": 1690
    },
    {
      "epoch": 0.7018633540372671,
      "grad_norm": 2.291398876690565,
      "learning_rate": 2.4729215439523945e-06,
      "loss": 0.3719,
      "step": 1695
    },
    {
      "epoch": 0.7039337474120083,
      "grad_norm": 2.5470639986375634,
      "learning_rate": 2.441800485558748e-06,
      "loss": 0.3676,
      "step": 1700
    },
    {
      "epoch": 0.7060041407867494,
      "grad_norm": 2.505975101239771,
      "learning_rate": 2.410813102974512e-06,
      "loss": 0.3657,
      "step": 1705
    },
    {
      "epoch": 0.7080745341614907,
      "grad_norm": 2.3892006117690796,
      "learning_rate": 2.3799610154101603e-06,
      "loss": 0.3688,
      "step": 1710
    },
    {
      "epoch": 0.7101449275362319,
      "grad_norm": 2.4856029935751884,
      "learning_rate": 2.3492458350064805e-06,
      "loss": 0.371,
      "step": 1715
    },
    {
      "epoch": 0.7122153209109731,
      "grad_norm": 2.2424244033057046,
      "learning_rate": 2.3186691667503315e-06,
      "loss": 0.3663,
      "step": 1720
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 2.279830929996751,
      "learning_rate": 2.288232608390778e-06,
      "loss": 0.3737,
      "step": 1725
    },
    {
      "epoch": 0.7163561076604554,
      "grad_norm": 2.6139030779076906,
      "learning_rate": 2.2579377503555967e-06,
      "loss": 0.3572,
      "step": 1730
    },
    {
      "epoch": 0.7184265010351967,
      "grad_norm": 2.3674642700390005,
      "learning_rate": 2.227786175668181e-06,
      "loss": 0.3697,
      "step": 1735
    },
    {
      "epoch": 0.7204968944099379,
      "grad_norm": 2.5964819008050344,
      "learning_rate": 2.1977794598648106e-06,
      "loss": 0.3644,
      "step": 1740
    },
    {
      "epoch": 0.722567287784679,
      "grad_norm": 2.43849421908629,
      "learning_rate": 2.167919170912333e-06,
      "loss": 0.3605,
      "step": 1745
    },
    {
      "epoch": 0.7246376811594203,
      "grad_norm": 2.334477611501331,
      "learning_rate": 2.138206869126222e-06,
      "loss": 0.3494,
      "step": 1750
    },
    {
      "epoch": 0.7267080745341615,
      "grad_norm": 2.402345510430816,
      "learning_rate": 2.108644107089055e-06,
      "loss": 0.3537,
      "step": 1755
    },
    {
      "epoch": 0.7287784679089027,
      "grad_norm": 2.380966085261301,
      "learning_rate": 2.0792324295693782e-06,
      "loss": 0.3629,
      "step": 1760
    },
    {
      "epoch": 0.7308488612836439,
      "grad_norm": 2.3999413524190483,
      "learning_rate": 2.0499733734409884e-06,
      "loss": 0.363,
      "step": 1765
    },
    {
      "epoch": 0.7329192546583851,
      "grad_norm": 2.412884814797146,
      "learning_rate": 2.020868467602623e-06,
      "loss": 0.3601,
      "step": 1770
    },
    {
      "epoch": 0.7349896480331263,
      "grad_norm": 2.5149920574704083,
      "learning_rate": 1.9919192328980703e-06,
      "loss": 0.3633,
      "step": 1775
    },
    {
      "epoch": 0.7370600414078675,
      "grad_norm": 2.396339910723278,
      "learning_rate": 1.9631271820367058e-06,
      "loss": 0.3461,
      "step": 1780
    },
    {
      "epoch": 0.7391304347826086,
      "grad_norm": 2.8475210222635052,
      "learning_rate": 1.9344938195144353e-06,
      "loss": 0.3543,
      "step": 1785
    },
    {
      "epoch": 0.7412008281573499,
      "grad_norm": 2.2976281524106157,
      "learning_rate": 1.9060206415350896e-06,
      "loss": 0.3594,
      "step": 1790
    },
    {
      "epoch": 0.7432712215320911,
      "grad_norm": 2.310572205925547,
      "learning_rate": 1.8777091359322386e-06,
      "loss": 0.3532,
      "step": 1795
    },
    {
