{
  "best_metric": 1.203166127204895,
  "best_model_checkpoint": "/Users/bbunzeck/Documents/german-llamas/cxn-llamas/cds-llama85/checkpoint-6003",
  "epoch": 0.9999250093738282,
  "eval_steps": 667,
  "global_step": 6667,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0014998125234345708,
      "grad_norm": 4.941356658935547,
      "learning_rate": 1.4999999999999999e-05,
      "loss": 4.5697,
      "step": 10
    },
    {
      "epoch": 0.0029996250468691415,
      "grad_norm": 3.326807737350464,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 4.207,
      "step": 20
    },
    {
      "epoch": 0.0044994375703037125,
      "grad_norm": 2.287435531616211,
      "learning_rate": 4.4999999999999996e-05,
      "loss": 3.7885,
      "step": 30
    },
    {
      "epoch": 0.005999250093738283,
      "grad_norm": 1.804822325706482,
      "learning_rate": 5.9999999999999995e-05,
      "loss": 3.5512,
      "step": 40
    },
    {
      "epoch": 0.0074990626171728535,
      "grad_norm": 1.4687652587890625,
      "learning_rate": 7.5e-05,
      "loss": 3.4201,
      "step": 50
    },
    {
      "epoch": 0.008998875140607425,
      "grad_norm": 1.0859005451202393,
      "learning_rate": 8.999999999999999e-05,
      "loss": 3.2891,
      "step": 60
    },
    {
      "epoch": 0.010498687664041995,
      "grad_norm": 0.9512425661087036,
      "learning_rate": 0.00010499999999999999,
      "loss": 3.1539,
      "step": 70
    },
    {
      "epoch": 0.011998500187476566,
      "grad_norm": 1.2439535856246948,
      "learning_rate": 0.00011999999999999999,
      "loss": 2.919,
      "step": 80
    },
    {
      "epoch": 0.013498312710911136,
      "grad_norm": 1.0196483135223389,
      "learning_rate": 0.000135,
      "loss": 2.728,
      "step": 90
    },
    {
      "epoch": 0.014998125234345707,
      "grad_norm": 3.3797309398651123,
      "learning_rate": 0.00015,
      "loss": 2.5985,
      "step": 100
    },
    {
      "epoch": 0.016497937757780277,
      "grad_norm": 1.4080921411514282,
      "learning_rate": 0.000165,
      "loss": 2.4991,
      "step": 110
    },
    {
      "epoch": 0.01799775028121485,
      "grad_norm": 1.177751898765564,
      "learning_rate": 0.00017999999999999998,
      "loss": 2.4234,
      "step": 120
    },
    {
      "epoch": 0.01949756280464942,
      "grad_norm": 1.9201545715332031,
      "learning_rate": 0.000195,
      "loss": 2.36,
      "step": 130
    },
    {
      "epoch": 0.02099737532808399,
      "grad_norm": 1.969179630279541,
      "learning_rate": 0.00020999999999999998,
      "loss": 2.3224,
      "step": 140
    },
    {
      "epoch": 0.02249718785151856,
      "grad_norm": 1.7428991794586182,
      "learning_rate": 0.000225,
      "loss": 2.2435,
      "step": 150
    },
    {
      "epoch": 0.023997000374953132,
      "grad_norm": 1.231548547744751,
      "learning_rate": 0.00023999999999999998,
      "loss": 2.1859,
      "step": 160
    },
    {
      "epoch": 0.0254968128983877,
      "grad_norm": 1.4751386642456055,
      "learning_rate": 0.00025499999999999996,
      "loss": 2.1358,
      "step": 170
    },
    {
      "epoch": 0.02699662542182227,
      "grad_norm": 1.457726001739502,
      "learning_rate": 0.00027,
      "loss": 2.0645,
      "step": 180
    },
    {
      "epoch": 0.028496437945256844,
      "grad_norm": 1.2510344982147217,
      "learning_rate": 0.000285,
      "loss": 2.0373,
      "step": 190
    },
    {
      "epoch": 0.029996250468691414,
      "grad_norm": 1.5325850248336792,
      "learning_rate": 0.0003,
      "loss": 1.9925,
      "step": 200
    },
    {
      "epoch": 0.031496062992125984,
      "grad_norm": 1.3607863187789917,
      "learning_rate": 0.0002999982300767559,
      "loss": 1.9542,
      "step": 210
    },
    {
      "epoch": 0.032995875515560553,
      "grad_norm": 1.13303804397583,
      "learning_rate": 0.000299992920348792,
      "loss": 1.9144,
      "step": 220
    },
    {
      "epoch": 0.03449568803899512,
      "grad_norm": 1.341699242591858,
      "learning_rate": 0.0002999840709414124,
      "loss": 1.9061,
      "step": 230
    },
    {
      "epoch": 0.0359955005624297,
      "grad_norm": 1.2910326719284058,
      "learning_rate": 0.0002999716820634541,
      "loss": 1.8744,
      "step": 240
    },
    {
      "epoch": 0.03749531308586427,
      "grad_norm": 1.3730792999267578,
      "learning_rate": 0.000299955754007282,
      "loss": 1.8369,
      "step": 250
    },
    {
      "epoch": 0.03899512560929884,
      "grad_norm": 1.2474159002304077,
      "learning_rate": 0.00029993628714878185,
      "loss": 1.796,
      "step": 260
    },
    {
      "epoch": 0.04049493813273341,
      "grad_norm": 1.2987885475158691,
      "learning_rate": 0.00029991328194735155,
      "loss": 1.8051,
      "step": 270
    },
    {
      "epoch": 0.04199475065616798,
      "grad_norm": 1.0884069204330444,
      "learning_rate": 0.0002998867389458904,
      "loss": 1.7658,
      "step": 280
    },
    {
      "epoch": 0.04349456317960255,
      "grad_norm": 1.3922066688537598,
      "learning_rate": 0.00029985665877078595,
      "loss": 1.7632,
      "step": 290
    },
    {
      "epoch": 0.04499437570303712,
      "grad_norm": 1.1491564512252808,
      "learning_rate": 0.0002998230421318997,
      "loss": 1.7608,
      "step": 300
    },
    {
      "epoch": 0.046494188226471694,
      "grad_norm": 1.2716453075408936,
      "learning_rate": 0.0002997858898225498,
      "loss": 1.7231,
      "step": 310
    },
    {
      "epoch": 0.047994000749906264,
      "grad_norm": 1.3676432371139526,
      "learning_rate": 0.0002997452027194928,
      "loss": 1.7303,
      "step": 320
    },
    {
      "epoch": 0.049493813273340834,
      "grad_norm": 1.1510239839553833,
      "learning_rate": 0.0002997009817829027,
      "loss": 1.7337,
      "step": 330
    },
    {
      "epoch": 0.0509936257967754,
      "grad_norm": 1.2083615064620972,
      "learning_rate": 0.0002996532280563483,
      "loss": 1.708,
      "step": 340
    },
    {
      "epoch": 0.05249343832020997,
      "grad_norm": 1.0882530212402344,
      "learning_rate": 0.0002996019426667687,
      "loss": 1.6717,
      "step": 350
    },
    {
      "epoch": 0.05399325084364454,
      "grad_norm": 1.2023850679397583,
      "learning_rate": 0.00029954712682444656,
      "loss": 1.6565,
      "step": 360
    },
    {
      "epoch": 0.05549306336707911,
      "grad_norm": 1.0490646362304688,
      "learning_rate": 0.0002994887818229797,
      "loss": 1.639,
      "step": 370
    },
    {
      "epoch": 0.05699287589051369,
      "grad_norm": 1.0732958316802979,
      "learning_rate": 0.0002994269090392505,
      "loss": 1.6685,
      "step": 380
    },
    {
      "epoch": 0.05849268841394826,
      "grad_norm": 1.1127232313156128,
      "learning_rate": 0.00029936150993339325,
      "loss": 1.6365,
      "step": 390
    },
    {
      "epoch": 0.05999250093738283,
      "grad_norm": 1.1355839967727661,
      "learning_rate": 0.0002992925860487599,
      "loss": 1.6495,
      "step": 400
    },
    {
      "epoch": 0.0614923134608174,
      "grad_norm": 1.0531744956970215,
      "learning_rate": 0.0002992201390118837,
      "loss": 1.6279,
      "step": 410
    },
    {
      "epoch": 0.06299212598425197,
      "grad_norm": 1.165711522102356,
      "learning_rate": 0.00029914417053244054,
      "loss": 1.6342,
      "step": 420
    },
    {
      "epoch": 0.06449193850768654,
      "grad_norm": 1.1082898378372192,
      "learning_rate": 0.00029906468240320874,
      "loss": 1.6184,
      "step": 430
    },
    {
      "epoch": 0.06599175103112111,
      "grad_norm": 1.07638680934906,
      "learning_rate": 0.00029898167650002676,
      "loss": 1.6078,
      "step": 440
    },
    {
      "epoch": 0.06749156355455568,
      "grad_norm": 0.9482781887054443,
      "learning_rate": 0.0002988951547817491,
      "loss": 1.6024,
      "step": 450
    },
    {
      "epoch": 0.06899137607799025,
      "grad_norm": 0.9475975036621094,
      "learning_rate": 0.00029880511929019965,
      "loss": 1.6056,
      "step": 460
    },
    {
      "epoch": 0.07049118860142482,
      "grad_norm": 1.2477397918701172,
      "learning_rate": 0.0002987115721501239,
      "loss": 1.5891,
      "step": 470
    },
    {
      "epoch": 0.0719910011248594,
      "grad_norm": 1.0981507301330566,
      "learning_rate": 0.00029861451556913865,
      "loss": 1.5971,
      "step": 480
    },
    {
      "epoch": 0.07349081364829396,
      "grad_norm": 1.0291006565093994,
      "learning_rate": 0.00029851395183767983,
      "loss": 1.5748,
      "step": 490
    },
    {
      "epoch": 0.07499062617172854,
      "grad_norm": 1.0406434535980225,
      "learning_rate": 0.00029840988332894864,
      "loss": 1.5739,
      "step": 500
    },
    {
      "epoch": 0.0764904386951631,
      "grad_norm": 0.9979759454727173,
      "learning_rate": 0.00029830231249885537,
      "loss": 1.5593,
      "step": 510
    },
    {
      "epoch": 0.07799025121859768,
      "grad_norm": 1.1204349994659424,
      "learning_rate": 0.00029819124188596146,
      "loss": 1.5598,
      "step": 520
    },
    {
      "epoch": 0.07949006374203224,
      "grad_norm": 1.0350697040557861,
      "learning_rate": 0.00029807667411141977,
      "loss": 1.5625,
      "step": 530
    },
    {
      "epoch": 0.08098987626546682,
      "grad_norm": 1.0373060703277588,
      "learning_rate": 0.0002979586118789125,
      "loss": 1.5413,
      "step": 540
    },
    {
      "epoch": 0.0824896887889014,
      "grad_norm": 0.9631483554840088,
      "learning_rate": 0.0002978370579745876,
      "loss": 1.5431,
      "step": 550
    },
    {
      "epoch": 0.08398950131233596,
      "grad_norm": 0.9262831211090088,
      "learning_rate": 0.00029771201526699264,
      "loss": 1.5288,
      "step": 560
    },
    {
      "epoch": 0.08548931383577053,
      "grad_norm": 1.0329458713531494,
      "learning_rate": 0.0002975834867070077,
      "loss": 1.5172,
      "step": 570
    },
    {
      "epoch": 0.0869891263592051,
      "grad_norm": 0.9907341599464417,
      "learning_rate": 0.00029745147532777514,
      "loss": 1.5118,
      "step": 580
    },
    {
      "epoch": 0.08848893888263967,
      "grad_norm": 0.9761767983436584,
      "learning_rate": 0.0002973159842446285,
      "loss": 1.5067,
      "step": 590
    },
    {
      "epoch": 0.08998875140607424,
      "grad_norm": 0.9076977968215942,
      "learning_rate": 0.00029717701665501865,
      "loss": 1.5173,
      "step": 600
    },
    {
      "epoch": 0.09148856392950881,
      "grad_norm": 0.9733080863952637,
      "learning_rate": 0.00029703457583843846,
      "loss": 1.5101,
      "step": 610
    },
    {
      "epoch": 0.09298837645294339,
      "grad_norm": 0.965733528137207,
      "learning_rate": 0.00029688866515634546,
      "loss": 1.5192,
      "step": 620
    },
    {
      "epoch": 0.09448818897637795,
      "grad_norm": 0.9676783680915833,
      "learning_rate": 0.00029673928805208237,
      "loss": 1.5123,
      "step": 630
    },
    {
      "epoch": 0.09598800149981253,
      "grad_norm": 0.9860779047012329,
      "learning_rate": 0.00029658644805079606,
      "loss": 1.5205,
      "step": 640
    },
    {
      "epoch": 0.09748781402324709,
      "grad_norm": 1.026779294013977,
      "learning_rate": 0.00029643014875935404,
      "loss": 1.5175,
      "step": 650
    },
    {
      "epoch": 0.09898762654668167,
      "grad_norm": 0.9782848954200745,
      "learning_rate": 0.00029627039386625976,
      "loss": 1.4966,
      "step": 660
    },
    {
      "epoch": 0.10003749531308587,
      "eval_loss": 1.540096402168274,
      "eval_runtime": 34.7434,
      "eval_samples_per_second": 719.561,
      "eval_steps_per_second": 89.945,
      "step": 667
    },
    {
      "epoch": 0.10048743907011623,
      "grad_norm": 1.0175178050994873,
      "learning_rate": 0.0002961071871415651,
      "loss": 1.4742,
      "step": 670
    },
    {
      "epoch": 0.1019872515935508,
      "grad_norm": 0.9086681008338928,
      "learning_rate": 0.00029594053243678175,
      "loss": 1.5138,
      "step": 680
    },
    {
      "epoch": 0.10348706411698538,
      "grad_norm": 1.0384730100631714,
      "learning_rate": 0.00029577043368479017,
      "loss": 1.4645,
      "step": 690
    },
    {
      "epoch": 0.10498687664041995,
      "grad_norm": 0.9737691283226013,
      "learning_rate": 0.0002955968948997469,
      "loss": 1.4812,
      "step": 700
    },
    {
      "epoch": 0.10648668916385452,
      "grad_norm": 0.9272564053535461,
      "learning_rate": 0.00029541992017698956,
      "loss": 1.4489,
      "step": 710
    },
    {
      "epoch": 0.10798650168728909,
      "grad_norm": 0.9667676687240601,
      "learning_rate": 0.0002952395136929406,
      "loss": 1.4726,
      "step": 720
    },
    {
      "epoch": 0.10948631421072366,
      "grad_norm": 0.8968635201454163,
      "learning_rate": 0.00029505567970500833,
      "loss": 1.4585,
      "step": 730
    },
    {
      "epoch": 0.11098612673415822,
      "grad_norm": 0.9180886745452881,
      "learning_rate": 0.0002948684225514868,
      "loss": 1.4507,
      "step": 740
    },
    {
      "epoch": 0.1124859392575928,
      "grad_norm": 0.9335956573486328,
      "learning_rate": 0.0002946777466514531,
      "loss": 1.4707,
      "step": 750
    },
    {
      "epoch": 0.11398575178102738,
      "grad_norm": 1.0144034624099731,
      "learning_rate": 0.00029448365650466336,
      "loss": 1.4697,
      "step": 760
    },
    {
      "epoch": 0.11548556430446194,
      "grad_norm": 1.003219485282898,
      "learning_rate": 0.0002942861566914465,
      "loss": 1.455,
      "step": 770
    },
    {
      "epoch": 0.11698537682789652,
      "grad_norm": 0.9664227962493896,
      "learning_rate": 0.0002940852518725959,
      "loss": 1.4472,
      "step": 780
    },
    {
      "epoch": 0.11848518935133108,
      "grad_norm": 0.9362863898277283,
      "learning_rate": 0.0002938809467892596,
      "loss": 1.4411,
      "step": 790
    },
    {
      "epoch": 0.11998500187476566,
      "grad_norm": 0.9592990279197693,
      "learning_rate": 0.0002936732462628287,
      "loss": 1.4459,
      "step": 800
    },
    {
      "epoch": 0.12148481439820022,
      "grad_norm": 0.9843602180480957,
      "learning_rate": 0.0002934621551948229,
      "loss": 1.4439,
      "step": 810
    },
    {
      "epoch": 0.1229846269216348,
      "grad_norm": 0.9243733286857605,
      "learning_rate": 0.0002932476785667754,
      "loss": 1.4312,
      "step": 820
    },
    {
      "epoch": 0.12448443944506937,
      "grad_norm": 0.9494661092758179,
      "learning_rate": 0.00029302982144011514,
      "loss": 1.457,
      "step": 830
    },
    {
