{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 198,
  "global_step": 1980,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0010108668182966893,
      "grad_norm": 11.463725090026855,
      "learning_rate": 0.0,
      "loss": 0.27058643102645874,
      "num_input_tokens_seen": 2014,
      "step": 1,
      "train_runtime": 30.365,
      "train_tokens_per_second": 66.326
    },
    {
      "epoch": 0.0020217336365933787,
      "grad_norm": 9.243507385253906,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.2549106180667877,
      "num_input_tokens_seen": 4302,
      "step": 2,
      "train_runtime": 32.7112,
      "train_tokens_per_second": 131.515
    },
    {
      "epoch": 0.003032600454890068,
      "grad_norm": 8.196733474731445,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.22403164207935333,
      "num_input_tokens_seen": 6256,
      "step": 3,
      "train_runtime": 35.0243,
      "train_tokens_per_second": 178.619
    },
    {
      "epoch": 0.004043467273186757,
      "grad_norm": 5.648019313812256,
      "learning_rate": 1.2e-05,
      "loss": 0.16279545426368713,
      "num_input_tokens_seen": 8140,
      "step": 4,
      "train_runtime": 37.2324,
      "train_tokens_per_second": 218.627
    },
    {
      "epoch": 0.005054334091483447,
      "grad_norm": 4.784319877624512,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 0.16342008113861084,
      "num_input_tokens_seen": 9800,
      "step": 5,
      "train_runtime": 39.4406,
      "train_tokens_per_second": 248.475
    },
    {
      "epoch": 0.006065200909780136,
      "grad_norm": 4.195488929748535,
      "learning_rate": 2e-05,
      "loss": 0.18118953704833984,
      "num_input_tokens_seen": 11810,
      "step": 6,
      "train_runtime": 41.6807,
      "train_tokens_per_second": 283.345
    },
    {
      "epoch": 0.0070760677280768255,
      "grad_norm": 3.3543472290039062,
      "learning_rate": 1.998987341772152e-05,
      "loss": 0.15088146924972534,
      "num_input_tokens_seen": 13386,
      "step": 7,
      "train_runtime": 43.8323,
      "train_tokens_per_second": 305.391
    },
    {
      "epoch": 0.008086934546373515,
      "grad_norm": 3.685063362121582,
      "learning_rate": 1.997974683544304e-05,
      "loss": 0.12043247371912003,
      "num_input_tokens_seen": 15448,
      "step": 8,
      "train_runtime": 46.1166,
      "train_tokens_per_second": 334.977
    },
    {
      "epoch": 0.009097801364670205,
      "grad_norm": 2.9997713565826416,
      "learning_rate": 1.996962025316456e-05,
      "loss": 0.136856809258461,
      "num_input_tokens_seen": 17280,
      "step": 9,
      "train_runtime": 48.3393,
      "train_tokens_per_second": 357.473
    },
    {
      "epoch": 0.010108668182966895,
      "grad_norm": 2.9768412113189697,
      "learning_rate": 1.9959493670886078e-05,
      "loss": 0.15534675121307373,
      "num_input_tokens_seen": 19228,
      "step": 10,
      "train_runtime": 50.6127,
      "train_tokens_per_second": 379.905
    },
    {
      "epoch": 0.011119535001263583,
      "grad_norm": 2.748751163482666,
      "learning_rate": 1.9949367088607597e-05,
      "loss": 0.144414022564888,
      "num_input_tokens_seen": 21282,
      "step": 11,
      "train_runtime": 52.9193,
      "train_tokens_per_second": 402.16
    },
    {
      "epoch": 0.012130401819560273,
      "grad_norm": 2.0463802814483643,
      "learning_rate": 1.9939240506329116e-05,
      "loss": 0.11977019906044006,
      "num_input_tokens_seen": 23866,
      "step": 12,
      "train_runtime": 55.2931,
      "train_tokens_per_second": 431.627
    },
    {
      "epoch": 0.013141268637856963,
      "grad_norm": 3.2663321495056152,
      "learning_rate": 1.9929113924050635e-05,
      "loss": 0.17596785724163055,
      "num_input_tokens_seen": 25162,
      "step": 13,
      "train_runtime": 57.3569,
      "train_tokens_per_second": 438.692
    },
    {
      "epoch": 0.014152135456153651,
      "grad_norm": 2.8759734630584717,
      "learning_rate": 1.9918987341772154e-05,
      "loss": 0.1420745849609375,
      "num_input_tokens_seen": 27000,
      "step": 14,
      "train_runtime": 59.5626,
      "train_tokens_per_second": 453.305
    },
    {
      "epoch": 0.015163002274450341,
      "grad_norm": 2.3515286445617676,
      "learning_rate": 1.9908860759493673e-05,
      "loss": 0.1690654456615448,
      "num_input_tokens_seen": 29284,
      "step": 15,
      "train_runtime": 61.9399,
      "train_tokens_per_second": 472.781
    },
    {
      "epoch": 0.01617386909274703,
      "grad_norm": 2.069561243057251,
      "learning_rate": 1.9898734177215192e-05,
      "loss": 0.11181537806987762,
      "num_input_tokens_seen": 30880,
      "step": 16,
      "train_runtime": 64.1141,
      "train_tokens_per_second": 481.642
    },
    {
      "epoch": 0.01718473591104372,
      "grad_norm": 2.4530253410339355,
      "learning_rate": 1.988860759493671e-05,
      "loss": 0.14298129081726074,
      "num_input_tokens_seen": 32636,
      "step": 17,
      "train_runtime": 66.333,
      "train_tokens_per_second": 492.002
    },
    {
      "epoch": 0.01819560272934041,
      "grad_norm": 2.4296536445617676,
      "learning_rate": 1.9878481012658227e-05,
      "loss": 0.1248965710401535,
      "num_input_tokens_seen": 34086,
      "step": 18,
      "train_runtime": 68.5037,
      "train_tokens_per_second": 497.579
    },
    {
      "epoch": 0.0192064695476371,
      "grad_norm": 2.3326025009155273,
      "learning_rate": 1.986835443037975e-05,
      "loss": 0.14833924174308777,
      "num_input_tokens_seen": 35890,
      "step": 19,
      "train_runtime": 70.7199,
      "train_tokens_per_second": 507.495
    },
    {
      "epoch": 0.02021733636593379,
      "grad_norm": 2.302140712738037,
      "learning_rate": 1.985822784810127e-05,
      "loss": 0.13441942632198334,
      "num_input_tokens_seen": 37808,
      "step": 20,
      "train_runtime": 73.1196,
      "train_tokens_per_second": 517.071
    },
    {
      "epoch": 0.02122820318423048,
      "grad_norm": 2.6388118267059326,
      "learning_rate": 1.9848101265822785e-05,
      "loss": 0.1305861622095108,
      "num_input_tokens_seen": 39772,
      "step": 21,
      "train_runtime": 75.3868,
      "train_tokens_per_second": 527.572
    },
    {
      "epoch": 0.022239070002527166,
      "grad_norm": 2.6078476905822754,
      "learning_rate": 1.9837974683544307e-05,
      "loss": 0.1466171145439148,
      "num_input_tokens_seen": 41968,
      "step": 22,
      "train_runtime": 77.6477,
      "train_tokens_per_second": 540.492
    },
    {
      "epoch": 0.023249936820823856,
      "grad_norm": 2.4036481380462646,
      "learning_rate": 1.9827848101265826e-05,
      "loss": 0.14058755338191986,
      "num_input_tokens_seen": 43884,
      "step": 23,
      "train_runtime": 79.8842,
      "train_tokens_per_second": 549.345
    },
    {
      "epoch": 0.024260803639120546,
      "grad_norm": 2.364380359649658,
      "learning_rate": 1.9817721518987342e-05,
      "loss": 0.1120375543832779,
      "num_input_tokens_seen": 45820,
      "step": 24,
      "train_runtime": 82.0906,
      "train_tokens_per_second": 558.163
    },
    {
      "epoch": 0.025271670457417236,
      "grad_norm": 2.6477315425872803,
      "learning_rate": 1.9807594936708864e-05,
      "loss": 0.14213895797729492,
      "num_input_tokens_seen": 47302,
      "step": 25,
      "train_runtime": 84.253,
      "train_tokens_per_second": 561.428
    },
    {
      "epoch": 0.026282537275713926,
      "grad_norm": 2.5072643756866455,
      "learning_rate": 1.979746835443038e-05,
      "loss": 0.13167902827262878,
      "num_input_tokens_seen": 48978,
      "step": 26,
      "train_runtime": 86.3872,
      "train_tokens_per_second": 566.959
    },
    {
      "epoch": 0.027293404094010616,
      "grad_norm": 2.603252649307251,
      "learning_rate": 1.97873417721519e-05,
      "loss": 0.1659485399723053,
      "num_input_tokens_seen": 50812,
      "step": 27,
      "train_runtime": 88.5512,
      "train_tokens_per_second": 573.815
    },
    {
      "epoch": 0.028304270912307302,
      "grad_norm": 2.4532132148742676,
      "learning_rate": 1.977721518987342e-05,
      "loss": 0.12353596091270447,
      "num_input_tokens_seen": 52666,
      "step": 28,
      "train_runtime": 90.7873,
      "train_tokens_per_second": 580.103
    },
    {
      "epoch": 0.029315137730603992,
      "grad_norm": 3.3822824954986572,
      "learning_rate": 1.9767088607594937e-05,
      "loss": 0.16516587138175964,
      "num_input_tokens_seen": 54330,
      "step": 29,
      "train_runtime": 92.9164,
      "train_tokens_per_second": 584.719
    },
    {
      "epoch": 0.030326004548900682,
      "grad_norm": 2.7512779235839844,
      "learning_rate": 1.9756962025316456e-05,
      "loss": 0.13730409741401672,
      "num_input_tokens_seen": 55990,
      "step": 30,
      "train_runtime": 95.0721,
      "train_tokens_per_second": 588.921
    },
    {
      "epoch": 0.03133687136719737,
      "grad_norm": 2.2253849506378174,
      "learning_rate": 1.974683544303798e-05,
      "loss": 0.10745520144701004,
      "num_input_tokens_seen": 57646,
      "step": 31,
      "train_runtime": 103.8486,
      "train_tokens_per_second": 555.096
    },
    {
      "epoch": 0.03234773818549406,
      "grad_norm": 2.0792880058288574,
      "learning_rate": 1.9736708860759494e-05,
      "loss": 0.11148789525032043,
      "num_input_tokens_seen": 59978,
      "step": 32,
      "train_runtime": 106.1404,
      "train_tokens_per_second": 565.082
    },
    {
      "epoch": 0.03335860500379075,
      "grad_norm": 2.7239630222320557,
      "learning_rate": 1.9726582278481014e-05,
      "loss": 0.1380208283662796,
      "num_input_tokens_seen": 61296,
      "step": 33,
      "train_runtime": 108.2579,
      "train_tokens_per_second": 566.204
    },
    {
      "epoch": 0.03436947182208744,
      "grad_norm": 2.196681499481201,
      "learning_rate": 1.9716455696202533e-05,
      "loss": 0.14132940769195557,
      "num_input_tokens_seen": 63650,
      "step": 34,
      "train_runtime": 110.5755,
      "train_tokens_per_second": 575.625
    },
    {
      "epoch": 0.03538033864038413,
      "grad_norm": 2.128329277038574,
      "learning_rate": 1.970632911392405e-05,
      "loss": 0.14076243340969086,
      "num_input_tokens_seen": 65694,
      "step": 35,
      "train_runtime": 112.8447,
      "train_tokens_per_second": 582.163
    },
    {
      "epoch": 0.03639120545868082,
      "grad_norm": 2.418405055999756,
      "learning_rate": 1.969620253164557e-05,
      "loss": 0.11391866207122803,
      "num_input_tokens_seen": 67306,
      "step": 36,
      "train_runtime": 115.0694,
      "train_tokens_per_second": 584.917
    },
    {
      "epoch": 0.037402072276977505,
      "grad_norm": 2.734203338623047,
      "learning_rate": 1.968607594936709e-05,
      "loss": 0.16631358861923218,
      "num_input_tokens_seen": 69350,
      "step": 37,
      "train_runtime": 117.2859,
      "train_tokens_per_second": 591.29
    },
    {
      "epoch": 0.0384129390952742,
      "grad_norm": 2.8153579235076904,
      "learning_rate": 1.967594936708861e-05,
      "loss": 0.1615174114704132,
      "num_input_tokens_seen": 71108,
      "step": 38,
      "train_runtime": 119.4274,
      "train_tokens_per_second": 595.408
    },
    {
      "epoch": 0.039423805913570885,
      "grad_norm": 2.186095952987671,
      "learning_rate": 1.9665822784810128e-05,
      "loss": 0.142922043800354,
      "num_input_tokens_seen": 72894,
      "step": 39,
      "train_runtime": 121.6064,
      "train_tokens_per_second": 599.426
    },
    {
      "epoch": 0.04043467273186758,
      "grad_norm": 1.8680264949798584,
      "learning_rate": 1.9655696202531647e-05,
      "loss": 0.11134858429431915,
      "num_input_tokens_seen": 74832,
      "step": 40,
      "train_runtime": 124.0373,
      "train_tokens_per_second": 603.302
    },
    {
      "epoch": 0.041445539550164265,
      "grad_norm": 2.2168729305267334,
      "learning_rate": 1.9645569620253166e-05,
      "loss": 0.14616534113883972,
      "num_input_tokens_seen": 76986,
      "step": 41,
      "train_runtime": 126.2916,
      "train_tokens_per_second": 609.589
    },
    {
      "epoch": 0.04245640636846096,
      "grad_norm": 3.064636707305908,
      "learning_rate": 1.9635443037974685e-05,
      "loss": 0.12268605083227158,
      "num_input_tokens_seen": 78728,
      "step": 42,
      "train_runtime": 128.4709,
      "train_tokens_per_second": 612.808
    },
    {
      "epoch": 0.043467273186757645,
      "grad_norm": 1.8882980346679688,
      "learning_rate": 1.9625316455696204e-05,
      "loss": 0.09156373143196106,
      "num_input_tokens_seen": 80352,
      "step": 43,
      "train_runtime": 130.6461,
      "train_tokens_per_second": 615.035
    },
    {
      "epoch": 0.04447814000505433,
      "grad_norm": 2.0294580459594727,
      "learning_rate": 1.9615189873417723e-05,
      "loss": 0.11838850378990173,
      "num_input_tokens_seen": 82186,
      "step": 44,
      "train_runtime": 132.8638,
      "train_tokens_per_second": 618.573
    },
    {
      "epoch": 0.045489006823351025,
      "grad_norm": 2.3659250736236572,
      "learning_rate": 1.9605063291139243e-05,
      "loss": 0.1352626085281372,
      "num_input_tokens_seen": 84134,
      "step": 45,
      "train_runtime": 135.0895,
      "train_tokens_per_second": 622.802
    },
    {
      "epoch": 0.04649987364164771,
      "grad_norm": 1.9668437242507935,
      "learning_rate": 1.959493670886076e-05,
      "loss": 0.13398335874080658,
      "num_input_tokens_seen": 86412,
      "step": 46,
      "train_runtime": 137.3768,
      "train_tokens_per_second": 629.014
    },
    {
      "epoch": 0.047510740459944405,
      "grad_norm": 2.4079082012176514,
      "learning_rate": 1.958481012658228e-05,
      "loss": 0.12897895276546478,
      "num_input_tokens_seen": 88426,
      "step": 47,
      "train_runtime": 139.6247,
      "train_tokens_per_second": 633.312
    },
    {
      "epoch": 0.04852160727824109,
      "grad_norm": 2.137667655944824,
      "learning_rate": 1.95746835443038e-05,
      "loss": 0.12803080677986145,
      "num_input_tokens_seen": 90310,
      "step": 48,
      "train_runtime": 141.8482,
      "train_tokens_per_second": 636.667
    },
    {
      "epoch": 0.04953247409653778,
      "grad_norm": 2.125237226486206,
      "learning_rate": 1.956455696202532e-05,
      "loss": 0.12782883644104004,
      "num_input_tokens_seen": 92530,
      "step": 49,
      "train_runtime": 144.1353,
      "train_tokens_per_second": 641.966
    },
    {
      "epoch": 0.05054334091483447,
      "grad_norm": 1.9926167726516724,
      "learning_rate": 1.9554430379746835e-05,
      "loss": 0.11848489195108414,
      "num_input_tokens_seen": 94718,
      "step": 50,
      "train_runtime": 146.4533,
      "train_tokens_per_second": 646.745
    },
    {
      "epoch": 0.05155420773313116,
      "grad_norm": 2.0517542362213135,
      "learning_rate": 1.9544303797468357e-05,
      "loss": 0.11716699600219727,
      "num_input_tokens_seen": 96530,
      "step": 51,
      "train_runtime": 148.6661,
      "train_tokens_per_second": 649.307
    },
    {
      "epoch": 0.05256507455142785,
      "grad_norm": 2.127108573913574,
      "learning_rate": 1.9534177215189876e-05,
      "loss": 0.11858700960874557,
      "num_input_tokens_seen": 98434,
      "step": 52,
      "train_runtime": 150.8912,
      "train_tokens_per_second": 652.351
    },
    {
      "epoch": 0.05357594136972454,
      "grad_norm": 2.1097190380096436,
      "learning_rate": 1.9524050632911392e-05,
      "loss": 0.11988560855388641,
      "num_input_tokens_seen": 100400,
      "step": 53,
      "train_runtime": 153.0729,