      "epoch": 0.7453416149068323,
      "grad_norm": 2.368001627955122,
      "learning_rate": 1.8495607820914451e-06,
      "loss": 0.3497,
      "step": 1800
    },
    {
      "epoch": 0.7474120082815735,
      "grad_norm": 2.3386052773816703,
      "learning_rate": 1.8215770508729602e-06,
      "loss": 0.3486,
      "step": 1805
    },
    {
      "epoch": 0.7494824016563147,
      "grad_norm": 2.2988828371262238,
      "learning_rate": 1.7937594045348666e-06,
      "loss": 0.3597,
      "step": 1810
    },
    {
      "epoch": 0.7515527950310559,
      "grad_norm": 2.5624431500836824,
      "learning_rate": 1.7661092966566728e-06,
      "loss": 0.3469,
      "step": 1815
    },
    {
      "epoch": 0.7536231884057971,
      "grad_norm": 2.4189462246329247,
      "learning_rate": 1.7386281720633569e-06,
      "loss": 0.3588,
      "step": 1820
    },
    {
      "epoch": 0.7556935817805382,
      "grad_norm": 2.3388294710115733,
      "learning_rate": 1.7113174667498612e-06,
      "loss": 0.3515,
      "step": 1825
    },
    {
      "epoch": 0.7577639751552795,
      "grad_norm": 2.3830945742175706,
      "learning_rate": 1.6841786078060674e-06,
      "loss": 0.3578,
      "step": 1830
    },
    {
      "epoch": 0.7598343685300207,
      "grad_norm": 2.3252394086801362,
      "learning_rate": 1.6572130133422204e-06,
      "loss": 0.3364,
      "step": 1835
    },
    {
      "epoch": 0.7619047619047619,
      "grad_norm": 2.5473883800899815,
      "learning_rate": 1.630422092414823e-06,
      "loss": 0.3539,
      "step": 1840
    },
    {
      "epoch": 0.7639751552795031,
      "grad_norm": 2.4523578357427596,
      "learning_rate": 1.6038072449530118e-06,
      "loss": 0.3485,
      "step": 1845
    },
    {
      "epoch": 0.7660455486542443,
      "grad_norm": 2.376864298598383,
      "learning_rate": 1.5773698616854065e-06,
      "loss": 0.3498,
      "step": 1850
    },
    {
      "epoch": 0.7681159420289855,
      "grad_norm": 2.2692629818358983,
      "learning_rate": 1.5511113240674364e-06,
      "loss": 0.3361,
      "step": 1855
    },
    {
      "epoch": 0.7701863354037267,
      "grad_norm": 2.2717444542780267,
      "learning_rate": 1.5250330042091543e-06,
      "loss": 0.3505,
      "step": 1860
    },
    {
      "epoch": 0.772256728778468,
      "grad_norm": 2.4283728013011143,
      "learning_rate": 1.4991362648035368e-06,
      "loss": 0.3382,
      "step": 1865
    },
    {
      "epoch": 0.7743271221532091,
      "grad_norm": 2.4498521782000666,
      "learning_rate": 1.4734224590552844e-06,
      "loss": 0.3507,
      "step": 1870
    },
    {
      "epoch": 0.7763975155279503,
      "grad_norm": 2.3883247184680823,
      "learning_rate": 1.4478929306101025e-06,
      "loss": 0.3352,
      "step": 1875
    },
    {
      "epoch": 0.7784679089026915,
      "grad_norm": 2.2773809296933565,
      "learning_rate": 1.4225490134844967e-06,
      "loss": 0.3458,
      "step": 1880
    },
    {
      "epoch": 0.7805383022774327,
      "grad_norm": 2.2048664941077636,
      "learning_rate": 1.3973920319960654e-06,
      "loss": 0.3413,
      "step": 1885
    },
    {
      "epoch": 0.782608695652174,
      "grad_norm": 2.4798855049257855,
      "learning_rate": 1.3724233006942972e-06,
      "loss": 0.3491,
      "step": 1890
    },
    {
      "epoch": 0.7846790890269151,
      "grad_norm": 2.4736183124807085,
      "learning_rate": 1.347644124291883e-06,
      "loss": 0.3424,
      "step": 1895
    },
    {
      "epoch": 0.7867494824016563,
      "grad_norm": 2.286682122373882,
      "learning_rate": 1.323055797596532e-06,
      "loss": 0.3429,
      "step": 1900
    },