      "epoch": 0.12598425196850394,
      "grad_norm": 0.8792492747306824,
      "learning_rate": 0.00029280858895604727,
      "loss": 1.4365,
      "step": 840
    },
    {
      "epoch": 0.1274840644919385,
      "grad_norm": 0.9076453447341919,
      "learning_rate": 0.0002925839863354322,
      "loss": 1.4302,
      "step": 850
    },
    {
      "epoch": 0.1289838770153731,
      "grad_norm": 0.9594305157661438,
      "learning_rate": 0.00029235601887866167,
      "loss": 1.4238,
      "step": 860
    },
    {
      "epoch": 0.13048368953880765,
      "grad_norm": 0.8772804737091064,
      "learning_rate": 0.00029212469196553456,
      "loss": 1.4174,
      "step": 870
    },
    {
      "epoch": 0.13198350206224221,
      "grad_norm": 0.9967095851898193,
      "learning_rate": 0.00029189001105512914,
      "loss": 1.4379,
      "step": 880
    },
    {
      "epoch": 0.13348331458567678,
      "grad_norm": 0.9739612936973572,
      "learning_rate": 0.0002916519816856748,
      "loss": 1.4396,
      "step": 890
    },
    {
      "epoch": 0.13498312710911137,
      "grad_norm": 0.9377513527870178,
      "learning_rate": 0.000291410609474421,
      "loss": 1.4512,
      "step": 900
    },
    {
      "epoch": 0.13648293963254593,
      "grad_norm": 0.9376864433288574,
      "learning_rate": 0.0002911659001175049,
      "loss": 1.4093,
      "step": 910
    },
    {
      "epoch": 0.1379827521559805,
      "grad_norm": 0.8760865330696106,
      "learning_rate": 0.000290917859389817,
      "loss": 1.4286,
      "step": 920
    },
    {
      "epoch": 0.13948256467941508,
      "grad_norm": 1.0221220254898071,
      "learning_rate": 0.0002906664931448645,
      "loss": 1.4424,
      "step": 930
    },
    {
      "epoch": 0.14098237720284965,
      "grad_norm": 0.8309196829795837,
      "learning_rate": 0.00029041180731463357,
      "loss": 1.426,
      "step": 940
    },
    {
      "epoch": 0.1424821897262842,
      "grad_norm": 0.9327015280723572,
      "learning_rate": 0.00029015380790944916,
      "loss": 1.4279,
      "step": 950
    },
    {
      "epoch": 0.1439820022497188,
      "grad_norm": 0.8863860964775085,
      "learning_rate": 0.0002898925010178332,
      "loss": 1.4184,
      "step": 960
    },
    {
      "epoch": 0.14548181477315336,
      "grad_norm": 0.9390380382537842,
      "learning_rate": 0.00028962789280636083,
      "loss": 1.4131,
      "step": 970
    },
    {
      "epoch": 0.14698162729658792,
      "grad_norm": 0.9246792197227478,
      "learning_rate": 0.00028935998951951515,
      "loss": 1.4148,
      "step": 980
    },
    {
      "epoch": 0.1484814398200225,
      "grad_norm": 0.8679428696632385,
      "learning_rate": 0.00028908879747953955,
      "loss": 1.405,
      "step": 990
    },
    {
      "epoch": 0.14998125234345708,
      "grad_norm": 0.9332796931266785,
      "learning_rate": 0.00028881432308628855,
      "loss": 1.3973,
      "step": 1000
    },
    {
      "epoch": 0.15148106486689164,
      "grad_norm": 0.9386698007583618,
      "learning_rate": 0.00028853657281707696,
      "loss": 1.4142,
      "step": 1010
    },
    {
      "epoch": 0.1529808773903262,
      "grad_norm": 0.8631579279899597,
      "learning_rate": 0.0002882555532265269,
      "loss": 1.4148,
      "step": 1020
    },
    {
      "epoch": 0.1544806899137608,
      "grad_norm": 0.892475962638855,
      "learning_rate": 0.0002879712709464131,
      "loss": 1.4077,
      "step": 1030
    },
    {
      "epoch": 0.15598050243719536,
      "grad_norm": 0.8318502306938171,
      "learning_rate": 0.0002876837326855064,
      "loss": 1.3898,
      "step": 1040
    },
    {
      "epoch": 0.15748031496062992,
      "grad_norm": 0.8452482223510742,
      "learning_rate": 0.00028739294522941555,
      "loss": 1.3882,
      "step": 1050
    },
    {
      "epoch": 0.15898012748406448,
      "grad_norm": 0.9173560738563538,
      "learning_rate": 0.00028709891544042687,
      "loss": 1.3831,
      "step": 1060
    },
    {
      "epoch": 0.16047994000749907,
      "grad_norm": 0.86043381690979,
      "learning_rate": 0.0002868016502573425,
      "loss": 1.3982,
      "step": 1070
    },
    {
      "epoch": 0.16197975253093364,
      "grad_norm": 0.8810198903083801,
      "learning_rate": 0.00028650115669531654,
      "loss": 1.3806,
      "step": 1080
    },
    {
      "epoch": 0.1634795650543682,
      "grad_norm": 0.966907799243927,
      "learning_rate": 0.00028619744184568946,
      "loss": 1.3766,
      "step": 1090
    },
    {
      "epoch": 0.1649793775778028,
      "grad_norm": 0.8428529500961304,
      "learning_rate": 0.00028589051287582093,
      "loss": 1.3873,
      "step": 1100
    },
    {
      "epoch": 0.16647919010123735,
      "grad_norm": 0.9754992723464966,
      "learning_rate": 0.0002855803770289206,
      "loss": 1.3956,
      "step": 1110
    },
    {
      "epoch": 0.1679790026246719,
      "grad_norm": 0.9287955164909363,
      "learning_rate": 0.0002852670416238769,
      "loss": 1.3714,
      "step": 1120
    },
    {
      "epoch": 0.16947881514810648,
      "grad_norm": 0.9051069617271423,
      "learning_rate": 0.0002849505140550848,
      "loss": 1.3866,
      "step": 1130
    },
    {
      "epoch": 0.17097862767154107,
      "grad_norm": 0.8417872190475464,
      "learning_rate": 0.00028463080179227105,
      "loss": 1.3827,
      "step": 1140
    },
    {
      "epoch": 0.17247844019497563,
      "grad_norm": 1.0358750820159912,
      "learning_rate": 0.00028430791238031775,
      "loss": 1.4054,
      "step": 1150
    },
    {
      "epoch": 0.1739782527184102,
      "grad_norm": 0.8023399710655212,
      "learning_rate": 0.00028398185343908464,
      "loss": 1.3819,
      "step": 1160
    },
    {
      "epoch": 0.17547806524184478,
      "grad_norm": 0.887250542640686,
      "learning_rate": 0.000283652632663229,
      "loss": 1.3997,
      "step": 1170
    },
    {
      "epoch": 0.17697787776527935,
      "grad_norm": 0.8158445358276367,
      "learning_rate": 0.0002833202578220242,
      "loss": 1.3762,
      "step": 1180
    },
    {
      "epoch": 0.1784776902887139,
      "grad_norm": 0.799800455570221,
      "learning_rate": 0.0002829847367591764,
      "loss": 1.3974,
      "step": 1190
    },
    {
      "epoch": 0.17997750281214847,
      "grad_norm": 0.8806690573692322,
      "learning_rate": 0.0002826460773926393,
      "loss": 1.3694,
      "step": 1200
    },
    {
      "epoch": 0.18147731533558306,
      "grad_norm": 0.9353827834129333,
      "learning_rate": 0.00028230428771442725,
      "loss": 1.3646,
      "step": 1210
    },
    {
      "epoch": 0.18297712785901762,
      "grad_norm": 0.8729731440544128,
      "learning_rate": 0.000281959375790427,
      "loss": 1.3693,
      "step": 1220
    },
    {
      "epoch": 0.1844769403824522,
      "grad_norm": 0.8897218108177185,
      "learning_rate": 0.0002816113497602069,
      "loss": 1.3737,
      "step": 1230
    },
    {
      "epoch": 0.18597675290588678,
      "grad_norm": 0.8456818461418152,
      "learning_rate": 0.0002812602178368251,
      "loss": 1.3628,
      "step": 1240
    },
    {
      "epoch": 0.18747656542932134,
      "grad_norm": 0.8601028323173523,
      "learning_rate": 0.00028090598830663566,
      "loss": 1.3691,
      "step": 1250
    },
    {
      "epoch": 0.1889763779527559,
      "grad_norm": 0.9002561569213867,
      "learning_rate": 0.00028054866952909296,
      "loss": 1.3955,
      "step": 1260
    },
    {
      "epoch": 0.19047619047619047,
      "grad_norm": 0.8424673080444336,
      "learning_rate": 0.00028018826993655445,
      "loss": 1.3606,
      "step": 1270
    },
    {
      "epoch": 0.19197600299962506,
      "grad_norm": 0.8275784254074097,
      "learning_rate": 0.00027982479803408166,
      "loss": 1.3566,
      "step": 1280
    },
    {
      "epoch": 0.19347581552305962,
      "grad_norm": 0.8737898468971252,
      "learning_rate": 0.00027945826239923955,
      "loss": 1.3677,
      "step": 1290
    },
    {
      "epoch": 0.19497562804649418,
      "grad_norm": 0.901017963886261,
      "learning_rate": 0.000279088671681894,
      "loss": 1.3691,
      "step": 1300
    },
    {
      "epoch": 0.19647544056992877,
      "grad_norm": 0.9295936822891235,
      "learning_rate": 0.0002787160346040076,
      "loss": 1.3403,
      "step": 1310
    },
    {
      "epoch": 0.19797525309336333,
      "grad_norm": 0.9055348038673401,
      "learning_rate": 0.00027834035995943413,
      "loss": 1.3562,
      "step": 1320
    },
    {
      "epoch": 0.1994750656167979,
      "grad_norm": 0.8951241970062256,
      "learning_rate": 0.00027796165661371074,
      "loss": 1.3415,
      "step": 1330
    },
    {
      "epoch": 0.20007499062617173,
      "eval_loss": 1.4021942615509033,
      "eval_runtime": 34.3883,
      "eval_samples_per_second": 726.991,
      "eval_steps_per_second": 90.874,
      "step": 1334
    },
    {
      "epoch": 0.20097487814023246,
      "grad_norm": 0.8198707699775696,
      "learning_rate": 0.00027757993350384873,
      "loss": 1.3479,
      "step": 1340
    },
    {
      "epoch": 0.20247469066366705,
      "grad_norm": 0.8340823650360107,
      "learning_rate": 0.00027719519963812286,
      "loss": 1.3498,
      "step": 1350
    },
    {
      "epoch": 0.2039745031871016,
      "grad_norm": 0.8691400289535522,
      "learning_rate": 0.00027680746409585865,
      "loss": 1.3531,
      "step": 1360
    },
    {
      "epoch": 0.20547431571053618,
      "grad_norm": 0.8769707679748535,
      "learning_rate": 0.00027641673602721805,
      "loss": 1.3337,
      "step": 1370
    },
    {
      "epoch": 0.20697412823397077,
      "grad_norm": 0.8912369012832642,
      "learning_rate": 0.00027602302465298367,
      "loss": 1.3398,
      "step": 1380
    },
    {
      "epoch": 0.20847394075740533,
      "grad_norm": 0.8747676014900208,
      "learning_rate": 0.0002756263392643409,
      "loss": 1.3427,
      "step": 1390
    },
    {
      "epoch": 0.2099737532808399,
      "grad_norm": 0.8138441443443298,
      "learning_rate": 0.0002752266892226591,
      "loss": 1.3397,
      "step": 1400
    },
    {
      "epoch": 0.21147356580427445,
      "grad_norm": 0.8546535968780518,
      "learning_rate": 0.0002748240839592701,
      "loss": 1.3423,
      "step": 1410
    },
    {
      "epoch": 0.21297337832770905,
      "grad_norm": 0.907558023929596,
      "learning_rate": 0.00027441853297524615,
      "loss": 1.3753,
      "step": 1420
    },
    {
      "epoch": 0.2144731908511436,
      "grad_norm": 0.8942850232124329,
      "learning_rate": 0.00027401004584117535,
      "loss": 1.3413,
      "step": 1430
    },
    {
      "epoch": 0.21597300337457817,
      "grad_norm": 0.8546082377433777,
      "learning_rate": 0.00027359863219693614,
      "loss": 1.3349,
      "step": 1440
    },
    {
      "epoch": 0.21747281589801276,
      "grad_norm": 0.8140226006507874,
      "learning_rate": 0.00027318430175146934,
      "loss": 1.3418,
      "step": 1450
    },
    {
      "epoch": 0.21897262842144732,
      "grad_norm": 0.8695687651634216,
      "learning_rate": 0.00027276706428254965,
      "loss": 1.345,
      "step": 1460
    },
    {
      "epoch": 0.2204724409448819,
      "grad_norm": 0.9050869941711426,
      "learning_rate": 0.00027234692963655407,
      "loss": 1.3443,
      "step": 1470
    },
    {
      "epoch": 0.22197225346831645,
      "grad_norm": 0.8348211646080017,
      "learning_rate": 0.00027192390772823045,
      "loss": 1.3453,
      "step": 1480
    },
    {
      "epoch": 0.22347206599175104,
      "grad_norm": 0.9440985918045044,
      "learning_rate": 0.00027149800854046283,
      "loss": 1.3336,
      "step": 1490
    },
    {
      "epoch": 0.2249718785151856,
      "grad_norm": 0.8988614678382874,
      "learning_rate": 0.0002710692421240362,
      "loss": 1.3411,
      "step": 1500
    },
    {
      "epoch": 0.22647169103862017,
      "grad_norm": 0.8702236413955688,
      "learning_rate": 0.0002706376185973991,
      "loss": 1.3423,
      "step": 1510
    },
    {
      "epoch": 0.22797150356205476,
      "grad_norm": 0.8790213465690613,
      "learning_rate": 0.0002702031481464252,
      "loss": 1.3192,
      "step": 1520
    },
    {
      "epoch": 0.22947131608548932,
      "grad_norm": 0.8360859155654907,
      "learning_rate": 0.00026976584102417233,
      "loss": 1.3411,
      "step": 1530
    },
    {
      "epoch": 0.23097112860892388,
      "grad_norm": 0.8287367820739746,
      "learning_rate": 0.0002693257075506411,
      "loss": 1.3423,
      "step": 1540
    },
    {
      "epoch": 0.23247094113235844,
      "grad_norm": 0.8342249989509583,
      "learning_rate": 0.00026888275811253105,
      "loss": 1.3485,
      "step": 1550
    },
    {
      "epoch": 0.23397075365579303,
      "grad_norm": 0.8485095500946045,
      "learning_rate": 0.00026843700316299564,
      "loss": 1.328,
      "step": 1560
    },
    {
      "epoch": 0.2354705661792276,
      "grad_norm": 0.8335973620414734,
      "learning_rate": 0.0002679884532213954,
      "loss": 1.3002,
      "step": 1570
    },
    {
      "epoch": 0.23697037870266216,
      "grad_norm": 0.8893833160400391,
      "learning_rate": 0.00026753711887304995,
      "loss": 1.3364,
      "step": 1580
    },
    {
      "epoch": 0.23847019122609675,
      "grad_norm": 0.9313494563102722,
      "learning_rate": 0.000267083010768988,
      "loss": 1.3419,
      "step": 1590
    },
    {
      "epoch": 0.2399700037495313,
      "grad_norm": 0.7982856631278992,
      "learning_rate": 0.0002666261396256961,
      "loss": 1.3219,
      "step": 1600
    },
    {
      "epoch": 0.24146981627296588,
      "grad_norm": 0.8474456071853638,
      "learning_rate": 0.0002661665162248656,
      "loss": 1.329,
      "step": 1610
    },
    {
      "epoch": 0.24296962879640044,
      "grad_norm": 0.8094434142112732,
      "learning_rate": 0.0002657041514131385,
      "loss": 1.3344,
      "step": 1620
    },
    {
      "epoch": 0.24446944131983503,
      "grad_norm": 0.8186250925064087,
      "learning_rate": 0.000265239056101851,
      "loss": 1.3266,
      "step": 1630
    },
    {
      "epoch": 0.2459692538432696,
      "grad_norm": 0.8790938854217529,
      "learning_rate": 0.0002647712412667765,
      "loss": 1.3141,
      "step": 1640
    },
    {
      "epoch": 0.24746906636670415,
      "grad_norm": 0.8377759456634521,
      "learning_rate": 0.00026430071794786644,
      "loss": 1.3285,
      "step": 1650
    },
    {
      "epoch": 0.24896887889013875,
      "grad_norm": 0.9131957292556763,
      "learning_rate": 0.00026382749724898955,
      "loss": 1.3029,
      "step": 1660
    },
    {
      "epoch": 0.2504686914135733,
      "grad_norm": 0.8385202884674072,