      "train_tokens_per_second": 655.897
    },
    {
      "epoch": 0.05458680818802123,
      "grad_norm": 2.1106865406036377,
      "learning_rate": 1.9513924050632914e-05,
      "loss": 0.10115162283182144,
      "num_input_tokens_seen": 101800,
      "step": 54,
      "train_runtime": 155.1902,
      "train_tokens_per_second": 655.969
    },
    {
      "epoch": 0.05559767500631792,
      "grad_norm": 2.2258450984954834,
      "learning_rate": 1.9503797468354433e-05,
      "loss": 0.12345483899116516,
      "num_input_tokens_seen": 103632,
      "step": 55,
      "train_runtime": 157.3676,
      "train_tokens_per_second": 658.535
    },
    {
      "epoch": 0.056608541824614604,
      "grad_norm": 2.647247552871704,
      "learning_rate": 1.949367088607595e-05,
      "loss": 0.15581361949443817,
      "num_input_tokens_seen": 105654,
      "step": 56,
      "train_runtime": 159.6032,
      "train_tokens_per_second": 661.979
    },
    {
      "epoch": 0.0576194086429113,
      "grad_norm": 3.035010814666748,
      "learning_rate": 1.948354430379747e-05,
      "loss": 0.14977982640266418,
      "num_input_tokens_seen": 107452,
      "step": 57,
      "train_runtime": 161.7704,
      "train_tokens_per_second": 664.225
    },
    {
      "epoch": 0.058630275461207984,
      "grad_norm": 1.9288145303726196,
      "learning_rate": 1.9473417721518987e-05,
      "loss": 0.0957871675491333,
      "num_input_tokens_seen": 109702,
      "step": 58,
      "train_runtime": 164.0524,
      "train_tokens_per_second": 668.701
    },
    {
      "epoch": 0.05964114227950468,
      "grad_norm": 2.5870797634124756,
      "learning_rate": 1.9463291139240506e-05,
      "loss": 0.13609547913074493,
      "num_input_tokens_seen": 111458,
      "step": 59,
      "train_runtime": 166.2055,
      "train_tokens_per_second": 670.604
    },
    {
      "epoch": 0.060652009097801364,
      "grad_norm": 2.4269790649414062,
      "learning_rate": 1.945316455696203e-05,
      "loss": 0.11794693022966385,
      "num_input_tokens_seen": 112866,
      "step": 60,
      "train_runtime": 168.2716,
      "train_tokens_per_second": 670.737
    },
    {
      "epoch": 0.06166287591609806,
      "grad_norm": 2.5546016693115234,
      "learning_rate": 1.9443037974683544e-05,
      "loss": 0.15975920855998993,
      "num_input_tokens_seen": 115128,
      "step": 61,
      "train_runtime": 176.8429,
      "train_tokens_per_second": 651.019
    },
    {
      "epoch": 0.06267374273439474,
      "grad_norm": 2.4380621910095215,
      "learning_rate": 1.9432911392405064e-05,
      "loss": 0.12921664118766785,
      "num_input_tokens_seen": 117058,
      "step": 62,
      "train_runtime": 179.0428,
      "train_tokens_per_second": 653.799
    },
    {
      "epoch": 0.06368460955269144,
      "grad_norm": 2.1433351039886475,
      "learning_rate": 1.9422784810126586e-05,
      "loss": 0.11755329370498657,
      "num_input_tokens_seen": 119258,
      "step": 63,
      "train_runtime": 181.3383,
      "train_tokens_per_second": 657.655
    },
    {
      "epoch": 0.06469547637098812,
      "grad_norm": 1.960486888885498,
      "learning_rate": 1.9412658227848102e-05,
      "loss": 0.10061006247997284,
      "num_input_tokens_seen": 121276,
      "step": 64,
      "train_runtime": 183.5726,
      "train_tokens_per_second": 660.643
    },
    {
      "epoch": 0.06570634318928481,
      "grad_norm": 2.0249807834625244,
      "learning_rate": 1.940253164556962e-05,
      "loss": 0.1087707057595253,
      "num_input_tokens_seen": 122996,
      "step": 65,
      "train_runtime": 185.7202,
      "train_tokens_per_second": 662.265
    },
    {
      "epoch": 0.0667172100075815,
      "grad_norm": 1.7995734214782715,
      "learning_rate": 1.939240506329114e-05,
      "loss": 0.12001064419746399,
      "num_input_tokens_seen": 125242,
      "step": 66,
      "train_runtime": 188.0206,
      "train_tokens_per_second": 666.108
    },
    {
      "epoch": 0.0677280768258782,
      "grad_norm": 2.116785764694214,
      "learning_rate": 1.938227848101266e-05,
      "loss": 0.11175717413425446,
      "num_input_tokens_seen": 127580,
      "step": 67,
      "train_runtime": 190.3261,
      "train_tokens_per_second": 670.323
    },
    {
      "epoch": 0.06873894364417488,
      "grad_norm": 2.242112874984741,
      "learning_rate": 1.9372151898734178e-05,
      "loss": 0.1137399822473526,
      "num_input_tokens_seen": 129804,
      "step": 68,
      "train_runtime": 192.5895,
      "train_tokens_per_second": 673.993
    },
    {
      "epoch": 0.06974981046247157,
      "grad_norm": 2.366602897644043,
      "learning_rate": 1.9362025316455697e-05,
      "loss": 0.17228052020072937,
      "num_input_tokens_seen": 131960,
      "step": 69,
      "train_runtime": 194.8769,
      "train_tokens_per_second": 677.145
    },
    {
      "epoch": 0.07076067728076826,
      "grad_norm": 2.067042589187622,
      "learning_rate": 1.9351898734177216e-05,
      "loss": 0.12803220748901367,
      "num_input_tokens_seen": 133790,
      "step": 70,
      "train_runtime": 197.0985,
      "train_tokens_per_second": 678.798
    },
    {
      "epoch": 0.07177154409906494,
      "grad_norm": 1.8924226760864258,
      "learning_rate": 1.9341772151898735e-05,
      "loss": 0.10274340957403183,
      "num_input_tokens_seen": 135344,
      "step": 71,
      "train_runtime": 199.2604,
      "train_tokens_per_second": 679.232
    },
    {
      "epoch": 0.07278241091736164,
      "grad_norm": 2.0558054447174072,
      "learning_rate": 1.9331645569620254e-05,
      "loss": 0.09446454793214798,
      "num_input_tokens_seen": 136974,
      "step": 72,
      "train_runtime": 201.387,
      "train_tokens_per_second": 680.153
    },
    {
      "epoch": 0.07379327773565833,
      "grad_norm": 1.939015507698059,
      "learning_rate": 1.9321518987341774e-05,
      "loss": 0.11452166736125946,
      "num_input_tokens_seen": 138742,
      "step": 73,
      "train_runtime": 203.5894,
      "train_tokens_per_second": 681.48
    },
    {
      "epoch": 0.07480414455395501,
      "grad_norm": 2.055706262588501,
      "learning_rate": 1.9311392405063293e-05,
      "loss": 0.11065662652254105,
      "num_input_tokens_seen": 140352,
      "step": 74,
      "train_runtime": 205.7367,
      "train_tokens_per_second": 682.192
    },
    {
      "epoch": 0.0758150113722517,
      "grad_norm": 1.903029441833496,
      "learning_rate": 1.930126582278481e-05,
      "loss": 0.08838392049074173,
      "num_input_tokens_seen": 142278,
      "step": 75,
      "train_runtime": 207.9342,
      "train_tokens_per_second": 684.245
    },
    {
      "epoch": 0.0768258781905484,
      "grad_norm": 2.230924606323242,
      "learning_rate": 1.929113924050633e-05,
      "loss": 0.11504680663347244,
      "num_input_tokens_seen": 144160,
      "step": 76,
      "train_runtime": 210.1283,
      "train_tokens_per_second": 686.057
    },
    {
      "epoch": 0.07783674500884509,
      "grad_norm": 1.9158000946044922,
      "learning_rate": 1.928101265822785e-05,
      "loss": 0.119988813996315,
      "num_input_tokens_seen": 146444,
      "step": 77,
      "train_runtime": 212.4393,
      "train_tokens_per_second": 689.345
    },
    {
      "epoch": 0.07884761182714177,
      "grad_norm": 2.004648447036743,
      "learning_rate": 1.927088607594937e-05,
      "loss": 0.10192504525184631,
      "num_input_tokens_seen": 148724,
      "step": 78,
      "train_runtime": 214.7221,
      "train_tokens_per_second": 692.635
    },
    {
      "epoch": 0.07985847864543846,
      "grad_norm": 2.0876219272613525,
      "learning_rate": 1.9260759493670888e-05,
      "loss": 0.12379962205886841,
      "num_input_tokens_seen": 150364,
      "step": 79,
      "train_runtime": 216.8343,
      "train_tokens_per_second": 693.451
    },
    {
      "epoch": 0.08086934546373516,
      "grad_norm": 2.624312400817871,
      "learning_rate": 1.9250632911392407e-05,
      "loss": 0.15576514601707458,
      "num_input_tokens_seen": 152144,
      "step": 80,
      "train_runtime": 219.0266,
      "train_tokens_per_second": 694.637
    },
    {
      "epoch": 0.08188021228203184,
      "grad_norm": 2.4926764965057373,
      "learning_rate": 1.9240506329113926e-05,
      "loss": 0.12395478785037994,
      "num_input_tokens_seen": 153538,
      "step": 81,
      "train_runtime": 221.128,
      "train_tokens_per_second": 694.34
    },
    {
      "epoch": 0.08289107910032853,
      "grad_norm": 2.0212905406951904,
      "learning_rate": 1.9230379746835445e-05,
      "loss": 0.11995596438646317,
      "num_input_tokens_seen": 155578,
      "step": 82,
      "train_runtime": 223.3235,
      "train_tokens_per_second": 696.648
    },
    {
      "epoch": 0.08390194591862522,
      "grad_norm": 2.133254051208496,
      "learning_rate": 1.9220253164556964e-05,
      "loss": 0.11523745954036713,
      "num_input_tokens_seen": 157254,
      "step": 83,
      "train_runtime": 225.4919,
      "train_tokens_per_second": 697.382
    },
    {
      "epoch": 0.08491281273692192,
      "grad_norm": 1.5193694829940796,
      "learning_rate": 1.9210126582278483e-05,
      "loss": 0.07533153146505356,
      "num_input_tokens_seen": 158702,
      "step": 84,
      "train_runtime": 227.5792,
      "train_tokens_per_second": 697.348
    },
    {
      "epoch": 0.0859236795552186,
      "grad_norm": 2.067779064178467,
      "learning_rate": 1.9200000000000003e-05,
      "loss": 0.11896681040525436,
      "num_input_tokens_seen": 160346,
      "step": 85,
      "train_runtime": 229.7672,
      "train_tokens_per_second": 697.863
    },
    {
      "epoch": 0.08693454637351529,
      "grad_norm": 2.41981840133667,
      "learning_rate": 1.918987341772152e-05,
      "loss": 0.15256190299987793,
      "num_input_tokens_seen": 162460,
      "step": 86,
      "train_runtime": 231.9979,
      "train_tokens_per_second": 700.265
    },
    {
      "epoch": 0.08794541319181198,
      "grad_norm": 1.9438633918762207,
      "learning_rate": 1.917974683544304e-05,
      "loss": 0.11282727122306824,
      "num_input_tokens_seen": 164650,
      "step": 87,
      "train_runtime": 234.2839,
      "train_tokens_per_second": 702.78
    },
    {
      "epoch": 0.08895628001010866,
      "grad_norm": 2.2633562088012695,
      "learning_rate": 1.916962025316456e-05,
      "loss": 0.1142498329281807,
      "num_input_tokens_seen": 166154,
      "step": 88,
      "train_runtime": 236.3963,
      "train_tokens_per_second": 702.862
    },
    {
      "epoch": 0.08996714682840536,
      "grad_norm": 2.4861838817596436,
      "learning_rate": 1.915949367088608e-05,
      "loss": 0.13935230672359467,
      "num_input_tokens_seen": 167978,
      "step": 89,
      "train_runtime": 238.6332,
      "train_tokens_per_second": 703.917
    },
    {
      "epoch": 0.09097801364670205,
      "grad_norm": 2.1855666637420654,
      "learning_rate": 1.9149367088607595e-05,
      "loss": 0.13527220487594604,
      "num_input_tokens_seen": 169662,
      "step": 90,
      "train_runtime": 240.7981,
      "train_tokens_per_second": 704.582
    },
    {
      "epoch": 0.09198888046499874,
      "grad_norm": 1.859151005744934,
      "learning_rate": 1.9139240506329117e-05,
      "loss": 0.12460088729858398,
      "num_input_tokens_seen": 171702,
      "step": 91,
      "train_runtime": 249.1994,
      "train_tokens_per_second": 689.015
    },
    {
      "epoch": 0.09299974728329542,
      "grad_norm": 1.901003122329712,
      "learning_rate": 1.9129113924050636e-05,
      "loss": 0.10142221301794052,
      "num_input_tokens_seen": 173154,
      "step": 92,
      "train_runtime": 251.3454,
      "train_tokens_per_second": 688.909
    },
    {
      "epoch": 0.09401061410159212,
      "grad_norm": 2.418314218521118,
      "learning_rate": 1.9118987341772152e-05,
      "loss": 0.1140369400382042,
      "num_input_tokens_seen": 175074,
      "step": 93,
      "train_runtime": 253.5251,
      "train_tokens_per_second": 690.559
    },
    {
      "epoch": 0.09502148091988881,
      "grad_norm": 1.9275882244110107,
      "learning_rate": 1.9108860759493674e-05,
      "loss": 0.09216680377721786,
      "num_input_tokens_seen": 177174,
      "step": 94,
      "train_runtime": 255.7891,
      "train_tokens_per_second": 692.656
    },
    {
      "epoch": 0.09603234773818549,
      "grad_norm": 2.2098522186279297,
      "learning_rate": 1.9098734177215193e-05,
      "loss": 0.10576708614826202,
      "num_input_tokens_seen": 178894,
      "step": 95,
      "train_runtime": 258.024,
      "train_tokens_per_second": 693.323
    },
    {
      "epoch": 0.09704321455648218,
      "grad_norm": 2.2304112911224365,
      "learning_rate": 1.908860759493671e-05,
      "loss": 0.10691983997821808,
      "num_input_tokens_seen": 180318,
      "step": 96,
      "train_runtime": 260.0918,
      "train_tokens_per_second": 693.286
    },
    {
      "epoch": 0.09805408137477888,
      "grad_norm": 2.067747116088867,
      "learning_rate": 1.9078481012658228e-05,
      "loss": 0.12733067572116852,
      "num_input_tokens_seen": 182496,
      "step": 97,
      "train_runtime": 262.3363,
      "train_tokens_per_second": 695.657
    },
    {
      "epoch": 0.09906494819307556,
      "grad_norm": 2.1281158924102783,
      "learning_rate": 1.9068354430379747e-05,
      "loss": 0.1434943974018097,
      "num_input_tokens_seen": 184870,
      "step": 98,
      "train_runtime": 264.6362,
      "train_tokens_per_second": 698.582
    },
    {
      "epoch": 0.10007581501137225,
      "grad_norm": 2.0052590370178223,
      "learning_rate": 1.9058227848101266e-05,
      "loss": 0.135469451546669,
      "num_input_tokens_seen": 187088,
      "step": 99,
      "train_runtime": 266.8744,
      "train_tokens_per_second": 701.034
    },
    {
      "epoch": 0.10108668182966894,
      "grad_norm": 2.2457380294799805,
      "learning_rate": 1.9048101265822785e-05,
      "loss": 0.13284868001937866,
      "num_input_tokens_seen": 188710,
      "step": 100,
      "train_runtime": 269.029,
      "train_tokens_per_second": 701.449
    },
    {
      "epoch": 0.10209754864796564,
      "grad_norm": 2.324469804763794,
      "learning_rate": 1.9037974683544304e-05,
      "loss": 0.15497452020645142,
      "num_input_tokens_seen": 190694,
      "step": 101,
      "train_runtime": 271.2959,
      "train_tokens_per_second": 702.9
    },
    {
      "epoch": 0.10310841546626232,
      "grad_norm": 2.3468329906463623,
      "learning_rate": 1.9027848101265824e-05,
      "loss": 0.13803139328956604,
      "num_input_tokens_seen": 192500,
      "step": 102,
      "train_runtime": 273.4579,
      "train_tokens_per_second": 703.948
    },
    {
      "epoch": 0.10411928228455901,
      "grad_norm": 2.050370454788208,
      "learning_rate": 1.9017721518987343e-05,
      "loss": 0.13963785767555237,
      "num_input_tokens_seen": 194224,
      "step": 103,
      "train_runtime": 275.6199,
      "train_tokens_per_second": 704.681
    },
    {
      "epoch": 0.1051301491028557,
      "grad_norm": 1.9895809888839722,
      "learning_rate": 1.9007594936708862e-05,
      "loss": 0.1277201622724533,
      "num_input_tokens_seen": 196598,
      "step": 104,
      "train_runtime": 277.9629,
      "train_tokens_per_second": 707.282
    },
    {
      "epoch": 0.10614101592115238,
      "grad_norm": 1.9515326023101807,
      "learning_rate": 1.899746835443038e-05,
      "loss": 0.1383090615272522,
      "num_input_tokens_seen": 199018,
      "step": 105,
      "train_runtime": 280.2932,
      "train_tokens_per_second": 710.035
    },
    {
      "epoch": 0.10715188273944908,
      "grad_norm": 2.397286891937256,
      "learning_rate": 1.89873417721519e-05,
      "loss": 0.12805454432964325,
      "num_input_tokens_seen": 200604,