    {
      "epoch": 0.7888198757763976,
      "grad_norm": 2.4109030453829137,
      "learning_rate": 1.2986596054433255e-06,
      "loss": 0.3539,
      "step": 1905
    },
    {
      "epoch": 0.7908902691511387,
      "grad_norm": 2.2527201235971157,
      "learning_rate": 1.2744568226275662e-06,
      "loss": 0.3413,
      "step": 1910
    },
    {
      "epoch": 0.7929606625258799,
      "grad_norm": 2.330930674364272,
      "learning_rate": 1.2504487138381782e-06,
      "loss": 0.3354,
      "step": 1915
    },
    {
      "epoch": 0.7950310559006211,
      "grad_norm": 2.479292963434977,
      "learning_rate": 1.2266365335916086e-06,
      "loss": 0.3426,
      "step": 1920
    },
    {
      "epoch": 0.7971014492753623,
      "grad_norm": 2.3313073595666416,
      "learning_rate": 1.203021526166287e-06,
      "loss": 0.3416,
      "step": 1925
    },
    {
      "epoch": 0.7991718426501035,
      "grad_norm": 2.5129741155310503,
      "learning_rate": 1.179604925537598e-06,
      "loss": 0.3306,
      "step": 1930
    },
    {
      "epoch": 0.8012422360248447,
      "grad_norm": 2.3092564913659834,
      "learning_rate": 1.1563879553134072e-06,
      "loss": 0.3332,
      "step": 1935
    },
    {
      "epoch": 0.8033126293995859,
      "grad_norm": 2.4752738147285607,
      "learning_rate": 1.1333718286701162e-06,
      "loss": 0.3361,
      "step": 1940
    },
    {
      "epoch": 0.8053830227743272,
      "grad_norm": 2.2093947429221275,
      "learning_rate": 1.110557748289275e-06,
      "loss": 0.335,
      "step": 1945
    },
    {
      "epoch": 0.8074534161490683,
      "grad_norm": 2.300100047692853,
      "learning_rate": 1.0879469062947369e-06,
      "loss": 0.3363,
      "step": 1950
    },
    {
      "epoch": 0.8095238095238095,
      "grad_norm": 2.43658926819036,
      "learning_rate": 1.0655404841903648e-06,
      "loss": 0.3374,
      "step": 1955
    },
    {
      "epoch": 0.8115942028985508,
      "grad_norm": 2.2921815004097876,
      "learning_rate": 1.0433396527982893e-06,
      "loss": 0.335,
      "step": 1960
    },
    {
      "epoch": 0.8136645962732919,
      "grad_norm": 2.310829513039428,
      "learning_rate": 1.0213455721977356e-06,
      "loss": 0.3335,
      "step": 1965
    },
    {
      "epoch": 0.8157349896480331,
      "grad_norm": 2.3643435111135434,
      "learning_rate": 9.995593916643981e-07,
      "loss": 0.3405,
      "step": 1970
    },
    {
      "epoch": 0.8178053830227743,
      "grad_norm": 2.3240624032063306,
      "learning_rate": 9.779822496103902e-07,
      "loss": 0.3299,
      "step": 1975
    },
    {
      "epoch": 0.8198757763975155,
      "grad_norm": 2.5318838270330195,
      "learning_rate": 9.566152735247576e-07,
      "loss": 0.3434,
      "step": 1980
    },
    {
      "epoch": 0.8219461697722568,
      "grad_norm": 2.3420007259490228,
      "learning_rate": 9.354595799145627e-07,
      "loss": 0.3421,
      "step": 1985
    },
    {
      "epoch": 0.8240165631469979,
      "grad_norm": 2.335473751400107,
      "learning_rate": 9.145162742465424e-07,
      "loss": 0.33,
      "step": 1990
    },
    {
      "epoch": 0.8260869565217391,
      "grad_norm": 2.5508864216623444,
      "learning_rate": 8.937864508893396e-07,
      "loss": 0.333,
      "step": 1995
    },
    {
      "epoch": 0.8281573498964804,
      "grad_norm": 2.488334856667467,
      "learning_rate": 8.732711930563259e-07,
      "loss": 0.3365,
      "step": 2000
    },
    {
      "epoch": 0.8302277432712215,
      "grad_norm": 2.2859380568393712,
      "learning_rate": 8.529715727489912e-07,
      "loss": 0.3318,
      "step": 2005
    },
    {