      "learning_rate": 0.00026335159033766996,
      "loss": 1.3329,
      "step": 1670
    },
    {
      "epoch": 0.25196850393700787,
      "grad_norm": 0.7972739338874817,
      "learning_rate": 0.0002628730084448239,
      "loss": 1.3253,
      "step": 1680
    },
    {
      "epoch": 0.25346831646044243,
      "grad_norm": 0.8282127380371094,
      "learning_rate": 0.000262391762864494,
      "loss": 1.3327,
      "step": 1690
    },
    {
      "epoch": 0.254968128983877,
      "grad_norm": 0.8004878163337708,
      "learning_rate": 0.00026190786495358366,
      "loss": 1.3186,
      "step": 1700
    },
    {
      "epoch": 0.25646794150731156,
      "grad_norm": 0.825681746006012,
      "learning_rate": 0.0002614213261315883,
      "loss": 1.3129,
      "step": 1710
    },
    {
      "epoch": 0.2579677540307462,
      "grad_norm": 0.8196373581886292,
      "learning_rate": 0.0002609321578803261,
      "loss": 1.3185,
      "step": 1720
    },
    {
      "epoch": 0.25946756655418074,
      "grad_norm": 0.8522502779960632,
      "learning_rate": 0.00026044037174366734,
      "loss": 1.3107,
      "step": 1730
    },
    {
      "epoch": 0.2609673790776153,
      "grad_norm": 0.8095912933349609,
      "learning_rate": 0.00025994597932726135,
      "loss": 1.3218,
      "step": 1740
    },
    {
      "epoch": 0.26246719160104987,
      "grad_norm": 0.8493536710739136,
      "learning_rate": 0.0002594489922982633,
      "loss": 1.3234,
      "step": 1750
    },
    {
      "epoch": 0.26396700412448443,
      "grad_norm": 0.8144869208335876,
      "learning_rate": 0.0002589494223850584,
      "loss": 1.3,
      "step": 1760
    },
    {
      "epoch": 0.265466816647919,
      "grad_norm": 0.8083682060241699,
      "learning_rate": 0.00025844728137698543,
      "loss": 1.3283,
      "step": 1770
    },
    {
      "epoch": 0.26696662917135355,
      "grad_norm": 0.8459505438804626,
      "learning_rate": 0.0002579425811240582,
      "loss": 1.319,
      "step": 1780
    },
    {
      "epoch": 0.26846644169478817,
      "grad_norm": 0.8262299299240112,
      "learning_rate": 0.00025743533353668626,
      "loss": 1.3089,
      "step": 1790
    },
    {
      "epoch": 0.26996625421822273,
      "grad_norm": 0.8823822736740112,
      "learning_rate": 0.0002569255505853934,
      "loss": 1.3157,
      "step": 1800
    },
    {
      "epoch": 0.2714660667416573,
      "grad_norm": 0.8654087781906128,
      "learning_rate": 0.0002564132443005356,
      "loss": 1.3113,
      "step": 1810
    },
    {
      "epoch": 0.27296587926509186,
      "grad_norm": 0.8660631775856018,
      "learning_rate": 0.00025589842677201693,
      "loss": 1.303,
      "step": 1820
    },
    {
      "epoch": 0.2744656917885264,
      "grad_norm": 0.8453037142753601,
      "learning_rate": 0.0002553811101490042,
      "loss": 1.3036,
      "step": 1830
    },
    {
      "epoch": 0.275965504311961,
      "grad_norm": 0.7789093255996704,
      "learning_rate": 0.00025486130663964016,
      "loss": 1.3064,
      "step": 1840
    },
    {
      "epoch": 0.27746531683539555,
      "grad_norm": 0.8085753321647644,
      "learning_rate": 0.00025433902851075584,
      "loss": 1.3135,
      "step": 1850
    },
    {
      "epoch": 0.27896512935883017,
      "grad_norm": 0.8380696773529053,
      "learning_rate": 0.0002538142880875805,
      "loss": 1.2949,
      "step": 1860
    },
    {
      "epoch": 0.28046494188226473,
      "grad_norm": 0.8569440245628357,
      "learning_rate": 0.00025328709775345105,
      "loss": 1.3174,
      "step": 1870
    },
    {
      "epoch": 0.2819647544056993,
      "grad_norm": 0.7963806390762329,
      "learning_rate": 0.0002527574699495199,
      "loss": 1.3079,
      "step": 1880
    },
    {
      "epoch": 0.28346456692913385,
      "grad_norm": 0.8502215147018433,
      "learning_rate": 0.00025222541717446117,
      "loss": 1.3019,
      "step": 1890
    },
    {
      "epoch": 0.2849643794525684,
      "grad_norm": 0.8192076086997986,
      "learning_rate": 0.00025169095198417584,
      "loss": 1.2963,
      "step": 1900
    },
    {
      "epoch": 0.286464191976003,
      "grad_norm": 0.8132408261299133,
      "learning_rate": 0.00025115408699149546,
      "loss": 1.3122,
      "step": 1910
    },
    {
      "epoch": 0.2879640044994376,
      "grad_norm": 0.9454010128974915,
      "learning_rate": 0.00025061483486588435,
      "loss": 1.3203,
      "step": 1920
    },
    {
      "epoch": 0.28946381702287216,
      "grad_norm": 0.7738835215568542,
      "learning_rate": 0.00025007320833314085,
      "loss": 1.2868,
      "step": 1930
    },
    {
      "epoch": 0.2909636295463067,
      "grad_norm": 0.7828739881515503,
      "learning_rate": 0.00024952922017509687,
      "loss": 1.3065,
      "step": 1940
    },
    {
      "epoch": 0.2924634420697413,
      "grad_norm": 0.8362312316894531,
      "learning_rate": 0.00024898288322931615,
      "loss": 1.2927,
      "step": 1950
    },
    {
      "epoch": 0.29396325459317585,
      "grad_norm": 0.80490642786026,
      "learning_rate": 0.00024843421038879147,
      "loss": 1.2976,
      "step": 1960
    },
    {
      "epoch": 0.2954630671166104,
      "grad_norm": 0.8396286368370056,
      "learning_rate": 0.0002478832146016404,
      "loss": 1.3003,
      "step": 1970
    },
    {
      "epoch": 0.296962879640045,
      "grad_norm": 0.8507101535797119,
      "learning_rate": 0.0002473299088707996,
      "loss": 1.298,
      "step": 1980
    },
    {
      "epoch": 0.2984626921634796,
      "grad_norm": 0.886080801486969,
      "learning_rate": 0.00024677430625371803,
      "loss": 1.3002,
      "step": 1990
    },
    {
      "epoch": 0.29996250468691416,
      "grad_norm": 0.8753901124000549,
      "learning_rate": 0.0002462164198620489,
      "loss": 1.3057,
      "step": 2000
    },
    {
      "epoch": 0.3001124859392576,
      "eval_loss": 1.337799072265625,
      "eval_runtime": 34.4016,
      "eval_samples_per_second": 726.711,
      "eval_steps_per_second": 90.839,
      "step": 2001
    },
    {
      "epoch": 0.3014623172103487,
      "grad_norm": 0.8367862701416016,
      "learning_rate": 0.00024565626286134003,
      "loss": 1.2853,
      "step": 2010
    },
    {
      "epoch": 0.3029621297337833,
      "grad_norm": 0.8323600888252258,
      "learning_rate": 0.0002450938484707234,
      "loss": 1.2831,
      "step": 2020
    },
    {
      "epoch": 0.30446194225721784,
      "grad_norm": 0.8133791089057922,
      "learning_rate": 0.0002445291899626031,
      "loss": 1.2837,
      "step": 2030
    },
    {
      "epoch": 0.3059617547806524,
      "grad_norm": 0.8445649743080139,
      "learning_rate": 0.000243962300662342,
      "loss": 1.3076,
      "step": 2040
    },
    {
      "epoch": 0.30746156730408697,
      "grad_norm": 0.8368006944656372,
      "learning_rate": 0.00024339319394794742,
      "loss": 1.3018,
      "step": 2050
    },
    {
      "epoch": 0.3089613798275216,
      "grad_norm": 0.8313194513320923,
      "learning_rate": 0.00024282188324975534,
      "loss": 1.2931,
      "step": 2060
    },
    {
      "epoch": 0.31046119235095615,
      "grad_norm": 0.8213202357292175,
      "learning_rate": 0.0002422483820501136,
      "loss": 1.2962,
      "step": 2070
    },
    {
      "epoch": 0.3119610048743907,
      "grad_norm": 0.8784091472625732,
      "learning_rate": 0.00024167270388306366,
      "loss": 1.288,
      "step": 2080
    },
    {
      "epoch": 0.3134608173978253,
      "grad_norm": 0.7992005348205566,
      "learning_rate": 0.00024109486233402102,
      "loss": 1.29,
      "step": 2090
    },
    {
      "epoch": 0.31496062992125984,
      "grad_norm": 0.7810459136962891,
      "learning_rate": 0.00024051487103945486,
      "loss": 1.2769,
      "step": 2100
    },
    {
      "epoch": 0.3164604424446944,
      "grad_norm": 0.7913289666175842,
      "learning_rate": 0.00023993274368656618,
      "loss": 1.2822,
      "step": 2110
    },
    {
      "epoch": 0.31796025496812896,
      "grad_norm": 0.7812384366989136,
      "learning_rate": 0.00023934849401296472,
      "loss": 1.2962,
      "step": 2120
    },
    {
      "epoch": 0.3194600674915636,
      "grad_norm": 0.8585249185562134,
      "learning_rate": 0.0002387621358063449,
      "loss": 1.2842,
      "step": 2130
    },
    {
      "epoch": 0.32095988001499814,
      "grad_norm": 0.8342397212982178,
      "learning_rate": 0.00023817368290416036,
      "loss": 1.287,
      "step": 2140
    },
    {
      "epoch": 0.3224596925384327,
      "grad_norm": 0.7857896089553833,
      "learning_rate": 0.00023758314919329726,
      "loss": 1.3053,
      "step": 2150
    },
    {
      "epoch": 0.32395950506186727,
      "grad_norm": 0.7923012971878052,
      "learning_rate": 0.00023699054860974682,
      "loss": 1.2731,
      "step": 2160
    },
    {
      "epoch": 0.32545931758530183,
      "grad_norm": 0.7796991467475891,
      "learning_rate": 0.00023639589513827636,
      "loss": 1.2716,
      "step": 2170
    },
    {
      "epoch": 0.3269591301087364,
      "grad_norm": 0.8867438435554504,
      "learning_rate": 0.0002357992028120993,
      "loss": 1.2908,
      "step": 2180
    },
    {
      "epoch": 0.32845894263217096,
      "grad_norm": 0.8547908067703247,
      "learning_rate": 0.00023520048571254378,
      "loss": 1.2772,
      "step": 2190
    },
    {
      "epoch": 0.3299587551556056,
      "grad_norm": 0.8480901122093201,
      "learning_rate": 0.00023459975796872063,
      "loss": 1.2716,
      "step": 2200
    },
    {
      "epoch": 0.33145856767904014,
      "grad_norm": 0.8187602758407593,
      "learning_rate": 0.0002339970337571899,
      "loss": 1.2724,
      "step": 2210
    },
    {
      "epoch": 0.3329583802024747,
      "grad_norm": 0.8206058740615845,
      "learning_rate": 0.000233392327301626,
      "loss": 1.3034,
      "step": 2220
    },
    {
      "epoch": 0.33445819272590926,
      "grad_norm": 0.76264888048172,
      "learning_rate": 0.0002327856528724825,
      "loss": 1.2576,
      "step": 2230
    },
    {
      "epoch": 0.3359580052493438,
      "grad_norm": 0.8335399627685547,
      "learning_rate": 0.0002321770247866551,
      "loss": 1.2857,
      "step": 2240
    },
    {
      "epoch": 0.3374578177727784,
      "grad_norm": 0.7656426429748535,
      "learning_rate": 0.00023156645740714368,
      "loss": 1.2978,
      "step": 2250
    },
    {
      "epoch": 0.33895763029621295,
      "grad_norm": 0.7743305563926697,
      "learning_rate": 0.00023095396514271355,
      "loss": 1.2803,
      "step": 2260
    },
    {
      "epoch": 0.34045744281964757,
      "grad_norm": 0.7768455147743225,
      "learning_rate": 0.0002303395624475553,
      "loss": 1.2978,
      "step": 2270
    },
    {
      "epoch": 0.34195725534308213,
      "grad_norm": 0.8204723596572876,
      "learning_rate": 0.00022972326382094378,
      "loss": 1.2708,
      "step": 2280
    },
    {
      "epoch": 0.3434570678665167,
      "grad_norm": 0.8377450108528137,
      "learning_rate": 0.00022910508380689584,
      "loss": 1.276,
      "step": 2290
    },
    {
      "epoch": 0.34495688038995126,
      "grad_norm": 0.7735800743103027,
      "learning_rate": 0.00022848503699382717,
      "loss": 1.2987,
      "step": 2300
    },
    {
      "epoch": 0.3464566929133858,
      "grad_norm": 0.8727670907974243,
      "learning_rate": 0.00022786313801420794,
      "loss": 1.267,
      "step": 2310
    },
    {
      "epoch": 0.3479565054368204,
      "grad_norm": 0.7944260835647583,
      "learning_rate": 0.0002272394015442177,
      "loss": 1.2937,
      "step": 2320
    },
    {
      "epoch": 0.34945631796025495,
      "grad_norm": 0.8214771747589111,
      "learning_rate": 0.0002266138423033987,
      "loss": 1.2879,
      "step": 2330
    },
    {
      "epoch": 0.35095613048368957,
      "grad_norm": 0.7794116139411926,
      "learning_rate": 0.00022598647505430895,
      "loss": 1.2599,
      "step": 2340
    },
    {
      "epoch": 0.35245594300712413,
      "grad_norm": 0.76594078540802,
      "learning_rate": 0.0002253573146021733,
      "loss": 1.2613,
      "step": 2350
    },
    {
      "epoch": 0.3539557555305587,
      "grad_norm": 0.8062904477119446,
      "learning_rate": 0.0002247263757945347,
      "loss": 1.2959,
      "step": 2360
    },
    {
      "epoch": 0.35545556805399325,
      "grad_norm": 0.8257420063018799,
      "learning_rate": 0.00022409367352090322,
      "loss": 1.2567,
      "step": 2370
    },
    {
      "epoch": 0.3569553805774278,
      "grad_norm": 0.8322898149490356,
      "learning_rate": 0.00022345922271240496,
      "loss": 1.2684,
      "step": 2380
    },
    {
      "epoch": 0.3584551931008624,
      "grad_norm": 0.8116471171379089,
      "learning_rate": 0.00022282303834142978,
      "loss": 1.2643,
      "step": 2390
    },
    {
      "epoch": 0.35995500562429694,
      "grad_norm": 0.8192791938781738,
      "learning_rate": 0.0002221851354212777,
      "loss": 1.2586,
      "step": 2400
    },
    {
      "epoch": 0.36145481814773156,
      "grad_norm": 0.7919474244117737,
      "learning_rate": 0.0002215455290058048,
      "loss": 1.2869,
      "step": 2410
    },
    {
      "epoch": 0.3629546306711661,
      "grad_norm": 0.8426802158355713,
      "learning_rate": 0.000220904234189068,
      "loss": 1.2589,
      "step": 2420
    },
    {
      "epoch": 0.3644544431946007,
      "grad_norm": 0.851420521736145,
      "learning_rate": 0.00022026126610496852,
      "loss": 1.2569,
      "step": 2430
    },
    {
      "epoch": 0.36595425571803525,
      "grad_norm": 0.8213547468185425,
      "learning_rate": 0.0002196166399268952,
      "loss": 1.2698,
      "step": 2440
    },
    {
      "epoch": 0.3674540682414698,
      "grad_norm": 0.7695969343185425,
      "learning_rate": 0.00021897037086736614,
      "loss": 1.2668,
      "step": 2450
    },
    {
      "epoch": 0.3689538807649044,
      "grad_norm": 0.7834669351577759,
      "learning_rate": 0.0002183224741776697,
      "loss": 1.2662,
      "step": 2460
    },
    {
      "epoch": 0.37045369328833894,
      "grad_norm": 0.7951564788818359,
      "learning_rate": 0.00021767296514750472,
      "loss": 1.2661,
      "step": 2470
    },
    {
      "epoch": 0.37195350581177355,
      "grad_norm": 0.7705678939819336,
      "learning_rate": 0.00021702185910461958,
      "loss": 1.2623,
      "step": 2480
    },
    {
      "epoch": 0.3734533183352081,
      "grad_norm": 0.850374162197113,
      "learning_rate": 0.00021636917141445056,
      "loss": 1.2386,
      "step": 2490
    },
    {
      "epoch": 0.3749531308586427,
      "grad_norm": 0.795702338218689,
      "learning_rate": 0.00021571491747975917,
      "loss": 1.2604,
      "step": 2500