      "step": 106,
      "train_runtime": 282.4289,
      "train_tokens_per_second": 710.281
    },
    {
      "epoch": 0.10816274955774577,
      "grad_norm": 2.6728909015655518,
      "learning_rate": 1.897721518987342e-05,
      "loss": 0.13824018836021423,
      "num_input_tokens_seen": 202286,
      "step": 107,
      "train_runtime": 284.5635,
      "train_tokens_per_second": 710.864
    },
    {
      "epoch": 0.10917361637604246,
      "grad_norm": 1.9050586223602295,
      "learning_rate": 1.8967088607594938e-05,
      "loss": 0.1345207244157791,
      "num_input_tokens_seen": 204376,
      "step": 108,
      "train_runtime": 286.8111,
      "train_tokens_per_second": 712.58
    },
    {
      "epoch": 0.11018448319433914,
      "grad_norm": 2.118673324584961,
      "learning_rate": 1.8956962025316457e-05,
      "loss": 0.1226341724395752,
      "num_input_tokens_seen": 206010,
      "step": 109,
      "train_runtime": 288.953,
      "train_tokens_per_second": 712.953
    },
    {
      "epoch": 0.11119535001263584,
      "grad_norm": 2.150815010070801,
      "learning_rate": 1.8946835443037976e-05,
      "loss": 0.16867713630199432,
      "num_input_tokens_seen": 207814,
      "step": 110,
      "train_runtime": 291.0932,
      "train_tokens_per_second": 713.909
    },
    {
      "epoch": 0.11220621683093253,
      "grad_norm": 2.179553747177124,
      "learning_rate": 1.8936708860759495e-05,
      "loss": 0.1012715995311737,
      "num_input_tokens_seen": 209512,
      "step": 111,
      "train_runtime": 293.2347,
      "train_tokens_per_second": 714.486
    },
    {
      "epoch": 0.11321708364922921,
      "grad_norm": 1.9366041421890259,
      "learning_rate": 1.8926582278481014e-05,
      "loss": 0.12508518993854523,
      "num_input_tokens_seen": 211092,
      "step": 112,
      "train_runtime": 295.3477,
      "train_tokens_per_second": 714.724
    },
    {
      "epoch": 0.1142279504675259,
      "grad_norm": 2.329840898513794,
      "learning_rate": 1.8916455696202533e-05,
      "loss": 0.13535994291305542,
      "num_input_tokens_seen": 212812,
      "step": 113,
      "train_runtime": 297.5256,
      "train_tokens_per_second": 715.273
    },
    {
      "epoch": 0.1152388172858226,
      "grad_norm": 2.201944589614868,
      "learning_rate": 1.8906329113924053e-05,
      "loss": 0.14602535963058472,
      "num_input_tokens_seen": 214682,
      "step": 114,
      "train_runtime": 299.704,
      "train_tokens_per_second": 716.313
    },
    {
      "epoch": 0.11624968410411929,
      "grad_norm": 2.1417429447174072,
      "learning_rate": 1.889620253164557e-05,
      "loss": 0.15375256538391113,
      "num_input_tokens_seen": 216826,
      "step": 115,
      "train_runtime": 301.9396,
      "train_tokens_per_second": 718.11
    },
    {
      "epoch": 0.11726055092241597,
      "grad_norm": 2.452090263366699,
      "learning_rate": 1.888607594936709e-05,
      "loss": 0.138523668050766,
      "num_input_tokens_seen": 218824,
      "step": 116,
      "train_runtime": 304.1672,
      "train_tokens_per_second": 719.42
    },
    {
      "epoch": 0.11827141774071266,
      "grad_norm": 2.308642625808716,
      "learning_rate": 1.887594936708861e-05,
      "loss": 0.13310624659061432,
      "num_input_tokens_seen": 220542,
      "step": 117,
      "train_runtime": 306.3643,
      "train_tokens_per_second": 719.869
    },
    {
      "epoch": 0.11928228455900936,
      "grad_norm": 2.0640058517456055,
      "learning_rate": 1.886582278481013e-05,
      "loss": 0.10749287158250809,
      "num_input_tokens_seen": 222116,
      "step": 118,
      "train_runtime": 308.5377,
      "train_tokens_per_second": 719.899
    },
    {
      "epoch": 0.12029315137730603,
      "grad_norm": 1.9837942123413086,
      "learning_rate": 1.8855696202531648e-05,
      "loss": 0.09972156584262848,
      "num_input_tokens_seen": 224040,
      "step": 119,
      "train_runtime": 310.774,
      "train_tokens_per_second": 720.91
    },
    {
      "epoch": 0.12130401819560273,
      "grad_norm": 2.2541520595550537,
      "learning_rate": 1.8845569620253167e-05,
      "loss": 0.12161248922348022,
      "num_input_tokens_seen": 225810,
      "step": 120,
      "train_runtime": 313.0174,
      "train_tokens_per_second": 721.398
    },
    {
      "epoch": 0.12231488501389942,
      "grad_norm": 2.0319716930389404,
      "learning_rate": 1.8835443037974686e-05,
      "loss": 0.08204134553670883,
      "num_input_tokens_seen": 227214,
      "step": 121,
      "train_runtime": 321.393,
      "train_tokens_per_second": 706.966
    },
    {
      "epoch": 0.12332575183219611,
      "grad_norm": 1.7932884693145752,
      "learning_rate": 1.8825316455696202e-05,
      "loss": 0.11774466186761856,
      "num_input_tokens_seen": 229020,
      "step": 122,
      "train_runtime": 323.5627,
      "train_tokens_per_second": 707.807
    },
    {
      "epoch": 0.1243366186504928,
      "grad_norm": 2.3838610649108887,
      "learning_rate": 1.8815189873417724e-05,
      "loss": 0.10385432839393616,
      "num_input_tokens_seen": 230962,
      "step": 123,
      "train_runtime": 325.8582,
      "train_tokens_per_second": 708.781
    },
    {
      "epoch": 0.1253474854687895,
      "grad_norm": 2.066030979156494,
      "learning_rate": 1.8805063291139243e-05,
      "loss": 0.12360858917236328,
      "num_input_tokens_seen": 232580,
      "step": 124,
      "train_runtime": 328.0183,
      "train_tokens_per_second": 709.046
    },
    {
      "epoch": 0.12635835228708617,
      "grad_norm": 2.22601318359375,
      "learning_rate": 1.879493670886076e-05,
      "loss": 0.12216442823410034,
      "num_input_tokens_seen": 234272,
      "step": 125,
      "train_runtime": 330.1693,
      "train_tokens_per_second": 709.551
    },
    {
      "epoch": 0.12736921910538287,
      "grad_norm": 1.7348881959915161,
      "learning_rate": 1.878481012658228e-05,
      "loss": 0.1055573970079422,
      "num_input_tokens_seen": 236512,
      "step": 126,
      "train_runtime": 332.4399,
      "train_tokens_per_second": 711.443
    },
    {
      "epoch": 0.12838008592367955,
      "grad_norm": 2.1512677669525146,
      "learning_rate": 1.87746835443038e-05,
      "loss": 0.1608087718486786,
      "num_input_tokens_seen": 238912,
      "step": 127,
      "train_runtime": 334.7416,
      "train_tokens_per_second": 713.721
    },
    {
      "epoch": 0.12939095274197623,
      "grad_norm": 2.6692278385162354,
      "learning_rate": 1.8764556962025316e-05,
      "loss": 0.14802397787570953,
      "num_input_tokens_seen": 241040,
      "step": 128,
      "train_runtime": 336.9632,
      "train_tokens_per_second": 715.33
    },
    {
      "epoch": 0.13040181956027294,
      "grad_norm": 2.1903557777404785,
      "learning_rate": 1.875443037974684e-05,
      "loss": 0.11743307113647461,
      "num_input_tokens_seen": 242810,
      "step": 129,
      "train_runtime": 339.1213,
      "train_tokens_per_second": 715.997
    },
    {
      "epoch": 0.13141268637856962,
      "grad_norm": 2.448732376098633,
      "learning_rate": 1.8744303797468355e-05,
      "loss": 0.14015504717826843,
      "num_input_tokens_seen": 244636,
      "step": 130,
      "train_runtime": 341.3609,
      "train_tokens_per_second": 716.649
    },
    {
      "epoch": 0.1324235531968663,
      "grad_norm": 2.160809278488159,
      "learning_rate": 1.8734177215189874e-05,
      "loss": 0.11315175145864487,
      "num_input_tokens_seen": 246324,
      "step": 131,
      "train_runtime": 343.5356,
      "train_tokens_per_second": 717.026
    },
    {
      "epoch": 0.133434420015163,
      "grad_norm": 2.168117046356201,
      "learning_rate": 1.8724050632911396e-05,
      "loss": 0.15828710794448853,
      "num_input_tokens_seen": 248122,
      "step": 132,
      "train_runtime": 345.7053,
      "train_tokens_per_second": 717.727
    },
    {
      "epoch": 0.1344452868334597,
      "grad_norm": 1.9508112668991089,
      "learning_rate": 1.8713924050632912e-05,
      "loss": 0.10569241642951965,
      "num_input_tokens_seen": 250064,
      "step": 133,
      "train_runtime": 347.9078,
      "train_tokens_per_second": 718.765
    },
    {
      "epoch": 0.1354561536517564,
      "grad_norm": 1.8830527067184448,
      "learning_rate": 1.870379746835443e-05,
      "loss": 0.10407204926013947,
      "num_input_tokens_seen": 251986,
      "step": 134,
      "train_runtime": 350.1121,
      "train_tokens_per_second": 719.729
    },
    {
      "epoch": 0.13646702047005307,
      "grad_norm": 2.082061529159546,
      "learning_rate": 1.8693670886075953e-05,
      "loss": 0.12073522061109543,
      "num_input_tokens_seen": 253914,
      "step": 135,
      "train_runtime": 352.3735,
      "train_tokens_per_second": 720.582
    },
    {
      "epoch": 0.13747788728834975,
      "grad_norm": 1.9986214637756348,
      "learning_rate": 1.868354430379747e-05,
      "loss": 0.10742796212434769,
      "num_input_tokens_seen": 255648,
      "step": 136,
      "train_runtime": 354.5771,
      "train_tokens_per_second": 720.994
    },
    {
      "epoch": 0.13848875410664646,
      "grad_norm": 1.891079068183899,
      "learning_rate": 1.8673417721518988e-05,
      "loss": 0.0955931767821312,
      "num_input_tokens_seen": 257298,
      "step": 137,
      "train_runtime": 356.6809,
      "train_tokens_per_second": 721.368
    },
    {
      "epoch": 0.13949962092494314,
      "grad_norm": 1.8957778215408325,
      "learning_rate": 1.8663291139240507e-05,
      "loss": 0.13465729355812073,
      "num_input_tokens_seen": 258954,
      "step": 138,
      "train_runtime": 358.8164,
      "train_tokens_per_second": 721.689
    },
    {
      "epoch": 0.14051048774323982,
      "grad_norm": 1.7684844732284546,
      "learning_rate": 1.8653164556962026e-05,
      "loss": 0.09110158681869507,
      "num_input_tokens_seen": 260652,
      "step": 139,
      "train_runtime": 360.9554,
      "train_tokens_per_second": 722.117
    },
    {
      "epoch": 0.14152135456153653,
      "grad_norm": 1.9227099418640137,
      "learning_rate": 1.8643037974683545e-05,
      "loss": 0.1080639511346817,
      "num_input_tokens_seen": 262534,
      "step": 140,
      "train_runtime": 363.1569,
      "train_tokens_per_second": 722.922
    },
    {
      "epoch": 0.1425322213798332,
      "grad_norm": 2.15092396736145,
      "learning_rate": 1.8632911392405064e-05,
      "loss": 0.09261924773454666,
      "num_input_tokens_seen": 264060,
      "step": 141,
      "train_runtime": 365.298,
      "train_tokens_per_second": 722.862
    },
    {
      "epoch": 0.1435430881981299,
      "grad_norm": 1.72467839717865,
      "learning_rate": 1.8622784810126584e-05,
      "loss": 0.08849076926708221,
      "num_input_tokens_seen": 265636,
      "step": 142,
      "train_runtime": 367.4008,
      "train_tokens_per_second": 723.014
    },
    {
      "epoch": 0.1445539550164266,
      "grad_norm": 2.267878532409668,
      "learning_rate": 1.8612658227848103e-05,
      "loss": 0.14270317554473877,
      "num_input_tokens_seen": 267074,
      "step": 143,
      "train_runtime": 369.471,
      "train_tokens_per_second": 722.855
    },
    {
      "epoch": 0.14556482183472327,
      "grad_norm": 1.9942142963409424,
      "learning_rate": 1.8602531645569622e-05,
      "loss": 0.10361195355653763,
      "num_input_tokens_seen": 269338,
      "step": 144,
      "train_runtime": 371.722,
      "train_tokens_per_second": 724.568
    },
    {
      "epoch": 0.14657568865301995,
      "grad_norm": 2.2510604858398438,
      "learning_rate": 1.859240506329114e-05,
      "loss": 0.12056334316730499,
      "num_input_tokens_seen": 271342,
      "step": 145,
      "train_runtime": 374.0016,
      "train_tokens_per_second": 725.51
    },
    {
      "epoch": 0.14758655547131666,
      "grad_norm": 2.6158652305603027,
      "learning_rate": 1.858227848101266e-05,
      "loss": 0.09602270275354385,
      "num_input_tokens_seen": 273232,
      "step": 146,
      "train_runtime": 376.2174,
      "train_tokens_per_second": 726.261
    },
    {
      "epoch": 0.14859742228961334,
      "grad_norm": 1.971649169921875,
      "learning_rate": 1.857215189873418e-05,
      "loss": 0.10333852469921112,
      "num_input_tokens_seen": 275106,
      "step": 147,
      "train_runtime": 378.462,
      "train_tokens_per_second": 726.905
    },
    {
      "epoch": 0.14960828910791002,
      "grad_norm": 1.993793249130249,
      "learning_rate": 1.8562025316455698e-05,
      "loss": 0.1353556513786316,
      "num_input_tokens_seen": 276940,
      "step": 148,
      "train_runtime": 380.6533,
      "train_tokens_per_second": 727.539
    },
    {
      "epoch": 0.15061915592620673,
      "grad_norm": 2.2878053188323975,
      "learning_rate": 1.8551898734177217e-05,
      "loss": 0.12384504079818726,
      "num_input_tokens_seen": 278862,
      "step": 149,
      "train_runtime": 382.9396,
      "train_tokens_per_second": 728.214
    },
    {
      "epoch": 0.1516300227445034,
      "grad_norm": 2.2018866539001465,
      "learning_rate": 1.8541772151898736e-05,
      "loss": 0.13707488775253296,
      "num_input_tokens_seen": 280454,
      "step": 150,
      "train_runtime": 385.055,
      "train_tokens_per_second": 728.348
    },
    {
      "epoch": 0.15264088956280011,
      "grad_norm": 2.000457763671875,
      "learning_rate": 1.8531645569620255e-05,
      "loss": 0.1024542897939682,
      "num_input_tokens_seen": 282348,
      "step": 151,
      "train_runtime": 393.534,
      "train_tokens_per_second": 717.468
    },
    {
      "epoch": 0.1536517563810968,
      "grad_norm": 1.7988492250442505,
      "learning_rate": 1.8521518987341774e-05,
      "loss": 0.09222530573606491,
      "num_input_tokens_seen": 284070,
      "step": 152,
      "train_runtime": 395.7276,
      "train_tokens_per_second": 717.842
    },
    {
      "epoch": 0.15466262319939347,
      "grad_norm": 2.2877094745635986,
      "learning_rate": 1.8511392405063293e-05,
      "loss": 0.145292267203331,
      "num_input_tokens_seen": 285946,
      "step": 153,
      "train_runtime": 397.9166,
      "train_tokens_per_second": 718.608
    },
    {
      "epoch": 0.15567349001769018,
      "grad_norm": 1.8731293678283691,
      "learning_rate": 1.850126582278481e-05,
      "loss": 0.12294542789459229,
      "num_input_tokens_seen": 287824,
      "step": 154,
      "train_runtime": 400.1086,
      "train_tokens_per_second": 719.365
    },
    {
      "epoch": 0.15668435683598686,
      "grad_norm": 1.6725560426712036,
      "learning_rate": 1.849113924050633e-05,
      "loss": 0.11795242875814438,
      "num_input_tokens_seen": 290194,
      "step": 155,
      "train_runtime": 402.4241,
      "train_tokens_per_second": 721.115
    },
    {
      "epoch": 0.15769522365428354,
      "grad_norm": 1.795425534248352,
      "learning_rate": 1.848101265822785e-05,
      "loss": 0.11238373816013336,
      "num_input_tokens_seen": 292346,
      "step": 156,
      "train_runtime": 404.6764,
      "train_tokens_per_second": 722.419
    },
    {
      "epoch": 0.15870609047258025,
      "grad_norm": 1.83493971824646,
      "learning_rate": 1.8470886075949366e-05,
      "loss": 0.13390424847602844,
      "num_input_tokens_seen": 294520,
      "step": 157,
      "train_runtime": 406.9315,
      "train_tokens_per_second": 723.758
    },
    {
      "epoch": 0.15971695729087693,
      "grad_norm": 2.047924041748047,
      "learning_rate": 1.846075949367089e-05,
      "loss": 0.12073490023612976,
      "num_input_tokens_seen": 296632,
      "step": 158,
      "train_runtime": 409.2316,
      "train_tokens_per_second": 724.851
    },
    {
      "epoch": 0.1607278241091736,
      "grad_norm": 1.8737225532531738,
      "learning_rate": 1.8450632911392408e-05,