      "epoch": 0.8322981366459627,
      "grad_norm": 2.5733587910845253,
      "learning_rate": 8.32888650700936e-07,
      "loss": 0.3358,
      "step": 2010
    },
    {
      "epoch": 0.8343685300207039,
      "grad_norm": 2.4265851072259124,
      "learning_rate": 8.130234763224342e-07,
      "loss": 0.335,
      "step": 2015
    },
    {
      "epoch": 0.8364389233954451,
      "grad_norm": 2.2464388261333257,
      "learning_rate": 7.93377087645607e-07,
      "loss": 0.3176,
      "step": 2020
    },
    {
      "epoch": 0.8385093167701864,
      "grad_norm": 2.3494420369788154,
      "learning_rate": 7.739505112701751e-07,
      "loss": 0.3332,
      "step": 2025
    },
    {
      "epoch": 0.8405797101449275,
      "grad_norm": 2.4054514152177653,
      "learning_rate": 7.547447623098191e-07,
      "loss": 0.3311,
      "step": 2030
    },
    {
      "epoch": 0.8426501035196687,
      "grad_norm": 2.261913018941359,
      "learning_rate": 7.357608443391312e-07,
      "loss": 0.3256,
      "step": 2035
    },
    {
      "epoch": 0.84472049689441,
      "grad_norm": 2.2421739628590043,
      "learning_rate": 7.169997493411774e-07,
      "loss": 0.3208,
      "step": 2040
    },
    {
      "epoch": 0.8467908902691511,
      "grad_norm": 2.3654167525865732,
      "learning_rate": 6.984624576556647e-07,
      "loss": 0.3218,
      "step": 2045
    },
    {
      "epoch": 0.8488612836438924,
      "grad_norm": 2.4284908975712574,
      "learning_rate": 6.801499379277115e-07,
      "loss": 0.3311,
      "step": 2050
    },
    {
      "epoch": 0.8509316770186336,
      "grad_norm": 2.22925170204331,
      "learning_rate": 6.620631470572314e-07,
      "loss": 0.3313,
      "step": 2055
    },
    {
      "epoch": 0.8530020703933747,
      "grad_norm": 2.5749952139899652,
      "learning_rate": 6.442030301489338e-07,
      "loss": 0.3301,
      "step": 2060
    },
    {
      "epoch": 0.855072463768116,
      "grad_norm": 2.2674939075313048,
      "learning_rate": 6.265705204629402e-07,
      "loss": 0.3081,
      "step": 2065
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 2.496812533980436,
      "learning_rate": 6.091665393660112e-07,
      "loss": 0.3262,
      "step": 2070
    },
    {
      "epoch": 0.8592132505175983,
      "grad_norm": 2.4717944573126287,
      "learning_rate": 5.919919962834069e-07,
      "loss": 0.3164,
      "step": 2075
    },
    {
      "epoch": 0.8612836438923396,
      "grad_norm": 2.3257141732938322,
      "learning_rate": 5.750477886513656e-07,
      "loss": 0.3314,
      "step": 2080
    },
    {
      "epoch": 0.8633540372670807,
      "grad_norm": 2.372908668701669,
      "learning_rate": 5.58334801870209e-07,
      "loss": 0.3257,
      "step": 2085
    },
    {
      "epoch": 0.865424430641822,
      "grad_norm": 2.344278855687297,
      "learning_rate": 5.418539092580727e-07,
      "loss": 0.3346,
      "step": 2090
    },
    {
      "epoch": 0.8674948240165632,
      "grad_norm": 2.3324906572581976,
      "learning_rate": 5.256059720052787e-07,
      "loss": 0.3131,
      "step": 2095
    },
    {
      "epoch": 0.8695652173913043,
      "grad_norm": 2.4514845156052503,
      "learning_rate": 5.095918391293303e-07,
      "loss": 0.3279,
      "step": 2100
    },
    {
      "epoch": 0.8716356107660456,
      "grad_norm": 2.3203875730343144,
      "learning_rate": 4.938123474305473e-07,
      "loss": 0.3195,
      "step": 2105
    },
    {
      "epoch": 0.8737060041407867,
      "grad_norm": 2.440973750364137,
      "learning_rate": 4.782683214483413e-07,
      "loss": 0.3217,
      "step": 2110
    },
    {
      "epoch": 0.8757763975155279,