    },
    {
      "epoch": 0.37645294338207724,
      "grad_norm": 0.8536216020584106,
      "learning_rate": 0.0002150591127402687,
      "loss": 1.2497,
      "step": 2510
    },
    {
      "epoch": 0.3779527559055118,
      "grad_norm": 0.813890278339386,
      "learning_rate": 0.00021440177267229984,
      "loss": 1.2505,
      "step": 2520
    },
    {
      "epoch": 0.37945256842894637,
      "grad_norm": 0.8229677081108093,
      "learning_rate": 0.00021374291278840546,
      "loss": 1.2634,
      "step": 2530
    },
    {
      "epoch": 0.38095238095238093,
      "grad_norm": 0.868016242980957,
      "learning_rate": 0.00021308254863700452,
      "loss": 1.2537,
      "step": 2540
    },
    {
      "epoch": 0.38245219347581555,
      "grad_norm": 0.8083469271659851,
      "learning_rate": 0.00021242069580201524,
      "loss": 1.2702,
      "step": 2550
    },
    {
      "epoch": 0.3839520059992501,
      "grad_norm": 0.7700805068016052,
      "learning_rate": 0.00021175736990248714,
      "loss": 1.2755,
      "step": 2560
    },
    {
      "epoch": 0.3854518185226847,
      "grad_norm": 0.787372350692749,
      "learning_rate": 0.00021109258659223254,
      "loss": 1.2581,
      "step": 2570
    },
    {
      "epoch": 0.38695163104611924,
      "grad_norm": 0.8132008910179138,
      "learning_rate": 0.00021042636155945723,
      "loss": 1.2408,
      "step": 2580
    },
    {
      "epoch": 0.3884514435695538,
      "grad_norm": 0.7871599793434143,
      "learning_rate": 0.00020975871052639024,
      "loss": 1.2622,
      "step": 2590
    },
    {
      "epoch": 0.38995125609298836,
      "grad_norm": 0.841528594493866,
      "learning_rate": 0.00020908964924891256,
      "loss": 1.2382,
      "step": 2600
    },
    {
      "epoch": 0.3914510686164229,
      "grad_norm": 0.7999451756477356,
      "learning_rate": 0.0002084191935161857,
      "loss": 1.2771,
      "step": 2610
    },
    {
      "epoch": 0.39295088113985754,
      "grad_norm": 0.7474108934402466,
      "learning_rate": 0.0002077473591502788,
      "loss": 1.2656,
      "step": 2620
    },
    {
      "epoch": 0.3944506936632921,
      "grad_norm": 0.8236092329025269,
      "learning_rate": 0.00020707416200579524,
      "loss": 1.2576,
      "step": 2630
    },
    {
      "epoch": 0.39595050618672667,
      "grad_norm": 0.8108281493186951,
      "learning_rate": 0.00020639961796949877,
      "loss": 1.2534,
      "step": 2640
    },
    {
      "epoch": 0.39745031871016123,
      "grad_norm": 0.8653910756111145,
      "learning_rate": 0.00020572374295993822,
      "loss": 1.2666,
      "step": 2650
    },
    {
      "epoch": 0.3989501312335958,
      "grad_norm": 0.8193902969360352,
      "learning_rate": 0.00020504655292707223,
      "loss": 1.2528,
      "step": 2660
    },
    {
      "epoch": 0.40014998125234347,
      "eval_loss": 1.300592303276062,
      "eval_runtime": 34.4278,
      "eval_samples_per_second": 726.158,
      "eval_steps_per_second": 90.77,
      "step": 2668
    },
    {
      "epoch": 0.40044994375703036,
      "grad_norm": 0.8101006150245667,
      "learning_rate": 0.00020436806385189246,
      "loss": 1.2646,
      "step": 2670
    },
    {
      "epoch": 0.4019497562804649,
      "grad_norm": 0.8838757276535034,
      "learning_rate": 0.00020368829174604667,
      "loss": 1.2667,
      "step": 2680
    },
    {
      "epoch": 0.40344956880389954,
      "grad_norm": 0.7894991636276245,
      "learning_rate": 0.00020300725265146093,
      "loss": 1.266,
      "step": 2690
    },
    {
      "epoch": 0.4049493813273341,
      "grad_norm": 0.8235330581665039,
      "learning_rate": 0.00020232496263996092,
      "loss": 1.2503,
      "step": 2700
    },
    {
      "epoch": 0.40644919385076866,
      "grad_norm": 0.7787916660308838,
      "learning_rate": 0.00020164143781289256,
      "loss": 1.2521,
      "step": 2710
    },
    {
      "epoch": 0.4079490063742032,
      "grad_norm": 0.8202505707740784,
      "learning_rate": 0.00020095669430074235,
      "loss": 1.2717,
      "step": 2720
    },
    {
      "epoch": 0.4094488188976378,
      "grad_norm": 0.7967453598976135,
      "learning_rate": 0.00020027074826275629,
      "loss": 1.2636,
      "step": 2730
    },
    {
      "epoch": 0.41094863142107235,
      "grad_norm": 0.7384628653526306,
      "learning_rate": 0.00019958361588655888,
      "loss": 1.2531,
      "step": 2740
    },
    {
      "epoch": 0.4124484439445069,
      "grad_norm": 0.8367340564727783,
      "learning_rate": 0.00019889531338777112,
      "loss": 1.262,
      "step": 2750
    },
    {
      "epoch": 0.41394825646794153,
      "grad_norm": 0.8179819583892822,
      "learning_rate": 0.0001982058570096274,
      "loss": 1.2453,
      "step": 2760
    },
    {
      "epoch": 0.4154480689913761,
      "grad_norm": 0.7929909825325012,
      "learning_rate": 0.00019751526302259271,
      "loss": 1.2332,
      "step": 2770
    },
    {
      "epoch": 0.41694788151481066,
      "grad_norm": 0.803432285785675,
      "learning_rate": 0.00019682354772397842,
      "loss": 1.2376,
      "step": 2780
    },
    {
      "epoch": 0.4184476940382452,
      "grad_norm": 0.7572347521781921,
      "learning_rate": 0.00019613072743755755,
      "loss": 1.2535,
      "step": 2790
    },
    {
      "epoch": 0.4199475065616798,
      "grad_norm": 0.7730807662010193,
      "learning_rate": 0.00019543681851317998,
      "loss": 1.2457,
      "step": 2800
    },
    {
      "epoch": 0.42144731908511435,
      "grad_norm": 0.8271188735961914,
      "learning_rate": 0.00019474183732638608,
      "loss": 1.2525,
      "step": 2810
    },
    {
      "epoch": 0.4229471316085489,
      "grad_norm": 0.8117753863334656,
      "learning_rate": 0.0001940458002780206,
      "loss": 1.2321,
      "step": 2820
    },
    {
      "epoch": 0.42444694413198353,
      "grad_norm": 0.8418020606040955,
      "learning_rate": 0.00019334872379384556,
      "loss": 1.2501,
      "step": 2830
    },
    {
      "epoch": 0.4259467566554181,
      "grad_norm": 0.8009008765220642,
      "learning_rate": 0.0001926506243241526,
      "loss": 1.2398,
      "step": 2840
    },
    {
      "epoch": 0.42744656917885265,
      "grad_norm": 0.7942110300064087,
      "learning_rate": 0.00019195151834337473,
      "loss": 1.244,
      "step": 2850
    },
    {
      "epoch": 0.4289463817022872,
      "grad_norm": 0.8147196173667908,
      "learning_rate": 0.00019125142234969762,
      "loss": 1.2499,
      "step": 2860
    },
    {
      "epoch": 0.4304461942257218,
      "grad_norm": 0.843659520149231,
      "learning_rate": 0.00019055035286467034,
      "loss": 1.2424,
      "step": 2870
    },
    {
      "epoch": 0.43194600674915634,
      "grad_norm": 0.8289938569068909,
      "learning_rate": 0.00018984832643281513,
      "loss": 1.2529,
      "step": 2880
    },
    {
      "epoch": 0.4334458192725909,
      "grad_norm": 0.8647093772888184,
      "learning_rate": 0.00018914535962123735,
      "loss": 1.2579,
      "step": 2890
    },
    {
      "epoch": 0.4349456317960255,
      "grad_norm": 0.7539538145065308,
      "learning_rate": 0.00018844146901923436,
      "loss": 1.2408,
      "step": 2900
    },
    {
      "epoch": 0.4364454443194601,
      "grad_norm": 0.8231214284896851,
      "learning_rate": 0.000187736671237904,
      "loss": 1.2346,
      "step": 2910
    },
    {
      "epoch": 0.43794525684289465,
      "grad_norm": 0.7767258882522583,
      "learning_rate": 0.0001870309829097526,
      "loss": 1.2379,
      "step": 2920
    },
    {
      "epoch": 0.4394450693663292,
      "grad_norm": 0.8100400567054749,
      "learning_rate": 0.00018632442068830244,
      "loss": 1.2248,
      "step": 2930
    },
    {
      "epoch": 0.4409448818897638,
      "grad_norm": 0.7970197796821594,
      "learning_rate": 0.00018561700124769892,
      "loss": 1.2312,
      "step": 2940
    },
    {
      "epoch": 0.44244469441319834,
      "grad_norm": 0.8453084826469421,
      "learning_rate": 0.0001849087412823168,
      "loss": 1.2379,
      "step": 2950
    },
    {
      "epoch": 0.4439445069366329,
      "grad_norm": 0.7835219502449036,
      "learning_rate": 0.00018419965750636645,
      "loss": 1.2377,
      "step": 2960
    },
    {
      "epoch": 0.4454443194600675,
      "grad_norm": 0.9176828861236572,
      "learning_rate": 0.00018348976665349932,
      "loss": 1.2322,
      "step": 2970
    },
    {
      "epoch": 0.4469441319835021,
      "grad_norm": 0.8404021859169006,
      "learning_rate": 0.00018277908547641294,
      "loss": 1.2364,
      "step": 2980
    },
    {
      "epoch": 0.44844394450693664,
      "grad_norm": 0.8273976445198059,
      "learning_rate": 0.00018206763074645588,
      "loss": 1.2242,
      "step": 2990
    },
    {
      "epoch": 0.4499437570303712,
      "grad_norm": 0.7463936805725098,
      "learning_rate": 0.0001813554192532316,
      "loss": 1.2459,
      "step": 3000
    },
    {
      "epoch": 0.45144356955380577,
      "grad_norm": 0.7790806293487549,
      "learning_rate": 0.00018064246780420245,
      "loss": 1.2473,
      "step": 3010
    },
    {
      "epoch": 0.45294338207724033,
      "grad_norm": 0.766952633857727,
      "learning_rate": 0.000179928793224293,
      "loss": 1.219,
      "step": 3020
    },
    {
      "epoch": 0.4544431946006749,
      "grad_norm": 0.8558096289634705,
      "learning_rate": 0.00017921441235549295,
      "loss": 1.2413,
      "step": 3030
    },
    {
      "epoch": 0.4559430071241095,
      "grad_norm": 0.8494218587875366,
      "learning_rate": 0.00017849934205645967,
      "loss": 1.2492,
      "step": 3040
    },
    {
      "epoch": 0.4574428196475441,
      "grad_norm": 0.8279352784156799,
      "learning_rate": 0.00017778359920212047,
      "loss": 1.2509,
      "step": 3050
    },
    {
      "epoch": 0.45894263217097864,
      "grad_norm": 0.7695614695549011,
      "learning_rate": 0.0001770672006832741,
      "loss": 1.2375,
      "step": 3060
    },
    {
      "epoch": 0.4604424446944132,
      "grad_norm": 0.8258097767829895,
      "learning_rate": 0.00017635016340619255,
      "loss": 1.2286,
      "step": 3070
    },
    {
      "epoch": 0.46194225721784776,
      "grad_norm": 0.7904211282730103,
      "learning_rate": 0.00017563250429222173,
      "loss": 1.2527,
      "step": 3080
    },
    {
      "epoch": 0.4634420697412823,
      "grad_norm": 0.8494943976402283,
      "learning_rate": 0.00017491424027738216,
      "loss": 1.2484,
      "step": 3090
    },
    {
      "epoch": 0.4649418822647169,
      "grad_norm": 0.8655872941017151,
      "learning_rate": 0.0001741953883119696,
      "loss": 1.2172,
      "step": 3100
    },
    {
      "epoch": 0.4664416947881515,
      "grad_norm": 0.8473703861236572,
      "learning_rate": 0.00017347596536015472,
      "loss": 1.2376,
      "step": 3110
    },
    {
      "epoch": 0.46794150731158607,
      "grad_norm": 0.8274358510971069,
      "learning_rate": 0.00017275598839958296,
      "loss": 1.2458,
      "step": 3120
    },
    {
      "epoch": 0.46944131983502063,
      "grad_norm": 0.8414213061332703,
      "learning_rate": 0.00017203547442097369,
      "loss": 1.233,
      "step": 3130
    },
    {
      "epoch": 0.4709411323584552,
      "grad_norm": 0.8282588720321655,
      "learning_rate": 0.0001713144404277195,
      "loss": 1.2398,
      "step": 3140
    },
    {
      "epoch": 0.47244094488188976,
      "grad_norm": 0.8157975077629089,
      "learning_rate": 0.0001705929034354846,
      "loss": 1.2236,
      "step": 3150
    },
    {
      "epoch": 0.4739407574053243,
      "grad_norm": 0.8301715850830078,
      "learning_rate": 0.0001698708804718037,
      "loss": 1.2214,
      "step": 3160
    },
    {
      "epoch": 0.4754405699287589,
      "grad_norm": 0.7506479620933533,
      "learning_rate": 0.00016914838857567979,
      "loss": 1.2332,
      "step": 3170
    },
    {
      "epoch": 0.4769403824521935,
      "grad_norm": 0.8075399994850159,
      "learning_rate": 0.00016842544479718215,
      "loss": 1.2344,
      "step": 3180
    },
    {
      "epoch": 0.47844019497562806,
      "grad_norm": 0.9165489673614502,
      "learning_rate": 0.00016770206619704412,
      "loss": 1.2393,
      "step": 3190
    },
    {
      "epoch": 0.4799400074990626,
      "grad_norm": 0.7626012563705444,
      "learning_rate": 0.0001669782698462603,
      "loss": 1.2273,
      "step": 3200
    },
    {
      "epoch": 0.4814398200224972,
      "grad_norm": 0.8293296694755554,
      "learning_rate": 0.00016625407282568394,
      "loss": 1.2378,
      "step": 3210
    },
    {
      "epoch": 0.48293963254593175,
      "grad_norm": 0.7572868466377258,
      "learning_rate": 0.00016552949222562352,
      "loss": 1.2449,
      "step": 3220
    },
    {
      "epoch": 0.4844394450693663,
      "grad_norm": 0.7891345024108887,
      "learning_rate": 0.00016480454514543962,
      "loss": 1.2336,
      "step": 3230
    },
    {
      "epoch": 0.4859392575928009,
      "grad_norm": 0.7002502083778381,
      "learning_rate": 0.00016407924869314144,
      "loss": 1.2249,
      "step": 3240
    },
    {
      "epoch": 0.4874390701162355,
      "grad_norm": 0.811424732208252,
      "learning_rate": 0.00016335361998498296,
      "loss": 1.2053,
      "step": 3250
    },
    {
      "epoch": 0.48893888263967006,
      "grad_norm": 0.8211063146591187,
      "learning_rate": 0.00016262767614505912,
      "loss": 1.2139,
      "step": 3260
    },
    {
      "epoch": 0.4904386951631046,
      "grad_norm": 0.8250209093093872,
      "learning_rate": 0.00016190143430490152,
      "loss": 1.2163,
      "step": 3270
    },
    {
      "epoch": 0.4919385076865392,
      "grad_norm": 0.7996110320091248,
      "learning_rate": 0.00016117491160307445,
      "loss": 1.2361,
      "step": 3280
    },
    {
      "epoch": 0.49343832020997375,
      "grad_norm": 0.8566946983337402,
      "learning_rate": 0.00016044812518477007,
      "loss": 1.2353,
      "step": 3290
    },
    {
      "epoch": 0.4949381327334083,
      "grad_norm": 0.7891775369644165,
      "learning_rate": 0.00015972109220140402,
      "loss": 1.2174,
      "step": 3300
    },
    {
      "epoch": 0.49643794525684287,
      "grad_norm": 0.769114077091217,
      "learning_rate": 0.0001589938298102108,
      "loss": 1.2293,
      "step": 3310
    },
    {
      "epoch": 0.4979377577802775,
      "grad_norm": 0.8417636156082153,
      "learning_rate": 0.0001582663551738384,
      "loss": 1.2303,
      "step": 3320
    },
    {
      "epoch": 0.49943757030371205,
      "grad_norm": 0.7648106813430786,
      "learning_rate": 0.00015753868545994378,
      "loss": 1.2287,
      "step": 3330
    },
    {
      "epoch": 0.5001874765654293,