      "loss": 0.10665791481733322,
      "num_input_tokens_seen": 298670,
      "step": 159,
      "train_runtime": 411.5096,
      "train_tokens_per_second": 725.791
    },
    {
      "epoch": 0.1617386909274703,
      "grad_norm": 2.387766122817993,
      "learning_rate": 1.8440506329113924e-05,
      "loss": 0.13044655323028564,
      "num_input_tokens_seen": 300156,
      "step": 160,
      "train_runtime": 413.6019,
      "train_tokens_per_second": 725.712
    },
    {
      "epoch": 0.162749557745767,
      "grad_norm": 1.9572023153305054,
      "learning_rate": 1.8430379746835446e-05,
      "loss": 0.1392357498407364,
      "num_input_tokens_seen": 301862,
      "step": 161,
      "train_runtime": 415.7351,
      "train_tokens_per_second": 726.092
    },
    {
      "epoch": 0.16376042456406367,
      "grad_norm": 1.8741393089294434,
      "learning_rate": 1.8420253164556962e-05,
      "loss": 0.10934767872095108,
      "num_input_tokens_seen": 303282,
      "step": 162,
      "train_runtime": 417.8117,
      "train_tokens_per_second": 725.882
    },
    {
      "epoch": 0.16477129138236038,
      "grad_norm": 1.9068574905395508,
      "learning_rate": 1.841012658227848e-05,
      "loss": 0.11664220690727234,
      "num_input_tokens_seen": 304986,
      "step": 163,
      "train_runtime": 419.9562,
      "train_tokens_per_second": 726.233
    },
    {
      "epoch": 0.16578215820065706,
      "grad_norm": 2.239553213119507,
      "learning_rate": 1.8400000000000003e-05,
      "loss": 0.10898025333881378,
      "num_input_tokens_seen": 306760,
      "step": 164,
      "train_runtime": 422.106,
      "train_tokens_per_second": 726.737
    },
    {
      "epoch": 0.16679302501895377,
      "grad_norm": 2.0420379638671875,
      "learning_rate": 1.838987341772152e-05,
      "loss": 0.12377582490444183,
      "num_input_tokens_seen": 308746,
      "step": 165,
      "train_runtime": 424.3356,
      "train_tokens_per_second": 727.599
    },
    {
      "epoch": 0.16780389183725045,
      "grad_norm": 2.119110345840454,
      "learning_rate": 1.8379746835443038e-05,
      "loss": 0.12563012540340424,
      "num_input_tokens_seen": 310382,
      "step": 166,
      "train_runtime": 426.5045,
      "train_tokens_per_second": 727.734
    },
    {
      "epoch": 0.16881475865554713,
      "grad_norm": 2.29335355758667,
      "learning_rate": 1.836962025316456e-05,
      "loss": 0.14566722512245178,
      "num_input_tokens_seen": 312192,
      "step": 167,
      "train_runtime": 428.7343,
      "train_tokens_per_second": 728.171
    },
    {
      "epoch": 0.16982562547384383,
      "grad_norm": 2.2466704845428467,
      "learning_rate": 1.8359493670886076e-05,
      "loss": 0.12062112987041473,
      "num_input_tokens_seen": 313830,
      "step": 168,
      "train_runtime": 431.0223,
      "train_tokens_per_second": 728.106
    },
    {
      "epoch": 0.1708364922921405,
      "grad_norm": 1.407857894897461,
      "learning_rate": 1.8349367088607595e-05,
      "loss": 0.08882243931293488,
      "num_input_tokens_seen": 316264,
      "step": 169,
      "train_runtime": 433.3712,
      "train_tokens_per_second": 729.776
    },
    {
      "epoch": 0.1718473591104372,
      "grad_norm": 1.7413135766983032,
      "learning_rate": 1.8339240506329115e-05,
      "loss": 0.1120813861489296,
      "num_input_tokens_seen": 318554,
      "step": 170,
      "train_runtime": 435.6513,
      "train_tokens_per_second": 731.213
    },
    {
      "epoch": 0.1728582259287339,
      "grad_norm": 2.6182568073272705,
      "learning_rate": 1.8329113924050634e-05,
      "loss": 0.13866037130355835,
      "num_input_tokens_seen": 320016,
      "step": 171,
      "train_runtime": 437.7569,
      "train_tokens_per_second": 731.036
    },
    {
      "epoch": 0.17386909274703058,
      "grad_norm": 1.7372453212738037,
      "learning_rate": 1.8318987341772153e-05,
      "loss": 0.1048712432384491,
      "num_input_tokens_seen": 321488,
      "step": 172,
      "train_runtime": 439.8334,
      "train_tokens_per_second": 730.931
    },
    {
      "epoch": 0.17487995956532726,
      "grad_norm": 1.8864858150482178,
      "learning_rate": 1.8308860759493672e-05,
      "loss": 0.12409922480583191,
      "num_input_tokens_seen": 323296,
      "step": 173,
      "train_runtime": 442.0048,
      "train_tokens_per_second": 731.431
    },
    {
      "epoch": 0.17589082638362397,
      "grad_norm": 2.2926132678985596,
      "learning_rate": 1.829873417721519e-05,
      "loss": 0.1726934313774109,
      "num_input_tokens_seen": 325238,
      "step": 174,
      "train_runtime": 444.2036,
      "train_tokens_per_second": 732.182
    },
    {
      "epoch": 0.17690169320192065,
      "grad_norm": 1.8553584814071655,
      "learning_rate": 1.828860759493671e-05,
      "loss": 0.11418695002794266,
      "num_input_tokens_seen": 327534,
      "step": 175,
      "train_runtime": 446.4571,
      "train_tokens_per_second": 733.629
    },
    {
      "epoch": 0.17791256002021733,
      "grad_norm": 1.7055572271347046,
      "learning_rate": 1.827848101265823e-05,
      "loss": 0.11833525449037552,
      "num_input_tokens_seen": 329454,
      "step": 176,
      "train_runtime": 448.6672,
      "train_tokens_per_second": 734.295
    },
    {
      "epoch": 0.17892342683851403,
      "grad_norm": 2.196890115737915,
      "learning_rate": 1.8268354430379748e-05,
      "loss": 0.12118027359247208,
      "num_input_tokens_seen": 331260,
      "step": 177,
      "train_runtime": 450.8396,
      "train_tokens_per_second": 734.762
    },
    {
      "epoch": 0.1799342936568107,
      "grad_norm": 2.3438303470611572,
      "learning_rate": 1.8258227848101267e-05,
      "loss": 0.1293737292289734,
      "num_input_tokens_seen": 333080,
      "step": 178,
      "train_runtime": 453.0017,
      "train_tokens_per_second": 735.273
    },
    {
      "epoch": 0.1809451604751074,
      "grad_norm": 2.1183717250823975,
      "learning_rate": 1.8248101265822786e-05,
      "loss": 0.10332482308149338,
      "num_input_tokens_seen": 334598,
      "step": 179,
      "train_runtime": 455.1101,
      "train_tokens_per_second": 735.202
    },
    {
      "epoch": 0.1819560272934041,
      "grad_norm": 1.9094011783599854,
      "learning_rate": 1.8237974683544305e-05,
      "loss": 0.08359847217798233,
      "num_input_tokens_seen": 336020,
      "step": 180,
      "train_runtime": 457.1868,
      "train_tokens_per_second": 734.973
    },
    {
      "epoch": 0.18296689411170078,
      "grad_norm": 2.663269281387329,
      "learning_rate": 1.8227848101265824e-05,
      "loss": 0.16630926728248596,
      "num_input_tokens_seen": 337744,
      "step": 181,
      "train_runtime": 465.7958,
      "train_tokens_per_second": 725.09
    },
    {
      "epoch": 0.18397776092999749,
      "grad_norm": 1.7838853597640991,
      "learning_rate": 1.8217721518987344e-05,
      "loss": 0.07612158358097076,
      "num_input_tokens_seen": 339166,
      "step": 182,
      "train_runtime": 467.9651,
      "train_tokens_per_second": 724.768
    },
    {
      "epoch": 0.18498862774829417,
      "grad_norm": 1.6300352811813354,
      "learning_rate": 1.8207594936708863e-05,
      "loss": 0.09971845149993896,
      "num_input_tokens_seen": 341148,
      "step": 183,
      "train_runtime": 470.2221,
      "train_tokens_per_second": 725.504
    },
    {
      "epoch": 0.18599949456659085,
      "grad_norm": 1.958196759223938,
      "learning_rate": 1.8197468354430382e-05,
      "loss": 0.1093597486615181,
      "num_input_tokens_seen": 342926,
      "step": 184,
      "train_runtime": 472.4691,
      "train_tokens_per_second": 725.817
    },
    {
      "epoch": 0.18701036138488755,
      "grad_norm": 1.9625788927078247,
      "learning_rate": 1.81873417721519e-05,
      "loss": 0.11427070200443268,
      "num_input_tokens_seen": 344884,
      "step": 185,
      "train_runtime": 474.6876,
      "train_tokens_per_second": 726.549
    },
    {
      "epoch": 0.18802122820318423,
      "grad_norm": 1.9605315923690796,
      "learning_rate": 1.817721518987342e-05,
      "loss": 0.1473732888698578,
      "num_input_tokens_seen": 346828,
      "step": 186,
      "train_runtime": 476.9824,
      "train_tokens_per_second": 727.129
    },
    {
      "epoch": 0.1890320950214809,
      "grad_norm": 1.6448954343795776,
      "learning_rate": 1.816708860759494e-05,
      "loss": 0.08606156706809998,
      "num_input_tokens_seen": 348858,
      "step": 187,
      "train_runtime": 479.2114,
      "train_tokens_per_second": 727.983
    },
    {
      "epoch": 0.19004296183977762,
      "grad_norm": 2.2308266162872314,
      "learning_rate": 1.8156962025316458e-05,
      "loss": 0.14672282338142395,
      "num_input_tokens_seen": 350544,
      "step": 188,
      "train_runtime": 481.3553,
      "train_tokens_per_second": 728.244
    },
    {
      "epoch": 0.1910538286580743,
      "grad_norm": 2.008155584335327,
      "learning_rate": 1.8146835443037977e-05,
      "loss": 0.13131436705589294,
      "num_input_tokens_seen": 352694,
      "step": 189,
      "train_runtime": 483.6153,
      "train_tokens_per_second": 729.286
    },
    {
      "epoch": 0.19206469547637098,
      "grad_norm": 1.6257904767990112,
      "learning_rate": 1.8136708860759496e-05,
      "loss": 0.11618009209632874,
      "num_input_tokens_seen": 354786,
      "step": 190,
      "train_runtime": 485.8682,
      "train_tokens_per_second": 730.21
    },
    {
      "epoch": 0.19307556229466769,
      "grad_norm": 2.078596830368042,
      "learning_rate": 1.8126582278481012e-05,
      "loss": 0.14265677332878113,
      "num_input_tokens_seen": 356610,
      "step": 191,
      "train_runtime": 488.0785,
      "train_tokens_per_second": 730.641
    },
    {
      "epoch": 0.19408642911296436,
      "grad_norm": 2.4707674980163574,
      "learning_rate": 1.8116455696202534e-05,
      "loss": 0.16185879707336426,
      "num_input_tokens_seen": 358904,
      "step": 192,
      "train_runtime": 490.3884,
      "train_tokens_per_second": 731.877
    },
    {
      "epoch": 0.19509729593126104,
      "grad_norm": 2.3911948204040527,
      "learning_rate": 1.8106329113924053e-05,
      "loss": 0.14641645550727844,
      "num_input_tokens_seen": 360522,
      "step": 193,
      "train_runtime": 492.5371,
      "train_tokens_per_second": 731.969
    },
    {
      "epoch": 0.19610816274955775,
      "grad_norm": 2.293113946914673,
      "learning_rate": 1.809620253164557e-05,
      "loss": 0.11854423582553864,
      "num_input_tokens_seen": 362594,
      "step": 194,
      "train_runtime": 494.7953,
      "train_tokens_per_second": 732.816
    },
    {
      "epoch": 0.19711902956785443,
      "grad_norm": 2.041398286819458,
      "learning_rate": 1.808607594936709e-05,
      "loss": 0.08841122686862946,
      "num_input_tokens_seen": 364210,
      "step": 195,
      "train_runtime": 496.965,
      "train_tokens_per_second": 732.868
    },
    {
      "epoch": 0.1981298963861511,
      "grad_norm": 1.735707402229309,
      "learning_rate": 1.807594936708861e-05,
      "loss": 0.10502967983484268,
      "num_input_tokens_seen": 366304,
      "step": 196,
      "train_runtime": 499.1902,
      "train_tokens_per_second": 733.796
    },
    {
      "epoch": 0.19914076320444782,
      "grad_norm": 1.6637848615646362,
      "learning_rate": 1.8065822784810126e-05,
      "loss": 0.10303906351327896,
      "num_input_tokens_seen": 368402,
      "step": 197,
      "train_runtime": 501.48,
      "train_tokens_per_second": 734.629
    },
    {
      "epoch": 0.2001516300227445,
      "grad_norm": 2.36517071723938,
      "learning_rate": 1.805569620253165e-05,
      "loss": 0.1280064731836319,
      "num_input_tokens_seen": 370366,
      "step": 198,
      "train_runtime": 503.6868,
      "train_tokens_per_second": 735.31
    },
    {
      "epoch": 0.2001516300227445,
      "eval_loss": 0.44587889313697815,
      "eval_runtime": 62.2769,
      "eval_samples_per_second": 14.114,
      "eval_steps_per_second": 7.065,
      "num_input_tokens_seen": 370366,
      "step": 198
    },
    {
      "epoch": 0.2011624968410412,
      "grad_norm": 2.617011547088623,
      "learning_rate": 1.8045569620253165e-05,
      "loss": 0.1196804791688919,
      "num_input_tokens_seen": 371792,
      "step": 199,
      "train_runtime": 569.0247,
      "train_tokens_per_second": 653.385
    },
    {
      "epoch": 0.20217336365933788,
      "grad_norm": 1.9300251007080078,
      "learning_rate": 1.8035443037974684e-05,
      "loss": 0.10957969725131989,
      "num_input_tokens_seen": 373630,
      "step": 200,
      "train_runtime": 571.262,
      "train_tokens_per_second": 654.043
    },
    {
      "epoch": 0.20318423047763456,
      "grad_norm": 2.0042214393615723,
      "learning_rate": 1.8025316455696206e-05,
      "loss": 0.09827364236116409,
      "num_input_tokens_seen": 375662,
      "step": 201,
      "train_runtime": 573.5096,
      "train_tokens_per_second": 655.023
    },
    {
      "epoch": 0.20419509729593127,
      "grad_norm": 2.0204923152923584,
      "learning_rate": 1.8015189873417722e-05,
      "loss": 0.09973488748073578,
      "num_input_tokens_seen": 377436,
      "step": 202,
      "train_runtime": 575.6869,
      "train_tokens_per_second": 655.627
    },
    {
      "epoch": 0.20520596411422795,
      "grad_norm": 2.118560791015625,
      "learning_rate": 1.800506329113924e-05,
      "loss": 0.13264839351177216,
      "num_input_tokens_seen": 379004,
      "step": 203,
      "train_runtime": 577.8081,
      "train_tokens_per_second": 655.934
    },
    {
      "epoch": 0.20621683093252463,
      "grad_norm": 2.4744246006011963,
      "learning_rate": 1.7994936708860763e-05,
      "loss": 0.12377555668354034,
      "num_input_tokens_seen": 380828,
      "step": 204,
      "train_runtime": 580.0246,
      "train_tokens_per_second": 656.572
    },
    {
      "epoch": 0.20722769775082134,
      "grad_norm": 1.9765815734863281,
      "learning_rate": 1.798481012658228e-05,
      "loss": 0.09765441715717316,
      "num_input_tokens_seen": 382788,
      "step": 205,
      "train_runtime": 582.2576,
      "train_tokens_per_second": 657.42
    },
    {
      "epoch": 0.20823856456911802,
      "grad_norm": 2.1058332920074463,
      "learning_rate": 1.7974683544303798e-05,
      "loss": 0.11945100128650665,
      "num_input_tokens_seen": 384470,
      "step": 206,
      "train_runtime": 584.4355,
      "train_tokens_per_second": 657.848
    },
    {
      "epoch": 0.2092494313874147,
      "grad_norm": 1.6983367204666138,
      "learning_rate": 1.7964556962025317e-05,
      "loss": 0.09330539405345917,
      "num_input_tokens_seen": 386452,
      "step": 207,
      "train_runtime": 586.6725,
      "train_tokens_per_second": 658.718
    },
    {
      "epoch": 0.2102602982057114,
      "grad_norm": 1.8551362752914429,
      "learning_rate": 1.7954430379746836e-05,
      "loss": 0.11788757890462875,
      "num_input_tokens_seen": 388504,
      "step": 208,
      "train_runtime": 588.9246,
      "train_tokens_per_second": 659.684
    },
    {
      "epoch": 0.21127116502400808,
      "grad_norm": 2.1599807739257812,
      "learning_rate": 1.7944303797468355e-05,
      "loss": 0.1124764159321785,
      "num_input_tokens_seen": 390166,
      "step": 209,
      "train_runtime": 591.0736,
      "train_tokens_per_second": 660.097
    },
    {
      "epoch": 0.21228203184230476,
      "grad_norm": 1.961949348449707,
      "learning_rate": 1.7934177215189875e-05,
      "loss": 0.10842674970626831,
      "num_input_tokens_seen": 391552,
      "step": 210,
      "train_runtime": 593.1879,
      "train_tokens_per_second": 660.081
    },
    {
      "epoch": 0.21329289866060147,