      "grad_norm": 2.424738960596876,
      "learning_rate": 4.629605734181331e-07,
      "loss": 0.3214,
      "step": 2115
    },
    {
      "epoch": 0.8778467908902692,
      "grad_norm": 2.3840075746764726,
      "learning_rate": 4.478899032289047e-07,
      "loss": 0.3219,
      "step": 2120
    },
    {
      "epoch": 0.8799171842650103,
      "grad_norm": 2.4240393254281516,
      "learning_rate": 4.3305709838140755e-07,
      "loss": 0.3259,
      "step": 2125
    },
    {
      "epoch": 0.8819875776397516,
      "grad_norm": 2.310253664818122,
      "learning_rate": 4.1846293394700645e-07,
      "loss": 0.3242,
      "step": 2130
    },
    {
      "epoch": 0.8840579710144928,
      "grad_norm": 2.517972583368381,
      "learning_rate": 4.041081725271856e-07,
      "loss": 0.3229,
      "step": 2135
    },
    {
      "epoch": 0.8861283643892339,
      "grad_norm": 2.428798401646251,
      "learning_rate": 3.8999356421369426e-07,
      "loss": 0.3271,
      "step": 2140
    },
    {
      "epoch": 0.8881987577639752,
      "grad_norm": 2.450001348873468,
      "learning_rate": 3.7611984654935707e-07,
      "loss": 0.3269,
      "step": 2145
    },
    {
      "epoch": 0.8902691511387164,
      "grad_norm": 2.3842094730267496,
      "learning_rate": 3.6248774448952695e-07,
      "loss": 0.3245,
      "step": 2150
    },
    {
      "epoch": 0.8923395445134575,
      "grad_norm": 2.3496216155823877,
      "learning_rate": 3.490979703642117e-07,
      "loss": 0.3195,
      "step": 2155
    },
    {
      "epoch": 0.8944099378881988,
      "grad_norm": 2.4989229920823774,
      "learning_rate": 3.35951223840848e-07,
      "loss": 0.3289,
      "step": 2160
    },
    {
      "epoch": 0.8964803312629399,
      "grad_norm": 2.215427120738685,
      "learning_rate": 3.2304819188773815e-07,
      "loss": 0.3137,
      "step": 2165
    },
    {
      "epoch": 0.8985507246376812,
      "grad_norm": 2.3056359567362383,
      "learning_rate": 3.1038954873816027e-07,
      "loss": 0.3181,
      "step": 2170
    },
    {
      "epoch": 0.9006211180124224,
      "grad_norm": 2.344264414609259,
      "learning_rate": 2.9797595585512986e-07,
      "loss": 0.3125,
      "step": 2175
    },
    {
      "epoch": 0.9026915113871635,
      "grad_norm": 2.2624241698250507,
      "learning_rate": 2.858080618968423e-07,
      "loss": 0.3265,
      "step": 2180
    },
    {
      "epoch": 0.9047619047619048,
      "grad_norm": 2.408379187199429,
      "learning_rate": 2.738865026827714e-07,
      "loss": 0.3304,
      "step": 2185
    },
    {
      "epoch": 0.906832298136646,
      "grad_norm": 2.375361674460972,
      "learning_rate": 2.6221190116045126e-07,
      "loss": 0.3105,
      "step": 2190
    },
    {
      "epoch": 0.9089026915113871,
      "grad_norm": 2.3986823821063723,
      "learning_rate": 2.507848673729224e-07,
      "loss": 0.3167,
      "step": 2195
    },
    {
      "epoch": 0.9109730848861284,
      "grad_norm": 2.300915767213228,
      "learning_rate": 2.3960599842685394e-07,
      "loss": 0.3174,
      "step": 2200
    },
    {
      "epoch": 0.9130434782608695,
      "grad_norm": 2.4752814085308525,
      "learning_rate": 2.2867587846134188e-07,
      "loss": 0.3205,
      "step": 2205
    },
    {
      "epoch": 0.9151138716356108,
      "grad_norm": 2.618957183855706,
      "learning_rate": 2.179950786173879e-07,
      "loss": 0.3105,
      "step": 2210
    },
    {
      "epoch": 0.917184265010352,
      "grad_norm": 2.4747630022844604,
      "learning_rate": 2.07564157008055e-07,
      "loss": 0.3206,
      "step": 2215
    },
    {
      "epoch": 0.9192546583850931,