      "eval_loss": 1.2699334621429443,
      "eval_runtime": 34.8375,
      "eval_samples_per_second": 717.618,
      "eval_steps_per_second": 89.702,
      "step": 3335
    },
    {
      "epoch": 0.5009373828271466,
      "grad_norm": 0.8185369372367859,
      "learning_rate": 0.00015681083784078748,
      "loss": 1.2221,
      "step": 3340
    },
    {
      "epoch": 0.5024371953505812,
      "grad_norm": 0.8825401067733765,
      "learning_rate": 0.00015608282949282844,
      "loss": 1.2339,
      "step": 3350
    },
    {
      "epoch": 0.5039370078740157,
      "grad_norm": 0.8161605596542358,
      "learning_rate": 0.00015535467759631862,
      "loss": 1.2352,
      "step": 3360
    },
    {
      "epoch": 0.5054368203974503,
      "grad_norm": 0.8088217973709106,
      "learning_rate": 0.00015462639933489753,
      "loss": 1.2212,
      "step": 3370
    },
    {
      "epoch": 0.5069366329208849,
      "grad_norm": 0.8628859519958496,
      "learning_rate": 0.00015389801189518693,
      "loss": 1.2222,
      "step": 3380
    },
    {
      "epoch": 0.5084364454443194,
      "grad_norm": 0.8813005089759827,
      "learning_rate": 0.00015316953246638482,
      "loss": 1.2165,
      "step": 3390
    },
    {
      "epoch": 0.509936257967754,
      "grad_norm": 0.7920242547988892,
      "learning_rate": 0.00015244097823986023,
      "loss": 1.2197,
      "step": 3400
    },
    {
      "epoch": 0.5114360704911886,
      "grad_norm": 0.8255246877670288,
      "learning_rate": 0.0001517123664087473,
      "loss": 1.2359,
      "step": 3410
    },
    {
      "epoch": 0.5129358830146231,
      "grad_norm": 0.8508349657058716,
      "learning_rate": 0.00015098371416753963,
      "loss": 1.2222,
      "step": 3420
    },
    {
      "epoch": 0.5144356955380578,
      "grad_norm": 0.870098352432251,
      "learning_rate": 0.00015025503871168432,
      "loss": 1.2107,
      "step": 3430
    },
    {
      "epoch": 0.5159355080614924,
      "grad_norm": 0.7887611985206604,
      "learning_rate": 0.00014952635723717642,
      "loss": 1.2469,
      "step": 3440
    },
    {
      "epoch": 0.5174353205849269,
      "grad_norm": 0.8463996052742004,
      "learning_rate": 0.000148797686940153,
      "loss": 1.2251,
      "step": 3450
    },
    {
      "epoch": 0.5189351331083615,
      "grad_norm": 0.8258456587791443,
      "learning_rate": 0.0001480690450164873,
      "loss": 1.2241,
      "step": 3460
    },
    {
      "epoch": 0.520434945631796,
      "grad_norm": 0.8315289616584778,
      "learning_rate": 0.00014734044866138312,
      "loss": 1.1998,
      "step": 3470
    },
    {
      "epoch": 0.5219347581552306,
      "grad_norm": 0.8053680062294006,
      "learning_rate": 0.00014661191506896867,
      "loss": 1.2062,
      "step": 3480
    },
    {
      "epoch": 0.5234345706786652,
      "grad_norm": 0.7926360368728638,
      "learning_rate": 0.0001458834614318912,
      "loss": 1.2092,
      "step": 3490
    },
    {
      "epoch": 0.5249343832020997,
      "grad_norm": 0.7700707912445068,
      "learning_rate": 0.00014515510494091102,
      "loss": 1.2077,
      "step": 3500
    },
    {
      "epoch": 0.5264341957255343,
      "grad_norm": 0.8049729466438293,
      "learning_rate": 0.00014442686278449588,
      "loss": 1.2134,
      "step": 3510
    },
    {
      "epoch": 0.5279340082489689,
      "grad_norm": 0.8229398727416992,
      "learning_rate": 0.00014369875214841548,
      "loss": 1.2235,
      "step": 3520
    },
    {
      "epoch": 0.5294338207724034,
      "grad_norm": 0.7805364727973938,
      "learning_rate": 0.0001429707902153355,
      "loss": 1.2315,
      "step": 3530
    },
    {
      "epoch": 0.530933633295838,
      "grad_norm": 0.7947820425033569,
      "learning_rate": 0.0001422429941644127,
      "loss": 1.2166,
      "step": 3540
    },
    {
      "epoch": 0.5324334458192725,
      "grad_norm": 0.7772228121757507,
      "learning_rate": 0.000141515381170889,
      "loss": 1.2267,
      "step": 3550
    },
    {
      "epoch": 0.5339332583427071,
      "grad_norm": 0.7693078517913818,
      "learning_rate": 0.00014078796840568647,
      "loss": 1.2173,
      "step": 3560
    },
    {
      "epoch": 0.5354330708661418,
      "grad_norm": 0.8406286239624023,
      "learning_rate": 0.0001400607730350018,
      "loss": 1.2161,
      "step": 3570
    },
    {
      "epoch": 0.5369328833895763,
      "grad_norm": 0.7883846163749695,
      "learning_rate": 0.0001393338122199016,
      "loss": 1.23,
      "step": 3580
    },
    {
      "epoch": 0.5384326959130109,
      "grad_norm": 0.7825552821159363,
      "learning_rate": 0.00013860710311591713,
      "loss": 1.2165,
      "step": 3590
    },
    {
      "epoch": 0.5399325084364455,
      "grad_norm": 0.9384158253669739,
      "learning_rate": 0.00013788066287263946,
      "loss": 1.2197,
      "step": 3600
    },
    {
      "epoch": 0.54143232095988,
      "grad_norm": 0.888299286365509,
      "learning_rate": 0.00013715450863331495,
      "loss": 1.2082,
      "step": 3610
    },
    {
      "epoch": 0.5429321334833146,
      "grad_norm": 0.8535633087158203,
      "learning_rate": 0.00013642865753444043,
      "loss": 1.212,
      "step": 3620
    },
    {
      "epoch": 0.5444319460067492,
      "grad_norm": 0.8287407159805298,
      "learning_rate": 0.000135703126705359,
      "loss": 1.2009,
      "step": 3630
    },
    {
      "epoch": 0.5459317585301837,
      "grad_norm": 0.7710711359977722,
      "learning_rate": 0.00013497793326785573,
      "loss": 1.2205,
      "step": 3640
    },
    {
      "epoch": 0.5474315710536183,
      "grad_norm": 0.7808111310005188,
      "learning_rate": 0.00013425309433575365,
      "loss": 1.2149,
      "step": 3650
    },
    {
      "epoch": 0.5489313835770528,
      "grad_norm": 0.7702746987342834,
      "learning_rate": 0.0001335286270145096,
      "loss": 1.197,
      "step": 3660
    },
    {
      "epoch": 0.5504311961004874,
      "grad_norm": 0.832952618598938,
      "learning_rate": 0.00013280454840081105,
      "loss": 1.2075,
      "step": 3670
    },
    {
      "epoch": 0.551931008623922,
      "grad_norm": 0.7749794125556946,
      "learning_rate": 0.0001320808755821722,
      "loss": 1.2136,
      "step": 3680
    },
    {
      "epoch": 0.5534308211473565,
      "grad_norm": 0.8737798929214478,
      "learning_rate": 0.00013135762563653097,
      "loss": 1.2017,
      "step": 3690
    },
    {
      "epoch": 0.5549306336707911,
      "grad_norm": 0.8434765934944153,
      "learning_rate": 0.00013063481563184589,
      "loss": 1.1912,
      "step": 3700
    },
    {
      "epoch": 0.5564304461942258,
      "grad_norm": 0.7952425479888916,
      "learning_rate": 0.00012991246262569327,
      "loss": 1.2148,
      "step": 3710
    },
    {
      "epoch": 0.5579302587176603,
      "grad_norm": 0.8273133039474487,
      "learning_rate": 0.00012919058366486492,
      "loss": 1.219,
      "step": 3720
    },
    {
      "epoch": 0.5594300712410949,
      "grad_norm": 0.86753249168396,
      "learning_rate": 0.00012846919578496545,
      "loss": 1.1893,
      "step": 3730
    },
    {
      "epoch": 0.5609298837645295,
      "grad_norm": 0.8196529746055603,
      "learning_rate": 0.00012774831601001054,
      "loss": 1.2166,
      "step": 3740
    },
    {
      "epoch": 0.562429696287964,
      "grad_norm": 0.9246373772621155,
      "learning_rate": 0.00012702796135202518,
      "loss": 1.2296,
      "step": 3750
    },
    {
      "epoch": 0.5639295088113986,
      "grad_norm": 0.8339414000511169,
      "learning_rate": 0.00012630814881064206,
      "loss": 1.2164,
      "step": 3760
    },
    {
      "epoch": 0.5654293213348331,
      "grad_norm": 0.8195322155952454,
      "learning_rate": 0.00012558889537270048,
      "loss": 1.2031,
      "step": 3770
    },
    {
      "epoch": 0.5669291338582677,
      "grad_norm": 0.8975194096565247,
      "learning_rate": 0.0001248702180118455,
      "loss": 1.2236,
      "step": 3780
    },
    {
      "epoch": 0.5684289463817023,
      "grad_norm": 0.7996919751167297,
      "learning_rate": 0.00012415213368812731,
      "loss": 1.1993,
      "step": 3790
    },
    {
      "epoch": 0.5699287589051368,
      "grad_norm": 0.7755063772201538,
      "learning_rate": 0.00012343465934760102,
      "loss": 1.2084,
      "step": 3800
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 0.8294559717178345,
      "learning_rate": 0.00012271781192192688,
      "loss": 1.2175,
      "step": 3810
    },
    {
      "epoch": 0.572928383952006,
      "grad_norm": 0.8047690391540527,
      "learning_rate": 0.00012200160832797046,
      "loss": 1.1986,
      "step": 3820
    },
    {
      "epoch": 0.5744281964754405,
      "grad_norm": 0.7778790593147278,
      "learning_rate": 0.0001212860654674036,
      "loss": 1.2211,
      "step": 3830
    },
    {
      "epoch": 0.5759280089988752,
      "grad_norm": 0.8431829810142517,
      "learning_rate": 0.00012057120022630546,
      "loss": 1.2089,
      "step": 3840
    },
    {
      "epoch": 0.5774278215223098,
      "grad_norm": 0.8307316899299622,
      "learning_rate": 0.00011985702947476424,
      "loss": 1.2035,
      "step": 3850
    },
    {
      "epoch": 0.5789276340457443,
      "grad_norm": 0.9299211502075195,
      "learning_rate": 0.00011914357006647877,
      "loss": 1.1933,
      "step": 3860
    },
    {
      "epoch": 0.5804274465691789,
      "grad_norm": 0.9003536105155945,
      "learning_rate": 0.00011843083883836084,
      "loss": 1.2093,
      "step": 3870
    },
    {
      "epoch": 0.5819272590926134,
      "grad_norm": 0.8364601731300354,
      "learning_rate": 0.0001177188526101381,
      "loss": 1.2051,
      "step": 3880
    },
    {
      "epoch": 0.583427071616048,
      "grad_norm": 0.877537190914154,
      "learning_rate": 0.00011700762818395682,
      "loss": 1.213,
      "step": 3890
    },
    {
      "epoch": 0.5849268841394826,
      "grad_norm": 0.820450484752655,
      "learning_rate": 0.0001162971823439856,
      "loss": 1.2025,
      "step": 3900
    },
    {
      "epoch": 0.5864266966629171,
      "grad_norm": 0.7949150800704956,
      "learning_rate": 0.00011558753185601922,
      "loss": 1.2006,
      "step": 3910
    },
    {
      "epoch": 0.5879265091863517,
      "grad_norm": 0.7832253575325012,
      "learning_rate": 0.00011487869346708289,
      "loss": 1.1894,
      "step": 3920
    },
    {
      "epoch": 0.5894263217097863,
      "grad_norm": 0.7712039351463318,
      "learning_rate": 0.00011417068390503716,
      "loss": 1.2076,
      "step": 3930
    },
    {
      "epoch": 0.5909261342332208,
      "grad_norm": 0.8228356838226318,
      "learning_rate": 0.00011346351987818307,
      "loss": 1.1907,
      "step": 3940
    },
    {
      "epoch": 0.5924259467566554,
      "grad_norm": 0.8279032111167908,
      "learning_rate": 0.00011275721807486805,
      "loss": 1.2132,
      "step": 3950
    },
    {
      "epoch": 0.59392575928009,
      "grad_norm": 0.8680859208106995,
      "learning_rate": 0.00011205179516309172,
      "loss": 1.199,
      "step": 3960
    },
    {
      "epoch": 0.5954255718035245,
      "grad_norm": 0.9166079759597778,
      "learning_rate": 0.00011134726779011288,
      "loss": 1.2,
      "step": 3970
    },
    {
      "epoch": 0.5969253843269592,
      "grad_norm": 0.8387262225151062,
      "learning_rate": 0.00011064365258205658,
      "loss": 1.1922,
      "step": 3980
    },
    {
      "epoch": 0.5984251968503937,
      "grad_norm": 0.7966826558113098,
      "learning_rate": 0.00010994096614352153,
      "loss": 1.2041,
      "step": 3990
    },
    {
      "epoch": 0.5999250093738283,
      "grad_norm": 0.8371879458427429,
      "learning_rate": 0.00010923922505718863,
      "loss": 1.2041,
      "step": 4000
    },
    {
      "epoch": 0.6002249718785152,
      "eval_loss": 1.2461844682693481,
      "eval_runtime": 34.5014,
      "eval_samples_per_second": 724.607,
      "eval_steps_per_second": 90.576,
      "step": 4002
    },
    {
      "epoch": 0.6014248218972629,
      "grad_norm": 0.8346044421195984,
      "learning_rate": 0.00010853844588342926,
      "loss": 1.1886,
      "step": 4010
    },
    {
      "epoch": 0.6029246344206974,
      "grad_norm": 0.7978150248527527,
      "learning_rate": 0.00010783864515991481,
      "loss": 1.2161,
      "step": 4020
    },
    {
      "epoch": 0.604424446944132,
      "grad_norm": 0.8575156331062317,
      "learning_rate": 0.00010713983940122617,
      "loss": 1.2171,
      "step": 4030
    },
    {
      "epoch": 0.6059242594675666,
      "grad_norm": 0.8124422430992126,
      "learning_rate": 0.00010644204509846398,
      "loss": 1.1864,
      "step": 4040
    },
    {
      "epoch": 0.6074240719910011,
      "grad_norm": 0.8295300006866455,
      "learning_rate": 0.00010574527871885977,
      "loss": 1.2312,
      "step": 4050
    },
    {
      "epoch": 0.6089238845144357,
      "grad_norm": 0.8077391982078552,
      "learning_rate": 0.00010504955670538699,
      "loss": 1.2036,
      "step": 4060
    },
    {
      "epoch": 0.6104236970378702,
      "grad_norm": 0.839034914970398,
      "learning_rate": 0.00010435489547637316,
      "loss": 1.2101,
      "step": 4070
    },
    {
      "epoch": 0.6119235095613048,
      "grad_norm": 0.7653098106384277,
      "learning_rate": 0.00010366131142511228,
      "loss": 1.2127,
      "step": 4080
    },
    {
      "epoch": 0.6134233220847394,
      "grad_norm": 0.7839916944503784,
      "learning_rate": 0.00010296882091947826,
      "loss": 1.1972,
      "step": 4090
    },
    {
      "epoch": 0.6149231346081739,
      "grad_norm": 0.8738676905632019,
      "learning_rate": 0.00010227744030153821,
      "loss": 1.1864,
      "step": 4100
    },
    {
      "epoch": 0.6164229471316085,
      "grad_norm": 0.8577001094818115,
      "learning_rate": 0.0001015871858871672,
      "loss": 1.2282,
      "step": 4110
    },
    {
      "epoch": 0.6179227596550432,
      "grad_norm": 0.8081793785095215,
      "learning_rate": 0.00010089807396566306,
      "loss": 1.2156,
      "step": 4120
    },
    {
      "epoch": 0.6194225721784777,
      "grad_norm": 0.8585608601570129,
      "learning_rate": 0.00010021012079936174,
      "loss": 1.1875,
      "step": 4130
    },
    {
      "epoch": 0.6209223847019123,
      "grad_norm": 0.8267788290977478,
      "learning_rate": 9.952334262325399e-05,
      "loss": 1.175,
      "step": 4140
    },
    {
      "epoch": 0.6224221972253469,
      "grad_norm": 0.840057373046875,
      "learning_rate": 9.883775564460193e-05,
      "loss": 1.1884,
      "step": 4150
    },
    {
      "epoch": 0.6239220097487814,
      "grad_norm": 0.8279968500137329,
      "learning_rate": 9.815337604255665e-05,