      "grad_norm": 1.7553244829177856,
      "learning_rate": 1.7924050632911394e-05,
      "loss": 0.12177146971225739,
      "num_input_tokens_seen": 393310,
      "step": 211,
      "train_runtime": 601.6506,
      "train_tokens_per_second": 653.718
    },
    {
      "epoch": 0.21430376547889815,
      "grad_norm": 1.8308024406433105,
      "learning_rate": 1.7913924050632913e-05,
      "loss": 0.1084492951631546,
      "num_input_tokens_seen": 395432,
      "step": 212,
      "train_runtime": 603.882,
      "train_tokens_per_second": 654.817
    },
    {
      "epoch": 0.21531463229719486,
      "grad_norm": 1.9430619478225708,
      "learning_rate": 1.7903797468354432e-05,
      "loss": 0.09723185002803802,
      "num_input_tokens_seen": 397112,
      "step": 213,
      "train_runtime": 606.0291,
      "train_tokens_per_second": 655.269
    },
    {
      "epoch": 0.21632549911549154,
      "grad_norm": 1.5587821006774902,
      "learning_rate": 1.789367088607595e-05,
      "loss": 0.08421872556209564,
      "num_input_tokens_seen": 399078,
      "step": 214,
      "train_runtime": 608.224,
      "train_tokens_per_second": 656.137
    },
    {
      "epoch": 0.21733636593378822,
      "grad_norm": 1.7721914052963257,
      "learning_rate": 1.788354430379747e-05,
      "loss": 0.10960154235363007,
      "num_input_tokens_seen": 401186,
      "step": 215,
      "train_runtime": 610.4651,
      "train_tokens_per_second": 657.181
    },
    {
      "epoch": 0.21834723275208492,
      "grad_norm": 2.3418967723846436,
      "learning_rate": 1.787341772151899e-05,
      "loss": 0.1320759654045105,
      "num_input_tokens_seen": 402874,
      "step": 216,
      "train_runtime": 612.6218,
      "train_tokens_per_second": 657.623
    },
    {
      "epoch": 0.2193580995703816,
      "grad_norm": 1.7354557514190674,
      "learning_rate": 1.7863291139240508e-05,
      "loss": 0.09888164699077606,
      "num_input_tokens_seen": 404962,
      "step": 217,
      "train_runtime": 614.8703,
      "train_tokens_per_second": 658.614
    },
    {
      "epoch": 0.22036896638867828,
      "grad_norm": 1.5967291593551636,
      "learning_rate": 1.7853164556962027e-05,
      "loss": 0.10974611341953278,
      "num_input_tokens_seen": 407460,
      "step": 218,
      "train_runtime": 617.2313,
      "train_tokens_per_second": 660.142
    },
    {
      "epoch": 0.221379833206975,
      "grad_norm": 1.9726117849349976,
      "learning_rate": 1.7843037974683546e-05,
      "loss": 0.12994709610939026,
      "num_input_tokens_seen": 409224,
      "step": 219,
      "train_runtime": 619.385,
      "train_tokens_per_second": 660.694
    },
    {
      "epoch": 0.22239070002527167,
      "grad_norm": 1.8002862930297852,
      "learning_rate": 1.7832911392405065e-05,
      "loss": 0.0879456102848053,
      "num_input_tokens_seen": 411000,
      "step": 220,
      "train_runtime": 621.5693,
      "train_tokens_per_second": 661.23
    },
    {
      "epoch": 0.22340156684356835,
      "grad_norm": 2.091977596282959,
      "learning_rate": 1.7822784810126584e-05,
      "loss": 0.13014501333236694,
      "num_input_tokens_seen": 412552,
      "step": 221,
      "train_runtime": 623.7077,
      "train_tokens_per_second": 661.451
    },
    {
      "epoch": 0.22441243366186506,
      "grad_norm": 2.036530017852783,
      "learning_rate": 1.7812658227848104e-05,
      "loss": 0.0972268208861351,
      "num_input_tokens_seen": 414648,
      "step": 222,
      "train_runtime": 625.9457,
      "train_tokens_per_second": 662.434
    },
    {
      "epoch": 0.22542330048016174,
      "grad_norm": 1.4234565496444702,
      "learning_rate": 1.780253164556962e-05,
      "loss": 0.09172986447811127,
      "num_input_tokens_seen": 416936,
      "step": 223,
      "train_runtime": 628.2903,
      "train_tokens_per_second": 663.604
    },
    {
      "epoch": 0.22643416729845842,
      "grad_norm": 1.8967690467834473,
      "learning_rate": 1.779240506329114e-05,
      "loss": 0.1082121804356575,
      "num_input_tokens_seen": 418750,
      "step": 224,
      "train_runtime": 630.4816,
      "train_tokens_per_second": 664.175
    },
    {
      "epoch": 0.22744503411675512,
      "grad_norm": 1.5741974115371704,
      "learning_rate": 1.778227848101266e-05,
      "loss": 0.10154334455728531,
      "num_input_tokens_seen": 420940,
      "step": 225,
      "train_runtime": 632.7984,
      "train_tokens_per_second": 665.204
    },
    {
      "epoch": 0.2284559009350518,
      "grad_norm": 2.162623643875122,
      "learning_rate": 1.7772151898734176e-05,
      "loss": 0.10570146888494492,
      "num_input_tokens_seen": 422712,
      "step": 226,
      "train_runtime": 634.9789,
      "train_tokens_per_second": 665.71
    },
    {
      "epoch": 0.22946676775334848,
      "grad_norm": 1.799513339996338,
      "learning_rate": 1.77620253164557e-05,
      "loss": 0.11576129496097565,
      "num_input_tokens_seen": 424832,
      "step": 227,
      "train_runtime": 637.2459,
      "train_tokens_per_second": 666.669
    },
    {
      "epoch": 0.2304776345716452,
      "grad_norm": 2.410942792892456,
      "learning_rate": 1.7751898734177218e-05,
      "loss": 0.13010409474372864,
      "num_input_tokens_seen": 426646,
      "step": 228,
      "train_runtime": 639.4501,
      "train_tokens_per_second": 667.208
    },
    {
      "epoch": 0.23148850138994187,
      "grad_norm": 1.8130074739456177,
      "learning_rate": 1.7741772151898734e-05,
      "loss": 0.12471440434455872,
      "num_input_tokens_seen": 428700,
      "step": 229,
      "train_runtime": 641.6914,
      "train_tokens_per_second": 668.078
    },
    {
      "epoch": 0.23249936820823858,
      "grad_norm": 2.150529384613037,
      "learning_rate": 1.7731645569620256e-05,
      "loss": 0.14512759447097778,
      "num_input_tokens_seen": 430232,
      "step": 230,
      "train_runtime": 643.8152,
      "train_tokens_per_second": 668.254
    },
    {
      "epoch": 0.23351023502653526,
      "grad_norm": 1.7570277452468872,
      "learning_rate": 1.7721518987341772e-05,
      "loss": 0.10100769251585007,
      "num_input_tokens_seen": 431962,
      "step": 231,
      "train_runtime": 645.9925,
      "train_tokens_per_second": 668.68
    },
    {
      "epoch": 0.23452110184483194,
      "grad_norm": 1.8429783582687378,
      "learning_rate": 1.771139240506329e-05,
      "loss": 0.12079241871833801,
      "num_input_tokens_seen": 434320,
      "step": 232,
      "train_runtime": 648.288,
      "train_tokens_per_second": 669.949
    },
    {
      "epoch": 0.23553196866312864,
      "grad_norm": 1.4480996131896973,
      "learning_rate": 1.7701265822784813e-05,
      "loss": 0.08419619500637054,
      "num_input_tokens_seen": 436184,
      "step": 233,
      "train_runtime": 650.4646,
      "train_tokens_per_second": 670.573
    },
    {
      "epoch": 0.23654283548142532,
      "grad_norm": 2.275111436843872,
      "learning_rate": 1.769113924050633e-05,
      "loss": 0.15849897265434265,
      "num_input_tokens_seen": 438052,
      "step": 234,
      "train_runtime": 652.6707,
      "train_tokens_per_second": 671.168
    },
    {
      "epoch": 0.237553702299722,
      "grad_norm": 1.7819358110427856,
      "learning_rate": 1.7681012658227848e-05,
      "loss": 0.12310174107551575,
      "num_input_tokens_seen": 440310,
      "step": 235,
      "train_runtime": 654.9677,
      "train_tokens_per_second": 672.262
    },
    {
      "epoch": 0.2385645691180187,
      "grad_norm": 2.355914354324341,
      "learning_rate": 1.767088607594937e-05,
      "loss": 0.1537615805864334,
      "num_input_tokens_seen": 442282,
      "step": 236,
      "train_runtime": 657.2043,
      "train_tokens_per_second": 672.975
    },
    {
      "epoch": 0.2395754359363154,
      "grad_norm": 2.181678056716919,
      "learning_rate": 1.7660759493670886e-05,
      "loss": 0.16061922907829285,
      "num_input_tokens_seen": 444134,
      "step": 237,
      "train_runtime": 659.4017,
      "train_tokens_per_second": 673.541
    },
    {
      "epoch": 0.24058630275461207,
      "grad_norm": 2.0904757976531982,
      "learning_rate": 1.7650632911392405e-05,
      "loss": 0.11011753976345062,
      "num_input_tokens_seen": 445702,
      "step": 238,
      "train_runtime": 661.5091,
      "train_tokens_per_second": 673.765
    },
    {
      "epoch": 0.24159716957290878,
      "grad_norm": 1.9311637878417969,
      "learning_rate": 1.7640506329113925e-05,
      "loss": 0.14344699680805206,
      "num_input_tokens_seen": 447646,
      "step": 239,
      "train_runtime": 663.7396,
      "train_tokens_per_second": 674.43
    },
    {
      "epoch": 0.24260803639120546,
      "grad_norm": 2.2915382385253906,
      "learning_rate": 1.7630379746835444e-05,
      "loss": 0.10493789613246918,
      "num_input_tokens_seen": 448988,
      "step": 240,
      "train_runtime": 665.8128,
      "train_tokens_per_second": 674.346
    },
    {
      "epoch": 0.24361890320950214,
      "grad_norm": 2.1953468322753906,
      "learning_rate": 1.7620253164556963e-05,
      "loss": 0.10852991789579391,
      "num_input_tokens_seen": 450462,
      "step": 241,
      "train_runtime": 674.1803,
      "train_tokens_per_second": 668.163
    },
    {
      "epoch": 0.24462977002779884,
      "grad_norm": 2.380145788192749,
      "learning_rate": 1.7610126582278482e-05,
      "loss": 0.10469101369380951,
      "num_input_tokens_seen": 451978,
      "step": 242,
      "train_runtime": 676.2928,
      "train_tokens_per_second": 668.317
    },
    {
      "epoch": 0.24564063684609552,
      "grad_norm": 2.1115336418151855,
      "learning_rate": 1.76e-05,
      "loss": 0.09177637100219727,
      "num_input_tokens_seen": 453610,
      "step": 243,
      "train_runtime": 678.4436,
      "train_tokens_per_second": 668.604
    },
    {
      "epoch": 0.24665150366439223,
      "grad_norm": 1.653169870376587,
      "learning_rate": 1.758987341772152e-05,
      "loss": 0.11477974057197571,
      "num_input_tokens_seen": 455582,
      "step": 244,
      "train_runtime": 680.6526,
      "train_tokens_per_second": 669.331
    },
    {
      "epoch": 0.2476623704826889,
      "grad_norm": 1.7820411920547485,
      "learning_rate": 1.757974683544304e-05,
      "loss": 0.07974696159362793,
      "num_input_tokens_seen": 457100,
      "step": 245,
      "train_runtime": 682.7487,
      "train_tokens_per_second": 669.5
    },
    {
      "epoch": 0.2486732373009856,
      "grad_norm": 1.890511155128479,
      "learning_rate": 1.7569620253164558e-05,
      "loss": 0.12120610475540161,
      "num_input_tokens_seen": 459110,
      "step": 246,
      "train_runtime": 684.9928,
      "train_tokens_per_second": 670.241
    },
    {
      "epoch": 0.2496841041192823,
      "grad_norm": 2.860581159591675,
      "learning_rate": 1.7559493670886077e-05,
      "loss": 0.15575653314590454,
      "num_input_tokens_seen": 460896,
      "step": 247,
      "train_runtime": 687.2072,
      "train_tokens_per_second": 670.68
    },
    {
      "epoch": 0.250694970937579,
      "grad_norm": 1.8721354007720947,
      "learning_rate": 1.7549367088607596e-05,
      "loss": 0.09407229721546173,
      "num_input_tokens_seen": 462456,
      "step": 248,
      "train_runtime": 689.3281,
      "train_tokens_per_second": 670.879
    },
    {
      "epoch": 0.2517058377558757,
      "grad_norm": 1.8691226243972778,
      "learning_rate": 1.7539240506329115e-05,
      "loss": 0.09890297055244446,
      "num_input_tokens_seen": 464000,
      "step": 249,
      "train_runtime": 691.4529,
      "train_tokens_per_second": 671.051
    },
    {
      "epoch": 0.25271670457417234,
      "grad_norm": 1.9749469757080078,
      "learning_rate": 1.7529113924050634e-05,
      "loss": 0.09291301667690277,
      "num_input_tokens_seen": 465810,
      "step": 250,
      "train_runtime": 693.6414,
      "train_tokens_per_second": 671.543
    },
    {
      "epoch": 0.25372757139246904,
      "grad_norm": 2.137969732284546,
      "learning_rate": 1.7518987341772154e-05,
      "loss": 0.12524446845054626,
      "num_input_tokens_seen": 468190,
      "step": 251,
      "train_runtime": 696.0274,
      "train_tokens_per_second": 672.66
    },
    {
      "epoch": 0.25473843821076575,
      "grad_norm": 1.6700412034988403,
      "learning_rate": 1.7508860759493673e-05,
      "loss": 0.09336793422698975,
      "num_input_tokens_seen": 470162,
      "step": 252,
      "train_runtime": 698.3222,
      "train_tokens_per_second": 673.274
    },
    {
      "epoch": 0.2557493050290624,
      "grad_norm": 2.4621388912200928,
      "learning_rate": 1.7498734177215192e-05,
      "loss": 0.14584292471408844,
      "num_input_tokens_seen": 472040,
      "step": 253,
      "train_runtime": 700.6228,
      "train_tokens_per_second": 673.743
    },
    {
      "epoch": 0.2567601718473591,
      "grad_norm": 2.0203704833984375,
      "learning_rate": 1.748860759493671e-05,
      "loss": 0.11945211887359619,
      "num_input_tokens_seen": 474150,
      "step": 254,
      "train_runtime": 702.8545,
      "train_tokens_per_second": 674.606
    },
    {
      "epoch": 0.2577710386656558,
      "grad_norm": 1.7765873670578003,
      "learning_rate": 1.747848101265823e-05,
      "loss": 0.0990256667137146,
      "num_input_tokens_seen": 476552,
      "step": 255,
      "train_runtime": 705.1809,
      "train_tokens_per_second": 675.787
    },
    {
      "epoch": 0.25878190548395247,
      "grad_norm": 2.020219087600708,
      "learning_rate": 1.746835443037975e-05,
      "loss": 0.11051210016012192,
      "num_input_tokens_seen": 478316,
      "step": 256,
      "train_runtime": 707.3766,
      "train_tokens_per_second": 676.183
    },
    {
      "epoch": 0.2597927723022492,
      "grad_norm": 1.8801995515823364,
      "learning_rate": 1.7458227848101268e-05,
      "loss": 0.12179265916347504,
      "num_input_tokens_seen": 480664,
      "step": 257,
      "train_runtime": 709.6671,
      "train_tokens_per_second": 677.309
    },
    {
      "epoch": 0.2608036391205459,
      "grad_norm": 2.045766592025757,
      "learning_rate": 1.7448101265822787e-05,
      "loss": 0.10970637202262878,
      "num_input_tokens_seen": 482522,
      "step": 258,
      "train_runtime": 711.8696,
      "train_tokens_per_second": 677.824
    },
    {
      "epoch": 0.26181450593884253,
      "grad_norm": 1.6100939512252808,
      "learning_rate": 1.7437974683544306e-05,
      "loss": 0.13290613889694214,
      "num_input_tokens_seen": 484608,
      "step": 259,
      "train_runtime": 714.1242,
      "train_tokens_per_second": 678.605
    },
    {
      "epoch": 0.26282537275713924,
      "grad_norm": 1.9544614553451538,
      "learning_rate": 1.7427848101265825e-05,
      "loss": 0.12113123387098312,
      "num_input_tokens_seen": 486676,
      "step": 260,
      "train_runtime": 716.3455,
      "train_tokens_per_second": 679.387
    },
    {
      "epoch": 0.26383623957543595,
      "grad_norm": 2.5813403129577637,
      "learning_rate": 1.7417721518987344e-05,
      "loss": 0.1483939290046692,
      "num_input_tokens_seen": 488252,
      "step": 261,
      "train_runtime": 718.4497,
      "train_tokens_per_second": 679.591
    },
    {
      "epoch": 0.2648471063937326,
      "grad_norm": 1.8588289022445679,
      "learning_rate": 1.7407594936708863e-05,
      "loss": 0.13005565106868744,
      "num_input_tokens_seen": 490056,
      "step": 262,
      "train_runtime": 720.6284,
      "train_tokens_per_second": 680.04
    },
    {
      "epoch": 0.2658579732120293,
      "grad_norm": 2.1340038776397705,
      "learning_rate": 1.739746835443038e-05,
      "loss": 0.14259295165538788,
      "num_input_tokens_seen": 492050,
      "step": 263,
      "train_runtime": 722.9093,