      "grad_norm": 2.35288736739077,
      "learning_rate": 1.9738365868930188e-07,
      "loss": 0.3243,
      "step": 2220
    },
    {
      "epoch": 0.9213250517598344,
      "grad_norm": 2.2353440364597175,
      "learning_rate": 1.8745411563150128e-07,
      "loss": 0.3216,
      "step": 2225
    },
    {
      "epoch": 0.9233954451345756,
      "grad_norm": 2.3590967551206834,
      "learning_rate": 1.7777604669164727e-07,
      "loss": 0.3169,
      "step": 2230
    },
    {
      "epoch": 0.9254658385093167,
      "grad_norm": 2.279766783196755,
      "learning_rate": 1.6834995758623685e-07,
      "loss": 0.3231,
      "step": 2235
    },
    {
      "epoch": 0.927536231884058,
      "grad_norm": 2.424705804446328,
      "learning_rate": 1.5917634086484778e-07,
      "loss": 0.3084,
      "step": 2240
    },
    {
      "epoch": 0.9296066252587992,
      "grad_norm": 2.284188592528974,
      "learning_rate": 1.5025567588439925e-07,
      "loss": 0.3207,
      "step": 2245
    },
    {
      "epoch": 0.9316770186335404,
      "grad_norm": 2.440907754264571,
      "learning_rate": 1.415884287841074e-07,
      "loss": 0.317,
      "step": 2250
    },
    {
      "epoch": 0.9337474120082816,
      "grad_norm": 2.4379439764772526,
      "learning_rate": 1.331750524611225e-07,
      "loss": 0.3072,
      "step": 2255
    },
    {
      "epoch": 0.9358178053830227,
      "grad_norm": 2.316142028863615,
      "learning_rate": 1.2501598654686532e-07,
      "loss": 0.3239,
      "step": 2260
    },
    {
      "epoch": 0.937888198757764,
      "grad_norm": 2.440325157626698,
      "learning_rate": 1.1711165738405749e-07,
      "loss": 0.3139,
      "step": 2265
    },
    {
      "epoch": 0.9399585921325052,
      "grad_norm": 2.5527087863191444,
      "learning_rate": 1.0946247800443899e-07,
      "loss": 0.3254,
      "step": 2270
    },
    {
      "epoch": 0.9420289855072463,
      "grad_norm": 2.3791340835281636,
      "learning_rate": 1.0206884810718964e-07,
      "loss": 0.3288,
      "step": 2275
    },
    {
      "epoch": 0.9440993788819876,
      "grad_norm": 2.4013244243260012,
      "learning_rate": 9.493115403803876e-08,
      "loss": 0.3273,
      "step": 2280
    },
    {
      "epoch": 0.9461697722567288,
      "grad_norm": 2.5204426609428237,
      "learning_rate": 8.804976876908178e-08,
      "loss": 0.3123,
      "step": 2285
    },
    {
      "epoch": 0.94824016563147,
      "grad_norm": 2.237652421457841,
      "learning_rate": 8.142505187928861e-08,
      "loss": 0.3179,
      "step": 2290
    },
    {
      "epoch": 0.9503105590062112,
      "grad_norm": 2.4231612218683343,
      "learning_rate": 7.505734953571486e-08,
      "loss": 0.3189,
      "step": 2295
    },
    {
      "epoch": 0.9523809523809523,
      "grad_norm": 2.2082939472481162,
      "learning_rate": 6.894699447541231e-08,
      "loss": 0.3098,
      "step": 2300
    },
    {
      "epoch": 0.9544513457556936,
      "grad_norm": 2.406819987358586,
      "learning_rate": 6.309430598804234e-08,
      "loss": 0.3035,
      "step": 2305
    },
    {
      "epoch": 0.9565217391304348,
      "grad_norm": 2.324311672019134,
      "learning_rate": 5.7499589899193134e-08,
      "loss": 0.3126,
      "step": 2310
    },
    {
      "epoch": 0.9585921325051759,
      "grad_norm": 2.513161161903101,
      "learning_rate": 5.216313855439858e-08,
      "loss": 0.3201,
      "step": 2315
    },
    {
      "epoch": 0.9606625258799172,
      "grad_norm": 2.4580927731114257,
      "learning_rate": 4.7085230803860513e-08,
      "loss": 0.3205,
      "step": 2320
    },
    {
      "epoch": 0.9627329192546584,
      "grad_norm": 2.446929983666297,