      "loss": 1.1902,
      "step": 4160
    },
    {
      "epoch": 0.625421822272216,
      "grad_norm": 0.7800420522689819,
      "learning_rate": 9.747021996777624e-05,
      "loss": 1.1982,
      "step": 4170
    },
    {
      "epoch": 0.6269216347956506,
      "grad_norm": 0.7733381390571594,
      "learning_rate": 9.678830354204504e-05,
      "loss": 1.2104,
      "step": 4180
    },
    {
      "epoch": 0.6284214473190851,
      "grad_norm": 0.8349838256835938,
      "learning_rate": 9.610764285789271e-05,
      "loss": 1.1976,
      "step": 4190
    },
    {
      "epoch": 0.6299212598425197,
      "grad_norm": 0.8445114493370056,
      "learning_rate": 9.542825397821485e-05,
      "loss": 1.1877,
      "step": 4200
    },
    {
      "epoch": 0.6314210723659542,
      "grad_norm": 0.8512496948242188,
      "learning_rate": 9.475015293589373e-05,
      "loss": 1.1979,
      "step": 4210
    },
    {
      "epoch": 0.6329208848893888,
      "grad_norm": 0.817348837852478,
      "learning_rate": 9.407335573341997e-05,
      "loss": 1.1888,
      "step": 4220
    },
    {
      "epoch": 0.6344206974128234,
      "grad_norm": 0.8680126070976257,
      "learning_rate": 9.339787834251489e-05,
      "loss": 1.1961,
      "step": 4230
    },
    {
      "epoch": 0.6359205099362579,
      "grad_norm": 0.82338547706604,
      "learning_rate": 9.272373670375362e-05,
      "loss": 1.196,
      "step": 4240
    },
    {
      "epoch": 0.6374203224596925,
      "grad_norm": 0.7532997131347656,
      "learning_rate": 9.205094672618889e-05,
      "loss": 1.1987,
      "step": 4250
    },
    {
      "epoch": 0.6389201349831272,
      "grad_norm": 0.8462244272232056,
      "learning_rate": 9.137952428697568e-05,
      "loss": 1.1696,
      "step": 4260
    },
    {
      "epoch": 0.6404199475065617,
      "grad_norm": 0.7773210406303406,
      "learning_rate": 9.070948523099643e-05,
      "loss": 1.1903,
      "step": 4270
    },
    {
      "epoch": 0.6419197600299963,
      "grad_norm": 0.8117038607597351,
      "learning_rate": 9.004084537048708e-05,
      "loss": 1.1968,
      "step": 4280
    },
    {
      "epoch": 0.6434195725534309,
      "grad_norm": 0.7861129641532898,
      "learning_rate": 8.937362048466404e-05,
      "loss": 1.1933,
      "step": 4290
    },
    {
      "epoch": 0.6449193850768654,
      "grad_norm": 0.777773916721344,
      "learning_rate": 8.870782631935184e-05,
      "loss": 1.2017,
      "step": 4300
    },
    {
      "epoch": 0.6464191976003,
      "grad_norm": 0.8288027048110962,
      "learning_rate": 8.804347858661131e-05,
      "loss": 1.1937,
      "step": 4310
    },
    {
      "epoch": 0.6479190101237345,
      "grad_norm": 0.8184522390365601,
      "learning_rate": 8.73805929643691e-05,
      "loss": 1.1911,
      "step": 4320
    },
    {
      "epoch": 0.6494188226471691,
      "grad_norm": 0.7932900190353394,
      "learning_rate": 8.67191850960475e-05,
      "loss": 1.1875,
      "step": 4330
    },
    {
      "epoch": 0.6509186351706037,
      "grad_norm": 0.8443998694419861,
      "learning_rate": 8.605927059019528e-05,
      "loss": 1.1897,
      "step": 4340
    },
    {
      "epoch": 0.6524184476940382,
      "grad_norm": 0.8229474425315857,
      "learning_rate": 8.540086502011935e-05,
      "loss": 1.2059,
      "step": 4350
    },
    {
      "epoch": 0.6539182602174728,
      "grad_norm": 0.7777069211006165,
      "learning_rate": 8.47439839235174e-05,
      "loss": 1.2047,
      "step": 4360
    },
    {
      "epoch": 0.6554180727409074,
      "grad_norm": 0.8484175205230713,
      "learning_rate": 8.408864280211115e-05,
      "loss": 1.1743,
      "step": 4370
    },
    {
      "epoch": 0.6569178852643419,
      "grad_norm": 0.8335903882980347,
      "learning_rate": 8.343485712128026e-05,
      "loss": 1.1826,
      "step": 4380
    },
    {
      "epoch": 0.6584176977877765,
      "grad_norm": 0.827379584312439,
      "learning_rate": 8.278264230969769e-05,
      "loss": 1.1924,
      "step": 4390
    },
    {
      "epoch": 0.6599175103112112,
      "grad_norm": 0.7646244168281555,
      "learning_rate": 8.213201375896563e-05,
      "loss": 1.1829,
      "step": 4400
    },
    {
      "epoch": 0.6614173228346457,
      "grad_norm": 0.7910548448562622,
      "learning_rate": 8.14829868232519e-05,
      "loss": 1.1861,
      "step": 4410
    },
    {
      "epoch": 0.6629171353580803,
      "grad_norm": 0.8238165974617004,
      "learning_rate": 8.083557681892797e-05,
      "loss": 1.1852,
      "step": 4420
    },
    {
      "epoch": 0.6644169478815148,
      "grad_norm": 0.8701585531234741,
      "learning_rate": 8.018979902420746e-05,
      "loss": 1.1935,
      "step": 4430
    },
    {
      "epoch": 0.6659167604049494,
      "grad_norm": 0.8514190316200256,
      "learning_rate": 7.954566867878538e-05,
      "loss": 1.1901,
      "step": 4440
    },
    {
      "epoch": 0.667416572928384,
      "grad_norm": 0.8281582593917847,
      "learning_rate": 7.890320098347861e-05,
      "loss": 1.1747,
      "step": 4450
    },
    {
      "epoch": 0.6689163854518185,
      "grad_norm": 0.8169477581977844,
      "learning_rate": 7.82624110998673e-05,
      "loss": 1.182,
      "step": 4460
    },
    {
      "epoch": 0.6704161979752531,
      "grad_norm": 0.9043864011764526,
      "learning_rate": 7.762331414993697e-05,
      "loss": 1.1811,
      "step": 4470
    },
    {
      "epoch": 0.6719160104986877,
      "grad_norm": 0.9023800492286682,
      "learning_rate": 7.698592521572155e-05,
      "loss": 1.182,
      "step": 4480
    },
    {
      "epoch": 0.6734158230221222,
      "grad_norm": 0.8538755178451538,
      "learning_rate": 7.635025933894747e-05,
      "loss": 1.2004,
      "step": 4490
    },
    {
      "epoch": 0.6749156355455568,
      "grad_norm": 0.8358992338180542,
      "learning_rate": 7.571633152067901e-05,
      "loss": 1.1949,
      "step": 4500
    },
    {
      "epoch": 0.6764154480689913,
      "grad_norm": 0.8562530279159546,
      "learning_rate": 7.508415672096389e-05,
      "loss": 1.1891,
      "step": 4510
    },
    {
      "epoch": 0.6779152605924259,
      "grad_norm": 0.867493212223053,
      "learning_rate": 7.445374985848035e-05,
      "loss": 1.172,
      "step": 4520
    },
    {
      "epoch": 0.6794150731158605,
      "grad_norm": 0.8086623549461365,
      "learning_rate": 7.382512581018514e-05,
      "loss": 1.2105,
      "step": 4530
    },
    {
      "epoch": 0.6809148856392951,
      "grad_norm": 0.8249533176422119,
      "learning_rate": 7.31982994109626e-05,
      "loss": 1.1874,
      "step": 4540
    },
    {
      "epoch": 0.6824146981627297,
      "grad_norm": 0.7889946103096008,
      "learning_rate": 7.25732854532741e-05,
      "loss": 1.1974,
      "step": 4550
    },
    {
      "epoch": 0.6839145106861643,
      "grad_norm": 0.8312809467315674,
      "learning_rate": 7.195009868680954e-05,
      "loss": 1.1875,
      "step": 4560
    },
    {
      "epoch": 0.6854143232095988,
      "grad_norm": 0.8431591987609863,
      "learning_rate": 7.13287538181387e-05,
      "loss": 1.1794,
      "step": 4570
    },
    {
      "epoch": 0.6869141357330334,
      "grad_norm": 0.8547993302345276,
      "learning_rate": 7.070926551036469e-05,
      "loss": 1.1723,
      "step": 4580
    },
    {
      "epoch": 0.688413948256468,
      "grad_norm": 0.8453377485275269,
      "learning_rate": 7.009164838277754e-05,
      "loss": 1.1835,
      "step": 4590
    },
    {
      "epoch": 0.6899137607799025,
      "grad_norm": 0.8473166823387146,
      "learning_rate": 6.947591701050932e-05,
      "loss": 1.2166,
      "step": 4600
    },
    {
      "epoch": 0.6914135733033371,
      "grad_norm": 0.882140040397644,
      "learning_rate": 6.886208592419043e-05,
      "loss": 1.1916,
      "step": 4610
    },
    {
      "epoch": 0.6929133858267716,
      "grad_norm": 0.7941833138465881,
      "learning_rate": 6.825016960960616e-05,
      "loss": 1.1955,
      "step": 4620
    },
    {
      "epoch": 0.6944131983502062,
      "grad_norm": 0.8405663967132568,
      "learning_rate": 6.764018250735532e-05,
      "loss": 1.1734,
      "step": 4630
    },
    {
      "epoch": 0.6959130108736408,
      "grad_norm": 0.8553961515426636,
      "learning_rate": 6.703213901250931e-05,
      "loss": 1.1743,
      "step": 4640
    },
    {
      "epoch": 0.6974128233970753,
      "grad_norm": 0.8203520178794861,
      "learning_rate": 6.64260534742723e-05,
      "loss": 1.181,
      "step": 4650
    },
    {
      "epoch": 0.6989126359205099,
      "grad_norm": 0.9138359427452087,
      "learning_rate": 6.582194019564266e-05,
      "loss": 1.1663,
      "step": 4660
    },
    {
      "epoch": 0.700262467191601,
      "eval_loss": 1.2265734672546387,
      "eval_runtime": 34.5391,
      "eval_samples_per_second": 723.818,
      "eval_steps_per_second": 90.477,
      "step": 4669
    },
    {
      "epoch": 0.7004124484439445,
      "grad_norm": 0.8695697784423828,
      "learning_rate": 6.521981343307554e-05,
      "loss": 1.186,
      "step": 4670
    },
    {
      "epoch": 0.7019122609673791,
      "grad_norm": 0.7855265140533447,
      "learning_rate": 6.461968739614639e-05,
      "loss": 1.1716,
      "step": 4680
    },
    {
      "epoch": 0.7034120734908137,
      "grad_norm": 0.8557237982749939,
      "learning_rate": 6.402157624721546e-05,
      "loss": 1.1854,
      "step": 4690
    },
    {
      "epoch": 0.7049118860142483,
      "grad_norm": 0.9032356142997742,
      "learning_rate": 6.342549410109372e-05,
      "loss": 1.1622,
      "step": 4700
    },
    {
      "epoch": 0.7064116985376828,
      "grad_norm": 0.845805287361145,
      "learning_rate": 6.283145502470976e-05,
      "loss": 1.1895,
      "step": 4710
    },
    {
      "epoch": 0.7079115110611174,
      "grad_norm": 0.8459937572479248,
      "learning_rate": 6.223947303677793e-05,
      "loss": 1.1948,
      "step": 4720
    },
    {
      "epoch": 0.709411323584552,
      "grad_norm": 0.7924135327339172,
      "learning_rate": 6.164956210746723e-05,
      "loss": 1.1826,
      "step": 4730
    },
    {
      "epoch": 0.7109111361079865,
      "grad_norm": 0.8918642997741699,
      "learning_rate": 6.106173615807186e-05,
      "loss": 1.1798,
      "step": 4740
    },
    {
      "epoch": 0.7124109486314211,
      "grad_norm": 0.7782163619995117,
      "learning_rate": 6.047600906068269e-05,
      "loss": 1.1677,
      "step": 4750
    },
    {
      "epoch": 0.7139107611548556,
      "grad_norm": 0.8455360531806946,
      "learning_rate": 5.989239463785971e-05,
      "loss": 1.1956,
      "step": 4760
    },
    {
      "epoch": 0.7154105736782902,
      "grad_norm": 0.8486021161079407,
      "learning_rate": 5.9310906662306125e-05,
      "loss": 1.1881,
      "step": 4770
    },
    {
      "epoch": 0.7169103862017248,
      "grad_norm": 0.8694136142730713,
      "learning_rate": 5.8731558856542935e-05,
      "loss": 1.1795,
      "step": 4780
    },
    {
      "epoch": 0.7184101987251593,
      "grad_norm": 0.8481091260910034,
      "learning_rate": 5.8154364892585574e-05,
      "loss": 1.1663,
      "step": 4790
    },
    {
      "epoch": 0.7199100112485939,
      "grad_norm": 0.7921723127365112,
      "learning_rate": 5.75793383916208e-05,
      "loss": 1.1648,
      "step": 4800
    },
    {
      "epoch": 0.7214098237720284,
      "grad_norm": 0.8086884021759033,
      "learning_rate": 5.70064929236855e-05,
      "loss": 1.181,
      "step": 4810
    },
    {
      "epoch": 0.7229096362954631,
      "grad_norm": 0.854877233505249,
      "learning_rate": 5.643584200734659e-05,
      "loss": 1.1877,
      "step": 4820
    },
    {
      "epoch": 0.7244094488188977,
      "grad_norm": 0.9068853259086609,
      "learning_rate": 5.586739910938161e-05,
      "loss": 1.184,
      "step": 4830
    },
    {
      "epoch": 0.7259092613423322,
      "grad_norm": 0.7965474724769592,
      "learning_rate": 5.5301177644461164e-05,
      "loss": 1.1629,
      "step": 4840
    },
    {
      "epoch": 0.7274090738657668,
      "grad_norm": 0.8608559966087341,
      "learning_rate": 5.4737190974832426e-05,
      "loss": 1.1807,
      "step": 4850
    },
    {
      "epoch": 0.7289088863892014,
      "grad_norm": 0.8948774337768555,
      "learning_rate": 5.417545241000353e-05,
      "loss": 1.1759,
      "step": 4860
    },
    {
      "epoch": 0.7304086989126359,
      "grad_norm": 0.9525447487831116,
      "learning_rate": 5.361597520642981e-05,
      "loss": 1.1643,
      "step": 4870
    },
    {
      "epoch": 0.7319085114360705,
      "grad_norm": 0.905668318271637,
      "learning_rate": 5.3058772567200595e-05,
      "loss": 1.1799,
      "step": 4880
    },
    {
      "epoch": 0.7334083239595051,
      "grad_norm": 0.8792032599449158,
      "learning_rate": 5.250385764172802e-05,
      "loss": 1.1766,
      "step": 4890
    },
    {
      "epoch": 0.7349081364829396,
      "grad_norm": 0.9094644784927368,
      "learning_rate": 5.195124352543636e-05,
      "loss": 1.192,
      "step": 4900
    },
    {
      "epoch": 0.7364079490063742,
      "grad_norm": 0.8372027277946472,
      "learning_rate": 5.140094325945323e-05,
      "loss": 1.1655,
      "step": 4910
    },
    {
      "epoch": 0.7379077615298087,
      "grad_norm": 0.9662985801696777,
      "learning_rate": 5.085296983030164e-05,
      "loss": 1.1926,
      "step": 4920
    },
    {
      "epoch": 0.7394075740532433,
      "grad_norm": 0.8691169023513794,
      "learning_rate": 5.030733616959384e-05,
      "loss": 1.1785,
      "step": 4930
    },
    {
      "epoch": 0.7409073865766779,
      "grad_norm": 0.8685835599899292,
      "learning_rate": 4.976405515372577e-05,
      "loss": 1.182,
      "step": 4940
    },
    {
      "epoch": 0.7424071991001124,
      "grad_norm": 0.8488807082176208,
      "learning_rate": 4.922313960357336e-05,
      "loss": 1.1782,
      "step": 4950
    },
    {
      "epoch": 0.7439070116235471,
      "grad_norm": 0.8728071451187134,
      "learning_rate": 4.868460228419003e-05,
      "loss": 1.1837,
      "step": 4960
    },
    {
      "epoch": 0.7454068241469817,
      "grad_norm": 0.8640130162239075,
      "learning_rate": 4.814845590450544e-05,
      "loss": 1.1762,
      "step": 4970
    },
    {
      "epoch": 0.7469066366704162,
      "grad_norm": 0.861436128616333,
      "learning_rate": 4.761471311702541e-05,
      "loss": 1.1605,
      "step": 4980
    },
    {
      "epoch": 0.7484064491938508,
      "grad_norm": 0.9322028756141663,
      "learning_rate": 4.70833865175334e-05,
      "loss": 1.1804,
      "step": 4990
    },
    {
      "epoch": 0.7499062617172854,
      "grad_norm": 0.8335092663764954,
      "learning_rate": 4.6554488644793555e-05,