      "train_tokens_per_second": 680.653
    },
    {
      "epoch": 0.266868840030326,
      "grad_norm": 1.7459964752197266,
      "learning_rate": 1.73873417721519e-05,
      "loss": 0.10176543891429901,
      "num_input_tokens_seen": 494088,
      "step": 264,
      "train_runtime": 725.2014,
      "train_tokens_per_second": 681.311
    },
    {
      "epoch": 0.26787970684862267,
      "grad_norm": 1.9116030931472778,
      "learning_rate": 1.737721518987342e-05,
      "loss": 0.14606161415576935,
      "num_input_tokens_seen": 496250,
      "step": 265,
      "train_runtime": 727.4671,
      "train_tokens_per_second": 682.161
    },
    {
      "epoch": 0.2688905736669194,
      "grad_norm": 1.8587530851364136,
      "learning_rate": 1.7367088607594936e-05,
      "loss": 0.1209803968667984,
      "num_input_tokens_seen": 498554,
      "step": 266,
      "train_runtime": 729.8733,
      "train_tokens_per_second": 683.069
    },
    {
      "epoch": 0.2699014404852161,
      "grad_norm": 2.1367876529693604,
      "learning_rate": 1.735696202531646e-05,
      "loss": 0.12396949529647827,
      "num_input_tokens_seen": 500336,
      "step": 267,
      "train_runtime": 732.0551,
      "train_tokens_per_second": 683.468
    },
    {
      "epoch": 0.2709123073035128,
      "grad_norm": 1.9846808910369873,
      "learning_rate": 1.7346835443037978e-05,
      "loss": 0.1330546736717224,
      "num_input_tokens_seen": 502388,
      "step": 268,
      "train_runtime": 734.2692,
      "train_tokens_per_second": 684.201
    },
    {
      "epoch": 0.27192317412180944,
      "grad_norm": 1.9799429178237915,
      "learning_rate": 1.7336708860759494e-05,
      "loss": 0.09931035339832306,
      "num_input_tokens_seen": 504016,
      "step": 269,
      "train_runtime": 736.4641,
      "train_tokens_per_second": 684.373
    },
    {
      "epoch": 0.27293404094010615,
      "grad_norm": 1.8704789876937866,
      "learning_rate": 1.7326582278481016e-05,
      "loss": 0.14354629814624786,
      "num_input_tokens_seen": 506228,
      "step": 270,
      "train_runtime": 738.8562,
      "train_tokens_per_second": 685.151
    },
    {
      "epoch": 0.27394490775840286,
      "grad_norm": 2.032816171646118,
      "learning_rate": 1.7316455696202532e-05,
      "loss": 0.09989900141954422,
      "num_input_tokens_seen": 507924,
      "step": 271,
      "train_runtime": 746.8839,
      "train_tokens_per_second": 680.058
    },
    {
      "epoch": 0.2749557745766995,
      "grad_norm": 1.7375068664550781,
      "learning_rate": 1.730632911392405e-05,
      "loss": 0.1109226867556572,
      "num_input_tokens_seen": 509700,
      "step": 272,
      "train_runtime": 749.0842,
      "train_tokens_per_second": 680.431
    },
    {
      "epoch": 0.2759666413949962,
      "grad_norm": 2.1660964488983154,
      "learning_rate": 1.729620253164557e-05,
      "loss": 0.11639153957366943,
      "num_input_tokens_seen": 511482,
      "step": 273,
      "train_runtime": 751.2666,
      "train_tokens_per_second": 680.826
    },
    {
      "epoch": 0.2769775082132929,
      "grad_norm": 1.8768203258514404,
      "learning_rate": 1.728607594936709e-05,
      "loss": 0.12145338207483292,
      "num_input_tokens_seen": 513976,
      "step": 274,
      "train_runtime": 753.6298,
      "train_tokens_per_second": 682.001
    },
    {
      "epoch": 0.2779883750315896,
      "grad_norm": 1.9955626726150513,
      "learning_rate": 1.7275949367088608e-05,
      "loss": 0.11777406185865402,
      "num_input_tokens_seen": 515382,
      "step": 275,
      "train_runtime": 755.7224,
      "train_tokens_per_second": 681.973
    },
    {
      "epoch": 0.2789992418498863,
      "grad_norm": 1.6571555137634277,
      "learning_rate": 1.7265822784810127e-05,
      "loss": 0.09360526502132416,
      "num_input_tokens_seen": 517038,
      "step": 276,
      "train_runtime": 757.9178,
      "train_tokens_per_second": 682.182
    },
    {
      "epoch": 0.280010108668183,
      "grad_norm": 1.8501170873641968,
      "learning_rate": 1.7255696202531646e-05,
      "loss": 0.12285037338733673,
      "num_input_tokens_seen": 519080,
      "step": 277,
      "train_runtime": 760.171,
      "train_tokens_per_second": 682.846
    },
    {
      "epoch": 0.28102097548647964,
      "grad_norm": 1.668231725692749,
      "learning_rate": 1.7245569620253165e-05,
      "loss": 0.12230192869901657,
      "num_input_tokens_seen": 520804,
      "step": 278,
      "train_runtime": 762.3715,
      "train_tokens_per_second": 683.137
    },
    {
      "epoch": 0.28203184230477635,
      "grad_norm": 1.6942402124404907,
      "learning_rate": 1.7235443037974685e-05,
      "loss": 0.1275428831577301,
      "num_input_tokens_seen": 523082,
      "step": 279,
      "train_runtime": 764.6501,
      "train_tokens_per_second": 684.08
    },
    {
      "epoch": 0.28304270912307306,
      "grad_norm": 1.873801827430725,
      "learning_rate": 1.7225316455696204e-05,
      "loss": 0.12711967527866364,
      "num_input_tokens_seen": 524988,
      "step": 280,
      "train_runtime": 766.8679,
      "train_tokens_per_second": 684.587
    },
    {
      "epoch": 0.2840535759413697,
      "grad_norm": 2.6262946128845215,
      "learning_rate": 1.7215189873417723e-05,
      "loss": 0.11261977255344391,
      "num_input_tokens_seen": 526770,
      "step": 281,
      "train_runtime": 769.0193,
      "train_tokens_per_second": 684.989
    },
    {
      "epoch": 0.2850644427596664,
      "grad_norm": 1.857487440109253,
      "learning_rate": 1.7205063291139242e-05,
      "loss": 0.10549350827932358,
      "num_input_tokens_seen": 528400,
      "step": 282,
      "train_runtime": 771.1449,
      "train_tokens_per_second": 685.215
    },
    {
      "epoch": 0.2860753095779631,
      "grad_norm": 1.779315710067749,
      "learning_rate": 1.719493670886076e-05,
      "loss": 0.09796726703643799,
      "num_input_tokens_seen": 530224,
      "step": 283,
      "train_runtime": 773.3346,
      "train_tokens_per_second": 685.633
    },
    {
      "epoch": 0.2870861763962598,
      "grad_norm": 1.9219661951065063,
      "learning_rate": 1.718481012658228e-05,
      "loss": 0.11788786947727203,
      "num_input_tokens_seen": 532286,
      "step": 284,
      "train_runtime": 775.5473,
      "train_tokens_per_second": 686.336
    },
    {
      "epoch": 0.2880970432145565,
      "grad_norm": 1.5746098756790161,
      "learning_rate": 1.71746835443038e-05,
      "loss": 0.08101858198642731,
      "num_input_tokens_seen": 534178,
      "step": 285,
      "train_runtime": 777.738,
      "train_tokens_per_second": 686.835
    },
    {
      "epoch": 0.2891079100328532,
      "grad_norm": 1.6413177251815796,
      "learning_rate": 1.7164556962025318e-05,
      "loss": 0.10468021035194397,
      "num_input_tokens_seen": 535870,
      "step": 286,
      "train_runtime": 779.8662,
      "train_tokens_per_second": 687.131
    },
    {
      "epoch": 0.29011877685114984,
      "grad_norm": 1.7688888311386108,
      "learning_rate": 1.7154430379746837e-05,
      "loss": 0.12281863391399384,
      "num_input_tokens_seen": 537938,
      "step": 287,
      "train_runtime": 782.0896,
      "train_tokens_per_second": 687.821
    },
    {
      "epoch": 0.29112964366944655,
      "grad_norm": 1.9296395778656006,
      "learning_rate": 1.7144303797468356e-05,
      "loss": 0.10115273296833038,
      "num_input_tokens_seen": 540342,
      "step": 288,
      "train_runtime": 784.4024,
      "train_tokens_per_second": 688.858
    },
    {
      "epoch": 0.29214051048774325,
      "grad_norm": 1.7572964429855347,
      "learning_rate": 1.7134177215189875e-05,
      "loss": 0.10689762234687805,
      "num_input_tokens_seen": 542224,
      "step": 289,
      "train_runtime": 786.6099,
      "train_tokens_per_second": 689.318
    },
    {
      "epoch": 0.2931513773060399,
      "grad_norm": 2.1670010089874268,
      "learning_rate": 1.7124050632911394e-05,
      "loss": 0.10287714004516602,
      "num_input_tokens_seen": 544196,
      "step": 290,
      "train_runtime": 788.8264,
      "train_tokens_per_second": 689.881
    },
    {
      "epoch": 0.2941622441243366,
      "grad_norm": 1.8853284120559692,
      "learning_rate": 1.7113924050632914e-05,
      "loss": 0.10559524595737457,
      "num_input_tokens_seen": 546366,
      "step": 291,
      "train_runtime": 791.1216,
      "train_tokens_per_second": 690.622
    },
    {
      "epoch": 0.2951731109426333,
      "grad_norm": 2.108994960784912,
      "learning_rate": 1.7103797468354433e-05,
      "loss": 0.10349194705486298,
      "num_input_tokens_seen": 548322,
      "step": 292,
      "train_runtime": 793.3614,
      "train_tokens_per_second": 691.138
    },
    {
      "epoch": 0.29618397776093,
      "grad_norm": 1.8746708631515503,
      "learning_rate": 1.7093670886075952e-05,
      "loss": 0.0990857183933258,
      "num_input_tokens_seen": 550206,
      "step": 293,
      "train_runtime": 795.6021,
      "train_tokens_per_second": 691.559
    },
    {
      "epoch": 0.2971948445792267,
      "grad_norm": 1.834490418434143,
      "learning_rate": 1.708354430379747e-05,
      "loss": 0.10397559404373169,
      "num_input_tokens_seen": 552534,
      "step": 294,
      "train_runtime": 797.8938,
      "train_tokens_per_second": 692.491
    },
    {
      "epoch": 0.2982057113975234,
      "grad_norm": 2.039677858352661,
      "learning_rate": 1.7073417721518987e-05,
      "loss": 0.13525766134262085,
      "num_input_tokens_seen": 554498,
      "step": 295,
      "train_runtime": 800.1127,
      "train_tokens_per_second": 693.025
    },
    {
      "epoch": 0.29921657821582004,
      "grad_norm": 2.3371741771698,
      "learning_rate": 1.706329113924051e-05,
      "loss": 0.09204934537410736,
      "num_input_tokens_seen": 555778,
      "step": 296,
      "train_runtime": 802.1777,
      "train_tokens_per_second": 692.836
    },
    {
      "epoch": 0.30022744503411675,
      "grad_norm": 2.206113338470459,
      "learning_rate": 1.7053164556962028e-05,
      "loss": 0.09678886830806732,
      "num_input_tokens_seen": 557814,
      "step": 297,
      "train_runtime": 804.415,
      "train_tokens_per_second": 693.441
    },
    {
      "epoch": 0.30123831185241345,
      "grad_norm": 1.7425931692123413,
      "learning_rate": 1.7043037974683544e-05,
      "loss": 0.10033717751502991,
      "num_input_tokens_seen": 560236,
      "step": 298,
      "train_runtime": 806.7918,
      "train_tokens_per_second": 694.4
    },
    {
      "epoch": 0.30224917867071016,
      "grad_norm": 1.9114035367965698,
      "learning_rate": 1.7032911392405066e-05,
      "loss": 0.08700454980134964,
      "num_input_tokens_seen": 561910,
      "step": 299,
      "train_runtime": 808.9448,
      "train_tokens_per_second": 694.621
    },
    {
      "epoch": 0.3032600454890068,
      "grad_norm": 1.9057053327560425,
      "learning_rate": 1.7022784810126585e-05,
      "loss": 0.08345802128314972,
      "num_input_tokens_seen": 563520,
      "step": 300,
      "train_runtime": 811.0375,
      "train_tokens_per_second": 694.814
    },
    {
      "epoch": 0.3042709123073035,
      "grad_norm": 2.048941135406494,
      "learning_rate": 1.70126582278481e-05,
      "loss": 0.1329759657382965,
      "num_input_tokens_seen": 565432,
      "step": 301,
      "train_runtime": 819.6872,
      "train_tokens_per_second": 689.814
    },
    {
      "epoch": 0.30528177912560023,
      "grad_norm": 1.8706474304199219,
      "learning_rate": 1.7002531645569623e-05,
      "loss": 0.09542995691299438,
      "num_input_tokens_seen": 567426,
      "step": 302,
      "train_runtime": 821.8921,
      "train_tokens_per_second": 690.39
    },
    {
      "epoch": 0.3062926459438969,
      "grad_norm": 1.629620909690857,
      "learning_rate": 1.699240506329114e-05,
      "loss": 0.08246508240699768,
      "num_input_tokens_seen": 569566,
      "step": 303,
      "train_runtime": 824.1375,
      "train_tokens_per_second": 691.106
    },
    {
      "epoch": 0.3073035127621936,
      "grad_norm": 2.172229766845703,
      "learning_rate": 1.6982278481012658e-05,
      "loss": 0.10861508548259735,
      "num_input_tokens_seen": 570788,
      "step": 304,
      "train_runtime": 826.1683,
      "train_tokens_per_second": 690.886
    },
    {
      "epoch": 0.3083143795804903,
      "grad_norm": 2.2898049354553223,
      "learning_rate": 1.697215189873418e-05,
      "loss": 0.13094933331012726,
      "num_input_tokens_seen": 572294,
      "step": 305,
      "train_runtime": 828.2848,
      "train_tokens_per_second": 690.939
    },
    {
      "epoch": 0.30932524639878695,
      "grad_norm": 1.6969938278198242,
      "learning_rate": 1.6962025316455696e-05,
      "loss": 0.143634632229805,
      "num_input_tokens_seen": 574604,
      "step": 306,
      "train_runtime": 830.5725,
      "train_tokens_per_second": 691.817
    },
    {
      "epoch": 0.31033611321708365,
      "grad_norm": 2.207098960876465,
      "learning_rate": 1.6951898734177216e-05,
      "loss": 0.11941751837730408,
      "num_input_tokens_seen": 576448,
      "step": 307,
      "train_runtime": 832.7564,
      "train_tokens_per_second": 692.217
    },
    {
      "epoch": 0.31134698003538036,
      "grad_norm": 2.441063404083252,
      "learning_rate": 1.6941772151898738e-05,
      "loss": 0.11468060314655304,
      "num_input_tokens_seen": 578256,
      "step": 308,
      "train_runtime": 834.9513,
      "train_tokens_per_second": 692.563
    },
    {
      "epoch": 0.312357846853677,
      "grad_norm": 2.1768319606781006,
      "learning_rate": 1.6931645569620254e-05,
      "loss": 0.14340467751026154,
      "num_input_tokens_seen": 580722,
      "step": 309,
      "train_runtime": 837.2782,
      "train_tokens_per_second": 693.583
    },
    {
      "epoch": 0.3133687136719737,
      "grad_norm": 1.9249969720840454,
      "learning_rate": 1.6921518987341773e-05,
      "loss": 0.11373825371265411,
      "num_input_tokens_seen": 582702,
      "step": 310,
      "train_runtime": 839.5223,
      "train_tokens_per_second": 694.088
    },
    {
      "epoch": 0.3143795804902704,
      "grad_norm": 2.24229097366333,
      "learning_rate": 1.6911392405063292e-05,
      "loss": 0.14747795462608337,
      "num_input_tokens_seen": 584540,
      "step": 311,
      "train_runtime": 841.7412,
      "train_tokens_per_second": 694.441
    },
    {
      "epoch": 0.3153904473085671,
      "grad_norm": 1.6668611764907837,
      "learning_rate": 1.690126582278481e-05,
      "loss": 0.10077495127916336,
      "num_input_tokens_seen": 586196,
      "step": 312,
      "train_runtime": 843.9087,
      "train_tokens_per_second": 694.62
    },
    {
      "epoch": 0.3164013141268638,
      "grad_norm": 1.5944832563400269,
      "learning_rate": 1.689113924050633e-05,
      "loss": 0.09194497764110565,
      "num_input_tokens_seen": 588350,
      "step": 313,
      "train_runtime": 846.338,
      "train_tokens_per_second": 695.171
    },
    {
      "epoch": 0.3174121809451605,
      "grad_norm": 1.9241758584976196,
      "learning_rate": 1.688101265822785e-05,
      "loss": 0.12003136426210403,
      "num_input_tokens_seen": 590494,
      "step": 314,
      "train_runtime": 848.5843,
      "train_tokens_per_second": 695.858
    },
    {
      "epoch": 0.31842304776345715,
      "grad_norm": 1.8242048025131226,
      "learning_rate": 1.6870886075949368e-05,
      "loss": 0.09342741966247559,
      "num_input_tokens_seen": 592296,
      "step": 315,
      "train_runtime": 850.8026,
      "train_tokens_per_second": 696.162
    },
    {
      "epoch": 0.31943391458175385,
      "grad_norm": 2.2841227054595947,
      "learning_rate": 1.6860759493670887e-05,
      "loss": 0.15476250648498535,