      "learning_rate": 4.2266131987880876e-08,
      "loss": 0.3206,
      "step": 2325
    },
    {
      "epoch": 0.9648033126293996,
      "grad_norm": 2.27874740868648,
      "learning_rate": 3.770609392299285e-08,
      "loss": 0.3131,
      "step": 2330
    },
    {
      "epoch": 0.9668737060041408,
      "grad_norm": 2.3849246524911107,
      "learning_rate": 3.340535488880525e-08,
      "loss": 0.3277,
      "step": 2335
    },
    {
      "epoch": 0.968944099378882,
      "grad_norm": 2.421083844991871,
      "learning_rate": 2.9364139615550846e-08,
      "loss": 0.3186,
      "step": 2340
    },
    {
      "epoch": 0.9710144927536232,
      "grad_norm": 2.4038973912823334,
      "learning_rate": 2.5582659272341827e-08,
      "loss": 0.3098,
      "step": 2345
    },
    {
      "epoch": 0.9730848861283644,
      "grad_norm": 2.5427340924398694,
      "learning_rate": 2.206111145613865e-08,
      "loss": 0.3156,
      "step": 2350
    },
    {
      "epoch": 0.9751552795031055,
      "grad_norm": 2.5031732620266753,
      "learning_rate": 1.8799680181421087e-08,
      "loss": 0.3302,
      "step": 2355
    },
    {
      "epoch": 0.9772256728778468,
      "grad_norm": 2.535683472708062,
      "learning_rate": 1.5798535870575893e-08,
      "loss": 0.3162,
      "step": 2360
    },
    {
      "epoch": 0.979296066252588,
      "grad_norm": 2.492815206410193,
      "learning_rate": 1.30578353449895e-08,
      "loss": 0.3112,
      "step": 2365
    },
    {
      "epoch": 0.9813664596273292,
      "grad_norm": 2.321359601019582,
      "learning_rate": 1.0577721816854015e-08,
      "loss": 0.3139,
      "step": 2370
    },
    {
      "epoch": 0.9834368530020704,
      "grad_norm": 2.376543987859838,
      "learning_rate": 8.35832488168431e-09,
      "loss": 0.318,
      "step": 2375
    },
    {
      "epoch": 0.9855072463768116,
      "grad_norm": 2.4905741784802693,
      "learning_rate": 6.399760511546227e-09,
      "loss": 0.3196,
      "step": 2380
    },
    {
      "epoch": 0.9875776397515528,
      "grad_norm": 2.4503057250197617,
      "learning_rate": 4.7021310489953065e-09,
      "loss": 0.3158,
      "step": 2385
    },
    {
      "epoch": 0.989648033126294,
      "grad_norm": 2.382766716100949,
      "learning_rate": 3.265525201731623e-09,
      "loss": 0.3154,
      "step": 2390
    },
    {
      "epoch": 0.9917184265010351,
      "grad_norm": 2.341666724587077,
      "learning_rate": 2.090018037960717e-09,
      "loss": 0.3122,
      "step": 2395
    },
    {
      "epoch": 0.9937888198757764,
      "grad_norm": 2.398969814052762,
      "learning_rate": 1.1756709824750633e-09,
      "loss": 0.3121,
      "step": 2400
    },
    {
      "epoch": 0.9958592132505176,
      "grad_norm": 2.320173528449576,
      "learning_rate": 5.225318134399749e-10,
      "loss": 0.3239,
      "step": 2405
    },
    {
      "epoch": 0.9979296066252588,
      "grad_norm": 2.3737353000585055,
      "learning_rate": 1.306346599011521e-10,
      "loss": 0.3218,
      "step": 2410
    },
    {
      "epoch": 1.0,
      "grad_norm": 2.319768008564562,
      "learning_rate": 0.0,
      "loss": 0.3215,
      "step": 2415
    },
    {
      "epoch": 1.0,
      "eval_runtime": 2.6812,
      "eval_samples_per_second": 3.73,
      "eval_steps_per_second": 1.119,
      "step": 2415
    },
    {
      "epoch": 1.0,
      "step": 2415,
      "total_flos": 252825934233600.0,
      "train_loss": 0.5143274352170419,
      "train_runtime": 26101.4191,
      "train_samples_per_second": 1.48,
      "train_steps_per_second": 0.093
    }
  ],
  "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
}