      "loss": 1.1822,
      "step": 5000
    },
    {
      "epoch": 0.7514060742407199,
      "grad_norm": 0.8994686603546143,
      "learning_rate": 4.602803198025429e-05,
      "loss": 1.1805,
      "step": 5010
    },
    {
      "epoch": 0.7529058867641545,
      "grad_norm": 0.9040279388427734,
      "learning_rate": 4.550402894775408e-05,
      "loss": 1.1593,
      "step": 5020
    },
    {
      "epoch": 0.754405699287589,
      "grad_norm": 0.8711543083190918,
      "learning_rate": 4.49824919132283e-05,
      "loss": 1.1526,
      "step": 5030
    },
    {
      "epoch": 0.7559055118110236,
      "grad_norm": 0.8705037832260132,
      "learning_rate": 4.446343318441719e-05,
      "loss": 1.1688,
      "step": 5040
    },
    {
      "epoch": 0.7574053243344582,
      "grad_norm": 0.8655483722686768,
      "learning_rate": 4.394686501057553e-05,
      "loss": 1.1758,
      "step": 5050
    },
    {
      "epoch": 0.7589051368578927,
      "grad_norm": 0.813109815120697,
      "learning_rate": 4.343279958218352e-05,
      "loss": 1.1772,
      "step": 5060
    },
    {
      "epoch": 0.7604049493813273,
      "grad_norm": 0.8437618613243103,
      "learning_rate": 4.29212490306592e-05,
      "loss": 1.1725,
      "step": 5070
    },
    {
      "epoch": 0.7619047619047619,
      "grad_norm": 0.908439576625824,
      "learning_rate": 4.241222542807211e-05,
      "loss": 1.1821,
      "step": 5080
    },
    {
      "epoch": 0.7634045744281964,
      "grad_norm": 0.8363406658172607,
      "learning_rate": 4.19057407868583e-05,
      "loss": 1.1671,
      "step": 5090
    },
    {
      "epoch": 0.7649043869516311,
      "grad_norm": 0.8935152888298035,
      "learning_rate": 4.140180705953689e-05,
      "loss": 1.1831,
      "step": 5100
    },
    {
      "epoch": 0.7664041994750657,
      "grad_norm": 0.9219470620155334,
      "learning_rate": 4.090043613842823e-05,
      "loss": 1.1837,
      "step": 5110
    },
    {
      "epoch": 0.7679040119985002,
      "grad_norm": 0.8783389329910278,
      "learning_rate": 4.0401639855372884e-05,
      "loss": 1.1915,
      "step": 5120
    },
    {
      "epoch": 0.7694038245219348,
      "grad_norm": 0.8177280426025391,
      "learning_rate": 3.990542998145262e-05,
      "loss": 1.1598,
      "step": 5130
    },
    {
      "epoch": 0.7709036370453693,
      "grad_norm": 0.8821897506713867,
      "learning_rate": 3.941181822671273e-05,
      "loss": 1.1794,
      "step": 5140
    },
    {
      "epoch": 0.7724034495688039,
      "grad_norm": 0.8683632612228394,
      "learning_rate": 3.892081623988541e-05,
      "loss": 1.1904,
      "step": 5150
    },
    {
      "epoch": 0.7739032620922385,
      "grad_norm": 0.800815999507904,
      "learning_rate": 3.8432435608115e-05,
      "loss": 1.1648,
      "step": 5160
    },
    {
      "epoch": 0.775403074615673,
      "grad_norm": 0.8233063817024231,
      "learning_rate": 3.794668785668465e-05,
      "loss": 1.1718,
      "step": 5170
    },
    {
      "epoch": 0.7769028871391076,
      "grad_norm": 0.8883649110794067,
      "learning_rate": 3.7463584448744186e-05,
      "loss": 1.1682,
      "step": 5180
    },
    {
      "epoch": 0.7784026996625422,
      "grad_norm": 0.8076657652854919,
      "learning_rate": 3.6983136785039636e-05,
      "loss": 1.1635,
      "step": 5190
    },
    {
      "epoch": 0.7799025121859767,
      "grad_norm": 0.7962640523910522,
      "learning_rate": 3.650535620364407e-05,
      "loss": 1.1731,
      "step": 5200
    },
    {
      "epoch": 0.7814023247094113,
      "grad_norm": 0.8250430822372437,
      "learning_rate": 3.603025397969037e-05,
      "loss": 1.1603,
      "step": 5210
    },
    {
      "epoch": 0.7829021372328459,
      "grad_norm": 0.9766924977302551,
      "learning_rate": 3.555784132510472e-05,
      "loss": 1.1686,
      "step": 5220
    },
    {
      "epoch": 0.7844019497562804,
      "grad_norm": 0.9148427844047546,
      "learning_rate": 3.508812938834227e-05,
      "loss": 1.1703,
      "step": 5230
    },
    {
      "epoch": 0.7859017622797151,
      "grad_norm": 0.8292137980461121,
      "learning_rate": 3.4621129254124106e-05,
      "loss": 1.1565,
      "step": 5240
    },
    {
      "epoch": 0.7874015748031497,
      "grad_norm": 0.8619733452796936,
      "learning_rate": 3.415685194317539e-05,
      "loss": 1.1571,
      "step": 5250
    },
    {
      "epoch": 0.7889013873265842,
      "grad_norm": 0.8798223733901978,
      "learning_rate": 3.3695308411965564e-05,
      "loss": 1.1738,
      "step": 5260
    },
    {
      "epoch": 0.7904011998500188,
      "grad_norm": 0.8004907369613647,
      "learning_rate": 3.323650955244951e-05,
      "loss": 1.1769,
      "step": 5270
    },
    {
      "epoch": 0.7919010123734533,
      "grad_norm": 0.9049323201179504,
      "learning_rate": 3.2780466191810905e-05,
      "loss": 1.1641,
      "step": 5280
    },
    {
      "epoch": 0.7934008248968879,
      "grad_norm": 0.8505051136016846,
      "learning_rate": 3.232718909220631e-05,
      "loss": 1.1765,
      "step": 5290
    },
    {
      "epoch": 0.7949006374203225,
      "grad_norm": 0.878035843372345,
      "learning_rate": 3.187668895051135e-05,
      "loss": 1.1665,
      "step": 5300
    },
    {
      "epoch": 0.796400449943757,
      "grad_norm": 0.8694955706596375,
      "learning_rate": 3.14289763980683e-05,
      "loss": 1.1904,
      "step": 5310
    },
    {
      "epoch": 0.7979002624671916,
      "grad_norm": 0.917580783367157,
      "learning_rate": 3.0984062000435276e-05,
      "loss": 1.1738,
      "step": 5320
    },
    {
      "epoch": 0.7994000749906262,
      "grad_norm": 0.9352427124977112,
      "learning_rate": 3.054195625713668e-05,
      "loss": 1.1685,
      "step": 5330
    },
    {
      "epoch": 0.8002999625046869,
      "eval_loss": 1.210194706916809,
      "eval_runtime": 35.2273,
      "eval_samples_per_second": 709.676,
      "eval_steps_per_second": 88.71,
      "step": 5336
    },
    {
      "epoch": 0.8008998875140607,
      "grad_norm": 0.845475971698761,
      "learning_rate": 3.0102669601415575e-05,
      "loss": 1.1577,
      "step": 5340
    },
    {
      "epoch": 0.8023997000374953,
      "grad_norm": 0.8200697898864746,
      "learning_rate": 2.966621239998755e-05,
      "loss": 1.1577,
      "step": 5350
    },
    {
      "epoch": 0.8038995125609298,
      "grad_norm": 0.8095065355300903,
      "learning_rate": 2.9232594952795818e-05,
      "loss": 1.1589,
      "step": 5360
    },
    {
      "epoch": 0.8053993250843644,
      "grad_norm": 0.8357236385345459,
      "learning_rate": 2.8801827492768352e-05,
      "loss": 1.1808,
      "step": 5370
    },
    {
      "epoch": 0.8068991376077991,
      "grad_norm": 0.8971685171127319,
      "learning_rate": 2.8373920185576375e-05,
      "loss": 1.1649,
      "step": 5380
    },
    {
      "epoch": 0.8083989501312336,
      "grad_norm": 0.8829404711723328,
      "learning_rate": 2.7948883129394467e-05,
      "loss": 1.1626,
      "step": 5390
    },
    {
      "epoch": 0.8098987626546682,
      "grad_norm": 0.8451800346374512,
      "learning_rate": 2.7526726354662104e-05,
      "loss": 1.155,
      "step": 5400
    },
    {
      "epoch": 0.8113985751781028,
      "grad_norm": 0.8615455031394958,
      "learning_rate": 2.7107459823847106e-05,
      "loss": 1.1607,
      "step": 5410
    },
    {
      "epoch": 0.8128983877015373,
      "grad_norm": 0.9241278767585754,
      "learning_rate": 2.6691093431210596e-05,
      "loss": 1.1771,
      "step": 5420
    },
    {
      "epoch": 0.8143982002249719,
      "grad_norm": 0.8356271982192993,
      "learning_rate": 2.6277637002573288e-05,
      "loss": 1.1755,
      "step": 5430
    },
    {
      "epoch": 0.8158980127484065,
      "grad_norm": 0.8474392294883728,
      "learning_rate": 2.586710029508375e-05,
      "loss": 1.1652,
      "step": 5440
    },
    {
      "epoch": 0.817397825271841,
      "grad_norm": 0.9105897545814514,
      "learning_rate": 2.54594929969881e-05,
      "loss": 1.158,
      "step": 5450
    },
    {
      "epoch": 0.8188976377952756,
      "grad_norm": 0.8202654123306274,
      "learning_rate": 2.5054824727401502e-05,
      "loss": 1.1632,
      "step": 5460
    },
    {
      "epoch": 0.8203974503187101,
      "grad_norm": 0.8397065997123718,
      "learning_rate": 2.46531050360809e-05,
      "loss": 1.166,
      "step": 5470
    },
    {
      "epoch": 0.8218972628421447,
      "grad_norm": 0.8714698553085327,
      "learning_rate": 2.4254343403199945e-05,
      "loss": 1.1749,
      "step": 5480
    },
    {
      "epoch": 0.8233970753655793,
      "grad_norm": 0.8937433958053589,
      "learning_rate": 2.3858549239125034e-05,
      "loss": 1.1631,
      "step": 5490
    },
    {
      "epoch": 0.8248968878890138,
      "grad_norm": 0.8778429627418518,
      "learning_rate": 2.346573188419341e-05,
      "loss": 1.1661,
      "step": 5500
    },
    {
      "epoch": 0.8263967004124484,
      "grad_norm": 0.7826377153396606,
      "learning_rate": 2.3075900608492637e-05,
      "loss": 1.1686,
      "step": 5510
    },
    {
      "epoch": 0.8278965129358831,
      "grad_norm": 1.108136773109436,
      "learning_rate": 2.2689064611641794e-05,
      "loss": 1.1926,
      "step": 5520
    },
    {
      "epoch": 0.8293963254593176,
      "grad_norm": 0.8729405403137207,
      "learning_rate": 2.230523302257461e-05,
      "loss": 1.1554,
      "step": 5530
    },
    {
      "epoch": 0.8308961379827522,
      "grad_norm": 0.824364185333252,
      "learning_rate": 2.192441489932372e-05,
      "loss": 1.1715,
      "step": 5540
    },
    {
      "epoch": 0.8323959505061868,
      "grad_norm": 0.8968133330345154,
      "learning_rate": 2.154661922880708e-05,
      "loss": 1.1549,
      "step": 5550
    },
    {
      "epoch": 0.8338957630296213,
      "grad_norm": 0.9061416983604431,
      "learning_rate": 2.117185492661592e-05,
      "loss": 1.1516,
      "step": 5560
    },
    {
      "epoch": 0.8353955755530559,
      "grad_norm": 0.8204578161239624,
      "learning_rate": 2.0800130836804214e-05,
      "loss": 1.1637,
      "step": 5570
    },
    {
      "epoch": 0.8368953880764904,
      "grad_norm": 0.8293260931968689,
      "learning_rate": 2.043145573168003e-05,
      "loss": 1.1604,
      "step": 5580
    },
    {
      "epoch": 0.838395200599925,
      "grad_norm": 0.8907037377357483,
      "learning_rate": 2.0065838311598543e-05,
      "loss": 1.1783,
      "step": 5590
    },
    {
      "epoch": 0.8398950131233596,
      "grad_norm": 0.8842608332633972,
      "learning_rate": 1.9703287204756757e-05,
      "loss": 1.1576,
      "step": 5600
    },
    {
      "epoch": 0.8413948256467941,
      "grad_norm": 0.8065600991249084,
      "learning_rate": 1.9343810966989716e-05,
      "loss": 1.182,
      "step": 5610
    },
    {
      "epoch": 0.8428946381702287,
      "grad_norm": 0.8144896626472473,
      "learning_rate": 1.8987418081568683e-05,
      "loss": 1.1482,
      "step": 5620
    },
    {
      "epoch": 0.8443944506936633,
      "grad_norm": 0.8756045699119568,
      "learning_rate": 1.8634116959001106e-05,
      "loss": 1.1623,
      "step": 5630
    },
    {
      "epoch": 0.8458942632170978,
      "grad_norm": 0.9638755321502686,
      "learning_rate": 1.828391593683185e-05,
      "loss": 1.1479,
      "step": 5640
    },
    {
      "epoch": 0.8473940757405324,
      "grad_norm": 0.8920614123344421,
      "learning_rate": 1.7936823279446676e-05,
      "loss": 1.1531,
      "step": 5650
    },
    {
      "epoch": 0.8488938882639671,
      "grad_norm": 0.8344776630401611,
      "learning_rate": 1.7592847177877008e-05,
      "loss": 1.1642,
      "step": 5660
    },
    {
      "epoch": 0.8503937007874016,
      "grad_norm": 0.8521091341972351,
      "learning_rate": 1.725199574960689e-05,
      "loss": 1.1456,
      "step": 5670
    },
    {
      "epoch": 0.8518935133108362,
      "grad_norm": 0.8243123292922974,
      "learning_rate": 1.6914277038381145e-05,
      "loss": 1.1689,
      "step": 5680
    },
    {
      "epoch": 0.8533933258342707,
      "grad_norm": 0.8362585306167603,
      "learning_rate": 1.6579699014015783e-05,
      "loss": 1.1565,
      "step": 5690
    },
    {
      "epoch": 0.8548931383577053,
      "grad_norm": 0.8934555053710938,
      "learning_rate": 1.6248269572209716e-05,
      "loss": 1.158,
      "step": 5700
    },
    {
      "epoch": 0.8563929508811399,
      "grad_norm": 0.9755041599273682,
      "learning_rate": 1.5919996534358635e-05,
      "loss": 1.1412,
      "step": 5710
    },
    {
      "epoch": 0.8578927634045744,
      "grad_norm": 0.9437934160232544,
      "learning_rate": 1.5594887647370263e-05,
      "loss": 1.1528,
      "step": 5720
    },
    {
      "epoch": 0.859392575928009,
      "grad_norm": 0.8423501253128052,
      "learning_rate": 1.527295058348154e-05,
      "loss": 1.1374,
      "step": 5730
    },
    {
      "epoch": 0.8608923884514436,
      "grad_norm": 0.8530197739601135,
      "learning_rate": 1.4954192940077809e-05,
      "loss": 1.153,
      "step": 5740
    },
    {
      "epoch": 0.8623922009748781,
      "grad_norm": 0.9238439798355103,
      "learning_rate": 1.463862223951317e-05,
      "loss": 1.1658,
      "step": 5750
    },
    {
      "epoch": 0.8638920134983127,
      "grad_norm": 0.82196444272995,
      "learning_rate": 1.4326245928933178e-05,
      "loss": 1.1657,
      "step": 5760
    },
    {
      "epoch": 0.8653918260217472,
      "grad_norm": 0.8516616821289062,
      "learning_rate": 1.4017071380099132e-05,
      "loss": 1.1585,
      "step": 5770
    },
    {
      "epoch": 0.8668916385451818,
      "grad_norm": 0.8456748127937317,
      "learning_rate": 1.3711105889213908e-05,
      "loss": 1.1606,
      "step": 5780
    },
    {
      "epoch": 0.8683914510686164,
      "grad_norm": 0.8640986680984497,
      "learning_rate": 1.3408356676750043e-05,
      "loss": 1.1806,
      "step": 5790
    },
    {
      "epoch": 0.869891263592051,
      "grad_norm": 0.8690096735954285,
      "learning_rate": 1.310883088727902e-05,
      "loss": 1.1607,
      "step": 5800
    },
    {
      "epoch": 0.8713910761154856,
      "grad_norm": 0.8273411393165588,
      "learning_rate": 1.2812535589303024e-05,
      "loss": 1.1436,
      "step": 5810
    },
    {
      "epoch": 0.8728908886389202,
      "grad_norm": 0.7910122871398926,
      "learning_rate": 1.2519477775087805e-05,
      "loss": 1.1706,
      "step": 5820
    },
    {
      "epoch": 0.8743907011623547,
      "grad_norm": 0.9277138710021973,
      "learning_rate": 1.222966436049786e-05,
      "loss": 1.1801,
      "step": 5830
    },
    {
      "epoch": 0.8758905136857893,
      "grad_norm": 0.861088752746582,