      "num_input_tokens_seen": 594442,
      "step": 316,
      "train_runtime": 853.081,
      "train_tokens_per_second": 696.818
    },
    {
      "epoch": 0.32044478140005056,
      "grad_norm": 1.6433976888656616,
      "learning_rate": 1.6850632911392406e-05,
      "loss": 0.1078416258096695,
      "num_input_tokens_seen": 596700,
      "step": 317,
      "train_runtime": 855.3709,
      "train_tokens_per_second": 697.592
    },
    {
      "epoch": 0.3214556482183472,
      "grad_norm": 1.7912700176239014,
      "learning_rate": 1.6840506329113925e-05,
      "loss": 0.11561059951782227,
      "num_input_tokens_seen": 598890,
      "step": 318,
      "train_runtime": 857.663,
      "train_tokens_per_second": 698.281
    },
    {
      "epoch": 0.3224665150366439,
      "grad_norm": 2.0721402168273926,
      "learning_rate": 1.6830379746835445e-05,
      "loss": 0.1092069149017334,
      "num_input_tokens_seen": 600748,
      "step": 319,
      "train_runtime": 859.9051,
      "train_tokens_per_second": 698.621
    },
    {
      "epoch": 0.3234773818549406,
      "grad_norm": 1.9189189672470093,
      "learning_rate": 1.6820253164556964e-05,
      "loss": 0.11300531029701233,
      "num_input_tokens_seen": 602598,
      "step": 320,
      "train_runtime": 862.0918,
      "train_tokens_per_second": 698.995
    },
    {
      "epoch": 0.3244882486732373,
      "grad_norm": 1.9352511167526245,
      "learning_rate": 1.6810126582278483e-05,
      "loss": 0.09355691075325012,
      "num_input_tokens_seen": 604068,
      "step": 321,
      "train_runtime": 864.2542,
      "train_tokens_per_second": 698.947
    },
    {
      "epoch": 0.325499115491534,
      "grad_norm": 1.7809242010116577,
      "learning_rate": 1.6800000000000002e-05,
      "loss": 0.09653531014919281,
      "num_input_tokens_seen": 605770,
      "step": 322,
      "train_runtime": 866.55,
      "train_tokens_per_second": 699.059
    },
    {
      "epoch": 0.3265099823098307,
      "grad_norm": 1.8602415323257446,
      "learning_rate": 1.678987341772152e-05,
      "loss": 0.12133066356182098,
      "num_input_tokens_seen": 607382,
      "step": 323,
      "train_runtime": 868.6691,
      "train_tokens_per_second": 699.21
    },
    {
      "epoch": 0.32752084912812734,
      "grad_norm": 2.417227268218994,
      "learning_rate": 1.677974683544304e-05,
      "loss": 0.13102781772613525,
      "num_input_tokens_seen": 609506,
      "step": 324,
      "train_runtime": 870.9771,
      "train_tokens_per_second": 699.796
    },
    {
      "epoch": 0.32853171594642405,
      "grad_norm": 1.4471842050552368,
      "learning_rate": 1.676962025316456e-05,
      "loss": 0.07626676559448242,
      "num_input_tokens_seen": 611400,
      "step": 325,
      "train_runtime": 873.1927,
      "train_tokens_per_second": 700.189
    },
    {
      "epoch": 0.32954258276472076,
      "grad_norm": 2.1252057552337646,
      "learning_rate": 1.6759493670886078e-05,
      "loss": 0.11368009448051453,
      "num_input_tokens_seen": 612836,
      "step": 326,
      "train_runtime": 875.2885,
      "train_tokens_per_second": 700.153
    },
    {
      "epoch": 0.3305534495830174,
      "grad_norm": 2.1375629901885986,
      "learning_rate": 1.6749367088607594e-05,
      "loss": 0.11263997852802277,
      "num_input_tokens_seen": 614146,
      "step": 327,
      "train_runtime": 877.3972,
      "train_tokens_per_second": 699.964
    },
    {
      "epoch": 0.3315643164013141,
      "grad_norm": 2.3300492763519287,
      "learning_rate": 1.6739240506329116e-05,
      "loss": 0.09814397245645523,
      "num_input_tokens_seen": 615256,
      "step": 328,
      "train_runtime": 879.4229,
      "train_tokens_per_second": 699.613
    },
    {
      "epoch": 0.3325751832196108,
      "grad_norm": 2.4924209117889404,
      "learning_rate": 1.6729113924050635e-05,
      "loss": 0.10129263997077942,
      "num_input_tokens_seen": 617376,
      "step": 329,
      "train_runtime": 881.6906,
      "train_tokens_per_second": 700.218
    },
    {
      "epoch": 0.33358605003790753,
      "grad_norm": 2.1257572174072266,
      "learning_rate": 1.671898734177215e-05,
      "loss": 0.13214707374572754,
      "num_input_tokens_seen": 619276,
      "step": 330,
      "train_runtime": 883.9211,
      "train_tokens_per_second": 700.601
    },
    {
      "epoch": 0.3345969168562042,
      "grad_norm": 1.9464322328567505,
      "learning_rate": 1.6708860759493674e-05,
      "loss": 0.1226983368396759,
      "num_input_tokens_seen": 621170,
      "step": 331,
      "train_runtime": 892.4949,
      "train_tokens_per_second": 695.993
    },
    {
      "epoch": 0.3356077836745009,
      "grad_norm": 2.048203945159912,
      "learning_rate": 1.6698734177215193e-05,
      "loss": 0.12150146812200546,
      "num_input_tokens_seen": 623702,
      "step": 332,
      "train_runtime": 894.9049,
      "train_tokens_per_second": 696.948
    },
    {
      "epoch": 0.3366186504927976,
      "grad_norm": 1.6402168273925781,
      "learning_rate": 1.668860759493671e-05,
      "loss": 0.08297025412321091,
      "num_input_tokens_seen": 625398,
      "step": 333,
      "train_runtime": 897.098,
      "train_tokens_per_second": 697.135
    },
    {
      "epoch": 0.33762951731109425,
      "grad_norm": 2.152194023132324,
      "learning_rate": 1.667848101265823e-05,
      "loss": 0.10354448854923248,
      "num_input_tokens_seen": 627014,
      "step": 334,
      "train_runtime": 899.2842,
      "train_tokens_per_second": 697.237
    },
    {
      "epoch": 0.33864038412939096,
      "grad_norm": 2.082301378250122,
      "learning_rate": 1.6668354430379746e-05,
      "loss": 0.10761665552854538,
      "num_input_tokens_seen": 628980,
      "step": 335,
      "train_runtime": 901.5443,
      "train_tokens_per_second": 697.67
    },
    {
      "epoch": 0.33965125094768767,
      "grad_norm": 2.3863511085510254,
      "learning_rate": 1.6658227848101266e-05,
      "loss": 0.12909376621246338,
      "num_input_tokens_seen": 630750,
      "step": 336,
      "train_runtime": 903.735,
      "train_tokens_per_second": 697.937
    },
    {
      "epoch": 0.3406621177659843,
      "grad_norm": 1.8604240417480469,
      "learning_rate": 1.6648101265822788e-05,
      "loss": 0.12805582582950592,
      "num_input_tokens_seen": 632848,
      "step": 337,
      "train_runtime": 906.0208,
      "train_tokens_per_second": 698.492
    },
    {
      "epoch": 0.341672984584281,
      "grad_norm": 2.243178367614746,
      "learning_rate": 1.6637974683544304e-05,
      "loss": 0.14108015596866608,
      "num_input_tokens_seen": 634710,
      "step": 338,
      "train_runtime": 908.2705,
      "train_tokens_per_second": 698.812
    },
    {
      "epoch": 0.34268385140257773,
      "grad_norm": 1.607914924621582,
      "learning_rate": 1.6627848101265823e-05,
      "loss": 0.08600659668445587,
      "num_input_tokens_seen": 636400,
      "step": 339,
      "train_runtime": 910.456,
      "train_tokens_per_second": 698.99
    },
    {
      "epoch": 0.3436947182208744,
      "grad_norm": 1.5468930006027222,
      "learning_rate": 1.6617721518987345e-05,
      "loss": 0.0988999679684639,
      "num_input_tokens_seen": 638424,
      "step": 340,
      "train_runtime": 912.7372,
      "train_tokens_per_second": 699.461
    },
    {
      "epoch": 0.3447055850391711,
      "grad_norm": 1.8652654886245728,
      "learning_rate": 1.660759493670886e-05,
      "loss": 0.11248230189085007,
      "num_input_tokens_seen": 639900,
      "step": 341,
      "train_runtime": 914.8966,
      "train_tokens_per_second": 699.423
    },
    {
      "epoch": 0.3457164518574678,
      "grad_norm": 1.8457918167114258,
      "learning_rate": 1.659746835443038e-05,
      "loss": 0.1206580251455307,
      "num_input_tokens_seen": 641800,
      "step": 342,
      "train_runtime": 917.1389,
      "train_tokens_per_second": 699.785
    },
    {
      "epoch": 0.34672731867576445,
      "grad_norm": 2.133465528488159,
      "learning_rate": 1.65873417721519e-05,
      "loss": 0.12648454308509827,
      "num_input_tokens_seen": 644028,
      "step": 343,
      "train_runtime": 919.5273,
      "train_tokens_per_second": 700.39
    },
    {
      "epoch": 0.34773818549406116,
      "grad_norm": 1.5376352071762085,
      "learning_rate": 1.6577215189873418e-05,
      "loss": 0.09699879586696625,
      "num_input_tokens_seen": 645904,
      "step": 344,
      "train_runtime": 921.7298,
      "train_tokens_per_second": 700.752
    },
    {
      "epoch": 0.34874905231235787,
      "grad_norm": 1.5410926342010498,
      "learning_rate": 1.6567088607594937e-05,
      "loss": 0.07477487623691559,
      "num_input_tokens_seen": 647174,
      "step": 345,
      "train_runtime": 923.8135,
      "train_tokens_per_second": 700.546
    },
    {
      "epoch": 0.3497599191306545,
      "grad_norm": 2.047024726867676,
      "learning_rate": 1.6556962025316456e-05,
      "loss": 0.13444937765598297,
      "num_input_tokens_seen": 648866,
      "step": 346,
      "train_runtime": 925.9563,
      "train_tokens_per_second": 700.752
    },
    {
      "epoch": 0.3507707859489512,
      "grad_norm": 1.7477526664733887,
      "learning_rate": 1.6546835443037976e-05,
      "loss": 0.11890025436878204,
      "num_input_tokens_seen": 651554,
      "step": 347,
      "train_runtime": 928.4416,
      "train_tokens_per_second": 701.772
    },
    {
      "epoch": 0.35178165276724793,
      "grad_norm": 2.191701889038086,
      "learning_rate": 1.6536708860759495e-05,
      "loss": 0.09472803771495819,
      "num_input_tokens_seen": 652762,
      "step": 348,
      "train_runtime": 930.5478,
      "train_tokens_per_second": 701.481
    },
    {
      "epoch": 0.3527925195855446,
      "grad_norm": 1.7229115962982178,
      "learning_rate": 1.6526582278481014e-05,
      "loss": 0.07826320827007294,
      "num_input_tokens_seen": 654322,
      "step": 349,
      "train_runtime": 932.6701,
      "train_tokens_per_second": 701.558
    },
    {
      "epoch": 0.3538033864038413,
      "grad_norm": 2.0089187622070312,
      "learning_rate": 1.6516455696202533e-05,
      "loss": 0.0905705913901329,
      "num_input_tokens_seen": 656156,
      "step": 350,
      "train_runtime": 934.8694,
      "train_tokens_per_second": 701.869
    },
    {
      "epoch": 0.354814253222138,
      "grad_norm": 1.7794486284255981,
      "learning_rate": 1.6506329113924052e-05,
      "loss": 0.10507946461439133,
      "num_input_tokens_seen": 657980,
      "step": 351,
      "train_runtime": 937.0891,
      "train_tokens_per_second": 702.153
    },
    {
      "epoch": 0.35582512004043465,
      "grad_norm": 2.046496629714966,
      "learning_rate": 1.649620253164557e-05,
      "loss": 0.12119658291339874,
      "num_input_tokens_seen": 659612,
      "step": 352,
      "train_runtime": 939.2753,
      "train_tokens_per_second": 702.256
    },
    {
      "epoch": 0.35683598685873136,
      "grad_norm": 2.325549364089966,
      "learning_rate": 1.648607594936709e-05,
      "loss": 0.13412690162658691,
      "num_input_tokens_seen": 661560,
      "step": 353,
      "train_runtime": 941.462,
      "train_tokens_per_second": 702.694
    },
    {
      "epoch": 0.35784685367702807,
      "grad_norm": 1.989941954612732,
      "learning_rate": 1.647594936708861e-05,
      "loss": 0.11104798316955566,
      "num_input_tokens_seen": 663524,
      "step": 354,
      "train_runtime": 943.7074,
      "train_tokens_per_second": 703.103
    },
    {
      "epoch": 0.3588577204953247,
      "grad_norm": 1.5388857126235962,
      "learning_rate": 1.6465822784810128e-05,
      "loss": 0.08549122512340546,
      "num_input_tokens_seen": 666486,
      "step": 355,
      "train_runtime": 946.1973,
      "train_tokens_per_second": 704.384
    },
    {
      "epoch": 0.3598685873136214,
      "grad_norm": 1.8720457553863525,
      "learning_rate": 1.6455696202531647e-05,
      "loss": 0.08153025060892105,
      "num_input_tokens_seen": 667976,
      "step": 356,
      "train_runtime": 948.3255,
      "train_tokens_per_second": 704.374
    },
    {
      "epoch": 0.36087945413191813,
      "grad_norm": 1.9136314392089844,
      "learning_rate": 1.6445569620253166e-05,
      "loss": 0.1059972494840622,
      "num_input_tokens_seen": 670104,
      "step": 357,
      "train_runtime": 950.605,
      "train_tokens_per_second": 704.924
    },
    {
      "epoch": 0.3618903209502148,
      "grad_norm": 2.0467233657836914,
      "learning_rate": 1.6435443037974685e-05,
      "loss": 0.08577902615070343,
      "num_input_tokens_seen": 671812,
      "step": 358,
      "train_runtime": 952.7626,
      "train_tokens_per_second": 705.12
    },
    {
      "epoch": 0.3629011877685115,
      "grad_norm": 2.021350145339966,
      "learning_rate": 1.6425316455696205e-05,
      "loss": 0.1395326554775238,
      "num_input_tokens_seen": 673840,
      "step": 359,
      "train_runtime": 955.0312,
      "train_tokens_per_second": 705.569
    },
    {
      "epoch": 0.3639120545868082,
      "grad_norm": 1.7927876710891724,
      "learning_rate": 1.6415189873417724e-05,
      "loss": 0.11609095335006714,
      "num_input_tokens_seen": 675820,
      "step": 360,
      "train_runtime": 957.2431,
      "train_tokens_per_second": 706.007
    },
    {
      "epoch": 0.36492292140510485,
      "grad_norm": 1.5044299364089966,
      "learning_rate": 1.6405063291139243e-05,
      "loss": 0.08769774436950684,
      "num_input_tokens_seen": 677590,
      "step": 361,
      "train_runtime": 965.8041,
      "train_tokens_per_second": 701.581
    },
    {
      "epoch": 0.36593378822340156,
      "grad_norm": 1.9887616634368896,
      "learning_rate": 1.6394936708860762e-05,
      "loss": 0.1319042146205902,
      "num_input_tokens_seen": 679656,
      "step": 362,
      "train_runtime": 968.1531,
      "train_tokens_per_second": 702.013
    },
    {
      "epoch": 0.36694465504169826,
      "grad_norm": 1.77889883518219,
      "learning_rate": 1.638481012658228e-05,
      "loss": 0.11687910556793213,
      "num_input_tokens_seen": 681542,
      "step": 363,
      "train_runtime": 970.4173,
      "train_tokens_per_second": 702.318
    },
    {
      "epoch": 0.36795552185999497,
      "grad_norm": 1.5798107385635376,
      "learning_rate": 1.63746835443038e-05,
      "loss": 0.08784738183021545,
      "num_input_tokens_seen": 683072,
      "step": 364,
      "train_runtime": 972.5466,
      "train_tokens_per_second": 702.354
    },
    {
      "epoch": 0.3689663886782916,
      "grad_norm": 1.9086544513702393,
      "learning_rate": 1.636455696202532e-05,
      "loss": 0.11384311318397522,
      "num_input_tokens_seen": 684780,
      "step": 365,
      "train_runtime": 974.7033,
      "train_tokens_per_second": 702.552
    },
    {
      "epoch": 0.36997725549658833,
      "grad_norm": 1.795804738998413,
      "learning_rate": 1.6354430379746838e-05,
      "loss": 0.14664557576179504,
      "num_input_tokens_seen": 687052,
      "step": 366,
      "train_runtime": 976.998,
      "train_tokens_per_second": 703.228
    },
    {
      "epoch": 0.37098812231488504,
      "grad_norm": 1.6334198713302612,
      "learning_rate": 1.6344303797468354e-05,
      "loss": 0.11831093579530716,
      "num_input_tokens_seen": 689058,
      "step": 367,
      "train_runtime": 979.2153,
      "train_tokens_per_second": 703.684
    },
    {
      "epoch": 0.3719989891331817,
      "grad_norm": 1.8755356073379517,
      "learning_rate": 1.6334177215189876e-05,
      "loss": 0.12714678049087524,
      "num_input_tokens_seen": 690736,
      "step": 368,
      "train_runtime": 981.383,
      "train_tokens_per_second": 703.839
    },
    {
      "epoch": 0.3730098559514784,
      "grad_norm": 1.7126693725585938,