      "learning_rate": 1.1943102184833165e-05,
      "loss": 1.1763,
      "step": 5840
    },
    {
      "epoch": 0.8773903262092239,
      "grad_norm": 0.8729887008666992,
      "learning_rate": 1.165979801066782e-05,
      "loss": 1.1571,
      "step": 5850
    },
    {
      "epoch": 0.8788901387326584,
      "grad_norm": 0.9203227162361145,
      "learning_rate": 1.1379758523690413e-05,
      "loss": 1.1756,
      "step": 5860
    },
    {
      "epoch": 0.880389951256093,
      "grad_norm": 0.8871293663978577,
      "learning_rate": 1.1102990332546175e-05,
      "loss": 1.1578,
      "step": 5870
    },
    {
      "epoch": 0.8818897637795275,
      "grad_norm": 0.8668881058692932,
      "learning_rate": 1.0829499968681204e-05,
      "loss": 1.1618,
      "step": 5880
    },
    {
      "epoch": 0.8833895763029621,
      "grad_norm": 0.8996675610542297,
      "learning_rate": 1.0559293886188246e-05,
      "loss": 1.1723,
      "step": 5890
    },
    {
      "epoch": 0.8848893888263967,
      "grad_norm": 0.892325758934021,
      "learning_rate": 1.029237846165426e-05,
      "loss": 1.1645,
      "step": 5900
    },
    {
      "epoch": 0.8863892013498312,
      "grad_norm": 0.8483745455741882,
      "learning_rate": 1.0028759994010071e-05,
      "loss": 1.1685,
      "step": 5910
    },
    {
      "epoch": 0.8878890138732658,
      "grad_norm": 0.8226345181465149,
      "learning_rate": 9.768444704381811e-06,
      "loss": 1.156,
      "step": 5920
    },
    {
      "epoch": 0.8893888263967004,
      "grad_norm": 0.9334031343460083,
      "learning_rate": 9.511438735943849e-06,
      "loss": 1.1732,
      "step": 5930
    },
    {
      "epoch": 0.890888638920135,
      "grad_norm": 0.9309195876121521,
      "learning_rate": 9.257748153773992e-06,
      "loss": 1.158,
      "step": 5940
    },
    {
      "epoch": 0.8923884514435696,
      "grad_norm": 0.899749219417572,
      "learning_rate": 9.007378944710431e-06,
      "loss": 1.1512,
      "step": 5950
    },
    {
      "epoch": 0.8938882639670042,
      "grad_norm": 0.8760167956352234,
      "learning_rate": 8.760337017210206e-06,
      "loss": 1.1453,
      "step": 5960
    },
    {
      "epoch": 0.8953880764904387,
      "grad_norm": 0.9029643535614014,
      "learning_rate": 8.516628201209985e-06,
      "loss": 1.1561,
      "step": 5970
    },
    {
      "epoch": 0.8968878890138733,
      "grad_norm": 0.8686094284057617,
      "learning_rate": 8.276258247988437e-06,
      "loss": 1.1569,
      "step": 5980
    },
    {
      "epoch": 0.8983877015373078,
      "grad_norm": 0.876531720161438,
      "learning_rate": 8.039232830030413e-06,
      "loss": 1.1651,
      "step": 5990
    },
    {
      "epoch": 0.8998875140607424,
      "grad_norm": 0.896596372127533,
      "learning_rate": 7.805557540893276e-06,
      "loss": 1.1709,
      "step": 6000
    },
    {
      "epoch": 0.9003374578177727,
      "eval_loss": 1.203166127204895,
      "eval_runtime": 35.8498,
      "eval_samples_per_second": 697.355,
      "eval_steps_per_second": 87.169,
      "step": 6003
    },
    {
      "epoch": 0.901387326584177,
      "grad_norm": 0.8913058638572693,
      "learning_rate": 7.575237895074637e-06,
      "loss": 1.1691,
      "step": 6010
    },
    {
      "epoch": 0.9028871391076115,
      "grad_norm": 0.8780670166015625,
      "learning_rate": 7.348279327882467e-06,
      "loss": 1.1685,
      "step": 6020
    },
    {
      "epoch": 0.9043869516310461,
      "grad_norm": 0.828803539276123,
      "learning_rate": 7.1246871953066666e-06,
      "loss": 1.1532,
      "step": 6030
    },
    {
      "epoch": 0.9058867641544807,
      "grad_norm": 0.8630168437957764,
      "learning_rate": 6.9044667738927365e-06,
      "loss": 1.1641,
      "step": 6040
    },
    {
      "epoch": 0.9073865766779152,
      "grad_norm": 0.8399310111999512,
      "learning_rate": 6.6876232606172255e-06,
      "loss": 1.1596,
      "step": 6050
    },
    {
      "epoch": 0.9088863892013498,
      "grad_norm": 0.8821493983268738,
      "learning_rate": 6.4741617727651626e-06,
      "loss": 1.1501,
      "step": 6060
    },
    {
      "epoch": 0.9103862017247843,
      "grad_norm": 0.948297381401062,
      "learning_rate": 6.264087347809188e-06,
      "loss": 1.1734,
      "step": 6070
    },
    {
      "epoch": 0.911886014248219,
      "grad_norm": 0.9395800232887268,
      "learning_rate": 6.0574049432907115e-06,
      "loss": 1.1738,
      "step": 6080
    },
    {
      "epoch": 0.9133858267716536,
      "grad_norm": 0.8580440878868103,
      "learning_rate": 5.854119436702976e-06,
      "loss": 1.1677,
      "step": 6090
    },
    {
      "epoch": 0.9148856392950881,
      "grad_norm": 0.9488980770111084,
      "learning_rate": 5.65423562537593e-06,
      "loss": 1.1638,
      "step": 6100
    },
    {
      "epoch": 0.9163854518185227,
      "grad_norm": 0.8783366680145264,
      "learning_rate": 5.4577582263629235e-06,
      "loss": 1.1676,
      "step": 6110
    },
    {
      "epoch": 0.9178852643419573,
      "grad_norm": 0.8629846572875977,
      "learning_rate": 5.264691876329474e-06,
      "loss": 1.1426,
      "step": 6120
    },
    {
      "epoch": 0.9193850768653918,
      "grad_norm": 0.8540226221084595,
      "learning_rate": 5.075041131443891e-06,
      "loss": 1.1582,
      "step": 6130
    },
    {
      "epoch": 0.9208848893888264,
      "grad_norm": 0.8809103965759277,
      "learning_rate": 4.88881046726966e-06,
      "loss": 1.1514,
      "step": 6140
    },
    {
      "epoch": 0.922384701912261,
      "grad_norm": 0.9011795520782471,
      "learning_rate": 4.706004278659831e-06,
      "loss": 1.1543,
      "step": 6150
    },
    {
      "epoch": 0.9238845144356955,
      "grad_norm": 0.8636093139648438,
      "learning_rate": 4.526626879653428e-06,
      "loss": 1.1694,
      "step": 6160
    },
    {
      "epoch": 0.9253843269591301,
      "grad_norm": 0.8122191429138184,
      "learning_rate": 4.350682503373437e-06,
      "loss": 1.1508,
      "step": 6170
    },
    {
      "epoch": 0.9268841394825647,
      "grad_norm": 0.9113324880599976,
      "learning_rate": 4.178175301927101e-06,
      "loss": 1.1767,
      "step": 6180
    },
    {
      "epoch": 0.9283839520059992,
      "grad_norm": 0.8934022784233093,
      "learning_rate": 4.009109346307792e-06,
      "loss": 1.162,
      "step": 6190
    },
    {
      "epoch": 0.9298837645294338,
      "grad_norm": 0.8794459700584412,
      "learning_rate": 3.8434886262991015e-06,
      "loss": 1.167,
      "step": 6200
    },
    {
      "epoch": 0.9313835770528683,
      "grad_norm": 0.9140746593475342,
      "learning_rate": 3.6813170503804834e-06,
      "loss": 1.1828,
      "step": 6210
    },
    {
      "epoch": 0.932883389576303,
      "grad_norm": 0.8573930859565735,
      "learning_rate": 3.522598445635172e-06,
      "loss": 1.138,
      "step": 6220
    },
    {
      "epoch": 0.9343832020997376,
      "grad_norm": 0.8401947021484375,
      "learning_rate": 3.3673365576598e-06,
      "loss": 1.1599,
      "step": 6230
    },
    {
      "epoch": 0.9358830146231721,
      "grad_norm": 0.9218401908874512,
      "learning_rate": 3.21553505047602e-06,
      "loss": 1.1699,
      "step": 6240
    },
    {
      "epoch": 0.9373828271466067,
      "grad_norm": 0.9082098603248596,
      "learning_rate": 3.067197506444058e-06,
      "loss": 1.1595,
      "step": 6250
    },
    {
      "epoch": 0.9388826396700413,
      "grad_norm": 0.9707618355751038,
      "learning_rate": 2.922327426178128e-06,
      "loss": 1.1417,
      "step": 6260
    },
    {
      "epoch": 0.9403824521934758,
      "grad_norm": 0.8763731718063354,
      "learning_rate": 2.7809282284638855e-06,
      "loss": 1.1839,
      "step": 6270
    },
    {
      "epoch": 0.9418822647169104,
      "grad_norm": 0.9058078527450562,
      "learning_rate": 2.643003250177672e-06,
      "loss": 1.147,
      "step": 6280
    },
    {
      "epoch": 0.943382077240345,
      "grad_norm": 0.817454993724823,
      "learning_rate": 2.5085557462078134e-06,
      "loss": 1.1457,
      "step": 6290
    },
    {
      "epoch": 0.9448818897637795,
      "grad_norm": 0.919282853603363,
      "learning_rate": 2.377588889377813e-06,
      "loss": 1.1738,
      "step": 6300
    },
    {
      "epoch": 0.9463817022872141,
      "grad_norm": 0.9078910946846008,
      "learning_rate": 2.2501057703714797e-06,
      "loss": 1.175,
      "step": 6310
    },
    {
      "epoch": 0.9478815148106486,
      "grad_norm": 0.894939661026001,
      "learning_rate": 2.1261093976599365e-06,
      "loss": 1.1704,
      "step": 6320
    },
    {
      "epoch": 0.9493813273340832,
      "grad_norm": 0.8727539777755737,
      "learning_rate": 2.005602697430675e-06,
      "loss": 1.1619,
      "step": 6330
    },
    {
      "epoch": 0.9508811398575178,
      "grad_norm": 0.8867236971855164,
      "learning_rate": 1.8885885135184963e-06,
      "loss": 1.1693,
      "step": 6340
    },
    {
      "epoch": 0.9523809523809523,
      "grad_norm": 0.8786768317222595,
      "learning_rate": 1.7750696073383974e-06,
      "loss": 1.1494,
      "step": 6350
    },
    {
      "epoch": 0.953880764904387,
      "grad_norm": 0.8469645380973816,
      "learning_rate": 1.6650486578203725e-06,
      "loss": 1.1619,
      "step": 6360
    },
    {
      "epoch": 0.9553805774278216,
      "grad_norm": 0.8775497674942017,
      "learning_rate": 1.558528261346248e-06,
      "loss": 1.1448,
      "step": 6370
    },
    {
      "epoch": 0.9568803899512561,
      "grad_norm": 0.9189411997795105,
      "learning_rate": 1.455510931688364e-06,
      "loss": 1.1539,
      "step": 6380
    },
    {
      "epoch": 0.9583802024746907,
      "grad_norm": 0.8403207659721375,
      "learning_rate": 1.3559990999502556e-06,
      "loss": 1.1644,
      "step": 6390
    },
    {
      "epoch": 0.9598800149981253,
      "grad_norm": 0.910416305065155,
      "learning_rate": 1.2599951145093157e-06,
      "loss": 1.1549,
      "step": 6400
    },
    {
      "epoch": 0.9613798275215598,
      "grad_norm": 0.8597015738487244,
      "learning_rate": 1.1675012409613715e-06,
      "loss": 1.1502,
      "step": 6410
    },
    {
      "epoch": 0.9628796400449944,
      "grad_norm": 0.8848598003387451,
      "learning_rate": 1.0785196620671455e-06,
      "loss": 1.1582,
      "step": 6420
    },
    {
      "epoch": 0.9643794525684289,
      "grad_norm": 0.8559622764587402,
      "learning_rate": 9.93052477700862e-07,
      "loss": 1.1679,
      "step": 6430
    },
    {
      "epoch": 0.9658792650918635,
      "grad_norm": 0.814378559589386,
      "learning_rate": 9.111017048005876e-07,
      "loss": 1.1639,
      "step": 6440
    },
    {
      "epoch": 0.9673790776152981,
      "grad_norm": 0.7913538217544556,
      "learning_rate": 8.326692773207189e-07,
      "loss": 1.1621,
      "step": 6450
    },
    {
      "epoch": 0.9688788901387326,
      "grad_norm": 0.947375476360321,
      "learning_rate": 7.577570461862359e-07,
      "loss": 1.158,
      "step": 6460
    },
    {
      "epoch": 0.9703787026621672,
      "grad_norm": 0.8357100486755371,
      "learning_rate": 6.863667792491534e-07,
      "loss": 1.1584,
      "step": 6470
    },
    {
      "epoch": 0.9718785151856018,
      "grad_norm": 0.880916953086853,
      "learning_rate": 6.185001612467044e-07,
      "loss": 1.1671,
      "step": 6480
    },
    {
      "epoch": 0.9733783277090363,
      "grad_norm": 0.8177213072776794,
      "learning_rate": 5.541587937616221e-07,
      "loss": 1.161,
      "step": 6490
    },
    {
      "epoch": 0.974878140232471,
      "grad_norm": 0.8296107053756714,
      "learning_rate": 4.933441951843198e-07,
      "loss": 1.1555,
      "step": 6500
    },
    {
      "epoch": 0.9763779527559056,
      "grad_norm": 0.8631340861320496,
      "learning_rate": 4.360578006770865e-07,
      "loss": 1.1624,
      "step": 6510
    },
    {
      "epoch": 0.9778777652793401,
      "grad_norm": 0.8406107425689697,
      "learning_rate": 3.82300962140214e-07,
      "loss": 1.1615,
      "step": 6520
    },
    {
      "epoch": 0.9793775778027747,
      "grad_norm": 0.9254620671272278,
      "learning_rate": 3.320749481800888e-07,
      "loss": 1.1597,
      "step": 6530
    },
    {
      "epoch": 0.9808773903262092,
      "grad_norm": 0.9091020822525024,
      "learning_rate": 2.8538094407919987e-07,
      "loss": 1.1537,
      "step": 6540
    },
    {
      "epoch": 0.9823772028496438,
      "grad_norm": 0.8602820634841919,
      "learning_rate": 2.4222005176829375e-07,
      "loss": 1.1485,
      "step": 6550
    },
    {
      "epoch": 0.9838770153730784,
      "grad_norm": 0.9516273736953735,
      "learning_rate": 2.025932898002458e-07,
      "loss": 1.1706,
      "step": 6560
    },
    {
      "epoch": 0.9853768278965129,
      "grad_norm": 0.8547878265380859,
      "learning_rate": 1.6650159332607939e-07,
      "loss": 1.1511,
      "step": 6570
    },
    {
      "epoch": 0.9868766404199475,
      "grad_norm": 0.9445357918739319,
      "learning_rate": 1.3394581407289996e-07,
      "loss": 1.1735,
      "step": 6580
    },
    {
      "epoch": 0.9883764529433821,
      "grad_norm": 0.9283078908920288,
      "learning_rate": 1.0492672032377803e-07,
      "loss": 1.1665,
      "step": 6590
    },
    {
      "epoch": 0.9898762654668166,
      "grad_norm": 0.9254633784294128,
      "learning_rate": 7.944499689961358e-08,
      "loss": 1.1533,
      "step": 6600
    },
    {
      "epoch": 0.9913760779902512,
      "grad_norm": 0.9120994806289673,
      "learning_rate": 5.7501245143015685e-08,
      "loss": 1.1616,
      "step": 6610
    },
    {
      "epoch": 0.9928758905136857,
      "grad_norm": 0.8868552446365356,
      "learning_rate": 3.9095982904080447e-08,
      "loss": 1.1591,
      "step": 6620
    },
    {
      "epoch": 0.9943757030371203,
      "grad_norm": 0.8874196410179138,
      "learning_rate": 2.4229644528150905e-08,
      "loss": 1.168,
      "step": 6630
    },
    {
      "epoch": 0.995875515560555,
      "grad_norm": 0.8636330366134644,
      "learning_rate": 1.290258084557516e-08,
      "loss": 1.1572,
      "step": 6640
    },
    {
      "epoch": 0.9973753280839895,
      "grad_norm": 0.8626487255096436,
      "learning_rate": 5.115059163496304e-09,
      "loss": 1.1482,
      "step": 6650
    },
    {
      "epoch": 0.9988751406074241,
      "grad_norm": 0.9289808869361877,
      "learning_rate": 8.672632594408646e-10,
      "loss": 1.1632,
      "step": 6660
    }
  ],
  "logging_steps": 10,
  "max_steps": 6667,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 667,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2426610074517504.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}