      "learning_rate": 1.6324050632911395e-05,
      "loss": 0.1069984883069992,
      "num_input_tokens_seen": 692596,
      "step": 369,
      "train_runtime": 983.6119,
      "train_tokens_per_second": 704.135
    },
    {
      "epoch": 0.3740207227697751,
      "grad_norm": 2.0715458393096924,
      "learning_rate": 1.631392405063291e-05,
      "loss": 0.12520043551921844,
      "num_input_tokens_seen": 694712,
      "step": 370,
      "train_runtime": 985.8509,
      "train_tokens_per_second": 704.683
    },
    {
      "epoch": 0.37503158958807176,
      "grad_norm": 1.7266870737075806,
      "learning_rate": 1.6303797468354434e-05,
      "loss": 0.08129256218671799,
      "num_input_tokens_seen": 696466,
      "step": 371,
      "train_runtime": 987.9997,
      "train_tokens_per_second": 704.925
    },
    {
      "epoch": 0.37604245640636846,
      "grad_norm": 1.6928164958953857,
      "learning_rate": 1.629367088607595e-05,
      "loss": 0.06832411140203476,
      "num_input_tokens_seen": 697804,
      "step": 372,
      "train_runtime": 990.1476,
      "train_tokens_per_second": 704.747
    },
    {
      "epoch": 0.37705332322466517,
      "grad_norm": 1.93890380859375,
      "learning_rate": 1.6283544303797468e-05,
      "loss": 0.1151122972369194,
      "num_input_tokens_seen": 699614,
      "step": 373,
      "train_runtime": 992.383,
      "train_tokens_per_second": 704.984
    },
    {
      "epoch": 0.3780641900429618,
      "grad_norm": 2.2426564693450928,
      "learning_rate": 1.627341772151899e-05,
      "loss": 0.1247732862830162,
      "num_input_tokens_seen": 700990,
      "step": 374,
      "train_runtime": 994.4776,
      "train_tokens_per_second": 704.883
    },
    {
      "epoch": 0.37907505686125853,
      "grad_norm": 1.6789480447769165,
      "learning_rate": 1.6263291139240506e-05,
      "loss": 0.1177583634853363,
      "num_input_tokens_seen": 703404,
      "step": 375,
      "train_runtime": 996.8104,
      "train_tokens_per_second": 705.655
    },
    {
      "epoch": 0.38008592367955524,
      "grad_norm": 2.000699996948242,
      "learning_rate": 1.6253164556962026e-05,
      "loss": 0.11292397230863571,
      "num_input_tokens_seen": 705094,
      "step": 376,
      "train_runtime": 998.969,
      "train_tokens_per_second": 705.822
    },
    {
      "epoch": 0.3810967904978519,
      "grad_norm": 2.2312796115875244,
      "learning_rate": 1.6243037974683548e-05,
      "loss": 0.13240455090999603,
      "num_input_tokens_seen": 706698,
      "step": 377,
      "train_runtime": 1001.102,
      "train_tokens_per_second": 705.92
    },
    {
      "epoch": 0.3821076573161486,
      "grad_norm": 1.6221849918365479,
      "learning_rate": 1.6232911392405064e-05,
      "loss": 0.0768461525440216,
      "num_input_tokens_seen": 708394,
      "step": 378,
      "train_runtime": 1003.2784,
      "train_tokens_per_second": 706.079
    },
    {
      "epoch": 0.3831185241344453,
      "grad_norm": 1.7459332942962646,
      "learning_rate": 1.6222784810126583e-05,
      "loss": 0.09541502594947815,
      "num_input_tokens_seen": 710096,
      "step": 379,
      "train_runtime": 1005.4229,
      "train_tokens_per_second": 706.266
    },
    {
      "epoch": 0.38412939095274196,
      "grad_norm": 1.5383440256118774,
      "learning_rate": 1.6212658227848102e-05,
      "loss": 0.07174061238765717,
      "num_input_tokens_seen": 711458,
      "step": 380,
      "train_runtime": 1007.5411,
      "train_tokens_per_second": 706.133
    },
    {
      "epoch": 0.38514025777103866,
      "grad_norm": 2.028183698654175,
      "learning_rate": 1.620253164556962e-05,
      "loss": 0.13594409823417664,
      "num_input_tokens_seen": 713746,
      "step": 381,
      "train_runtime": 1009.8843,
      "train_tokens_per_second": 706.76
    },
    {
      "epoch": 0.38615112458933537,
      "grad_norm": 1.3330520391464233,
      "learning_rate": 1.619240506329114e-05,
      "loss": 0.07100304961204529,
      "num_input_tokens_seen": 715434,
      "step": 382,
      "train_runtime": 1012.0256,
      "train_tokens_per_second": 706.933
    },
    {
      "epoch": 0.387161991407632,
      "grad_norm": 1.972743034362793,
      "learning_rate": 1.618227848101266e-05,
      "loss": 0.09860429167747498,
      "num_input_tokens_seen": 717224,
      "step": 383,
      "train_runtime": 1014.1947,
      "train_tokens_per_second": 707.186
    },
    {
      "epoch": 0.38817285822592873,
      "grad_norm": 2.5587127208709717,
      "learning_rate": 1.6172151898734178e-05,
      "loss": 0.13275621831417084,
      "num_input_tokens_seen": 718734,
      "step": 384,
      "train_runtime": 1016.3047,
      "train_tokens_per_second": 707.203
    },
    {
      "epoch": 0.38918372504422544,
      "grad_norm": 1.4623247385025024,
      "learning_rate": 1.6162025316455697e-05,
      "loss": 0.08318155258893967,
      "num_input_tokens_seen": 720200,
      "step": 385,
      "train_runtime": 1018.4238,
      "train_tokens_per_second": 707.171
    },
    {
      "epoch": 0.3901945918625221,
      "grad_norm": 1.7298856973648071,
      "learning_rate": 1.6151898734177216e-05,
      "loss": 0.11590967327356339,
      "num_input_tokens_seen": 722210,
      "step": 386,
      "train_runtime": 1020.6689,
      "train_tokens_per_second": 707.585
    },
    {
      "epoch": 0.3912054586808188,
      "grad_norm": 2.266411066055298,
      "learning_rate": 1.6141772151898735e-05,
      "loss": 0.11143288761377335,
      "num_input_tokens_seen": 724340,
      "step": 387,
      "train_runtime": 1022.9451,
      "train_tokens_per_second": 708.093
    },
    {
      "epoch": 0.3922163254991155,
      "grad_norm": 2.060002565383911,
      "learning_rate": 1.6131645569620255e-05,
      "loss": 0.1141839399933815,
      "num_input_tokens_seen": 725928,
      "step": 388,
      "train_runtime": 1025.0836,
      "train_tokens_per_second": 708.165
    },
    {
      "epoch": 0.39322719231741216,
      "grad_norm": 1.9606693983078003,
      "learning_rate": 1.6121518987341774e-05,
      "loss": 0.1146438866853714,
      "num_input_tokens_seen": 728306,
      "step": 389,
      "train_runtime": 1027.4277,
      "train_tokens_per_second": 708.864
    },
    {
      "epoch": 0.39423805913570886,
      "grad_norm": 1.9957023859024048,
      "learning_rate": 1.6111392405063293e-05,
      "loss": 0.12976482510566711,
      "num_input_tokens_seen": 730248,
      "step": 390,
      "train_runtime": 1029.6729,
      "train_tokens_per_second": 709.204
    },
    {
      "epoch": 0.39524892595400557,
      "grad_norm": 1.8542495965957642,
      "learning_rate": 1.6101265822784812e-05,
      "loss": 0.08730176091194153,
      "num_input_tokens_seen": 732022,
      "step": 391,
      "train_runtime": 1038.1427,
      "train_tokens_per_second": 705.127
    },
    {
      "epoch": 0.3962597927723022,
      "grad_norm": 1.9549685716629028,
      "learning_rate": 1.609113924050633e-05,
      "loss": 0.1409481018781662,
      "num_input_tokens_seen": 734084,
      "step": 392,
      "train_runtime": 1040.4066,
      "train_tokens_per_second": 705.574
    },
    {
      "epoch": 0.39727065959059893,
      "grad_norm": 1.9346776008605957,
      "learning_rate": 1.608101265822785e-05,
      "loss": 0.12512962520122528,
      "num_input_tokens_seen": 735922,
      "step": 393,
      "train_runtime": 1042.767,
      "train_tokens_per_second": 705.74
    },
    {
      "epoch": 0.39828152640889564,
      "grad_norm": 1.7594246864318848,
      "learning_rate": 1.607088607594937e-05,
      "loss": 0.11548963189125061,
      "num_input_tokens_seen": 737962,
      "step": 394,
      "train_runtime": 1044.9739,
      "train_tokens_per_second": 706.201
    },
    {
      "epoch": 0.39929239322719234,
      "grad_norm": 2.4287121295928955,
      "learning_rate": 1.6060759493670888e-05,
      "loss": 0.13242198526859283,
      "num_input_tokens_seen": 739550,
      "step": 395,
      "train_runtime": 1047.1342,
      "train_tokens_per_second": 706.261
    },
    {
      "epoch": 0.400303260045489,
      "grad_norm": 1.869313359260559,
      "learning_rate": 1.6050632911392404e-05,
      "loss": 0.10317026078701019,
      "num_input_tokens_seen": 741172,
      "step": 396,
      "train_runtime": 1049.292,
      "train_tokens_per_second": 706.354
    },
    {
      "epoch": 0.400303260045489,
      "eval_loss": 0.42201629281044006,
      "eval_runtime": 52.301,
      "eval_samples_per_second": 16.807,
      "eval_steps_per_second": 8.413,
      "num_input_tokens_seen": 741172,
      "step": 396
    },
    {
      "epoch": 0.4013141268637857,
      "grad_norm": 1.8665021657943726,
      "learning_rate": 1.6040506329113926e-05,
      "loss": 0.09587148576974869,
      "num_input_tokens_seen": 743004,
      "step": 397,
      "train_runtime": 1103.7991,
      "train_tokens_per_second": 673.133
    },
    {
      "epoch": 0.4023249936820824,
      "grad_norm": 1.5042665004730225,
      "learning_rate": 1.6030379746835445e-05,
      "loss": 0.10045738518238068,
      "num_input_tokens_seen": 744848,
      "step": 398,
      "train_runtime": 1105.9857,
      "train_tokens_per_second": 673.47
    },
    {
      "epoch": 0.40333586050037906,
      "grad_norm": 1.6995333433151245,
      "learning_rate": 1.602025316455696e-05,
      "loss": 0.09934995323419571,
      "num_input_tokens_seen": 746680,
      "step": 399,
      "train_runtime": 1108.1595,
      "train_tokens_per_second": 673.802
    },
    {
      "epoch": 0.40434672731867577,
      "grad_norm": 1.9656271934509277,
      "learning_rate": 1.6010126582278484e-05,
      "loss": 0.11542671918869019,
      "num_input_tokens_seen": 748696,
      "step": 400,
      "train_runtime": 1110.3632,
      "train_tokens_per_second": 674.28
    },
    {
      "epoch": 0.4053575941369725,
      "grad_norm": 1.6320098638534546,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 0.09155505895614624,
      "num_input_tokens_seen": 750616,
      "step": 401,
      "train_runtime": 1112.5727,
      "train_tokens_per_second": 674.667
    },
    {
      "epoch": 0.40636846095526913,
      "grad_norm": 2.454561948776245,
      "learning_rate": 1.598987341772152e-05,
      "loss": 0.13649535179138184,
      "num_input_tokens_seen": 752420,
      "step": 402,
      "train_runtime": 1114.7875,
      "train_tokens_per_second": 674.945
    },
    {
      "epoch": 0.40737932777356584,
      "grad_norm": 2.0516269207000732,
      "learning_rate": 1.597974683544304e-05,
      "loss": 0.1292344033718109,
      "num_input_tokens_seen": 754336,
      "step": 403,
      "train_runtime": 1117.0455,
      "train_tokens_per_second": 675.296
    },
    {
      "epoch": 0.40839019459186254,
      "grad_norm": 2.109393835067749,
      "learning_rate": 1.5969620253164557e-05,
      "loss": 0.10809981822967529,
      "num_input_tokens_seen": 756468,
      "step": 404,
      "train_runtime": 1119.3543,
      "train_tokens_per_second": 675.807
    },
    {
      "epoch": 0.4094010614101592,
      "grad_norm": 1.70687735080719,
      "learning_rate": 1.5959493670886076e-05,
      "loss": 0.11761138588190079,
      "num_input_tokens_seen": 758516,
      "step": 405,
      "train_runtime": 1121.6274,
      "train_tokens_per_second": 676.264
    },
    {
      "epoch": 0.4104119282284559,
      "grad_norm": 1.7862193584442139,
      "learning_rate": 1.5949367088607598e-05,
      "loss": 0.10888718068599701,
      "num_input_tokens_seen": 760364,
      "step": 406,
      "train_runtime": 1123.8172,
      "train_tokens_per_second": 676.59
    },
    {
      "epoch": 0.4114227950467526,
      "grad_norm": 1.822191596031189,
      "learning_rate": 1.5939240506329114e-05,
      "loss": 0.11547520756721497,
      "num_input_tokens_seen": 762296,
      "step": 407,
      "train_runtime": 1126.0691,
      "train_tokens_per_second": 676.953
    },
    {
      "epoch": 0.41243366186504926,
      "grad_norm": 1.9484902620315552,
      "learning_rate": 1.5929113924050633e-05,
      "loss": 0.09211841225624084,
      "num_input_tokens_seen": 764002,
      "step": 408,
      "train_runtime": 1128.2247,
      "train_tokens_per_second": 677.172
    },
    {
      "epoch": 0.41344452868334597,
      "grad_norm": 1.9491668939590454,
      "learning_rate": 1.5918987341772155e-05,
      "loss": 0.144774928689003,
      "num_input_tokens_seen": 766318,
      "step": 409,
      "train_runtime": 1130.5753,
      "train_tokens_per_second": 677.812
    },
    {
      "epoch": 0.4144553955016427,
      "grad_norm": 1.5557540655136108,
      "learning_rate": 1.590886075949367e-05,
      "loss": 0.09754966199398041,
      "num_input_tokens_seen": 768542,
      "step": 410,
      "train_runtime": 1132.8389,
      "train_tokens_per_second": 678.421
    },
    {
      "epoch": 0.41546626231993933,
      "grad_norm": 1.9235962629318237,
      "learning_rate": 1.589873417721519e-05,
      "loss": 0.09587027132511139,
      "num_input_tokens_seen": 770400,
      "step": 411,
      "train_runtime": 1135.0152,
      "train_tokens_per_second": 678.757
    },
    {
      "epoch": 0.41647712913823604,
      "grad_norm": 1.5368192195892334,
      "learning_rate": 1.588860759493671e-05,
      "loss": 0.08561362326145172,
      "num_input_tokens_seen": 773532,
      "step": 412,
      "train_runtime": 1137.5146,
      "train_tokens_per_second": 680.019
    },
    {
      "epoch": 0.41748799595653274,
      "grad_norm": 1.5619041919708252,
      "learning_rate": 1.5878481012658228e-05,
      "loss": 0.08629076182842255,
      "num_input_tokens_seen": 775576,
      "step": 413,
      "train_runtime": 1139.7497,
      "train_tokens_per_second": 680.479
    },
    {
      "epoch": 0.4184988627748294,
      "grad_norm": 1.8506624698638916,
      "learning_rate": 1.5868354430379747e-05,
      "loss": 0.11986834555864334,
      "num_input_tokens_seen": 777670,
      "step": 414,
      "train_runtime": 1142.0165,
      "train_tokens_per_second": 680.962
    },
    {
      "epoch": 0.4195097295931261,
      "grad_norm": 2.1654317378997803,
      "learning_rate": 1.5858227848101266e-05,
      "loss": 0.11728420853614807,
      "num_input_tokens_seen": 779290,
      "step": 415,
      "train_runtime": 1144.2012,
      "train_tokens_per_second": 681.078
    },
    {
      "epoch": 0.4205205964114228,
      "grad_norm": 2.0820565223693848,
      "learning_rate": 1.5848101265822786e-05,
      "loss": 0.1193307489156723,
      "num_input_tokens_seen": 781116,
      "step": 416,
      "train_runtime": 1146.3847,
      "train_tokens_per_second": 681.373
    },
    {
      "epoch": 0.42153146322971946,
      "grad_norm": 1.7809218168258667,
      "learning_rate": 1.5837974683544305e-05,
      "loss": 0.1084236353635788,
      "num_input_tokens_seen": 783262,
      "step": 417,
      "train_runtime": 1148.6239,
      "train_tokens_per_second": 681.913
    },
    {
      "epoch": 0.42254233004801617,
      "grad_norm": 1.6279551982879639,
      "learning_rate": 1.5827848101265824e-05,
      "loss": 0.09959094971418381,
      "num_input_tokens_seen": 785110,
      "step": 418,
      "train_runtime": 1150.8261,
      "train_tokens_per_second": 682.214
    },
    {
      "epoch": 0.4235531968663129,
      "grad_norm": 1.692862868309021,
      "learning_rate": 1.5817721518987343e-05,
      "loss": 0.11396458745002747,
      "num_input_tokens_seen": 787038,
      "step": 419,
      "train_runtime": 1153.0624,
      "train_tokens_per_second": 682.563
    },
    {
      "epoch": 0.4245640636846095,
      "grad_norm": 1.6826294660568237,
      "learning_rate": 1.5807594936708862e-05,
      "loss": 0.11504846066236496,
      "num_input_tokens_seen": 788866,
      "step": 420,
      "train_runtime": 1155.2779,
      "train_tokens_per_second": 682.837
    },
    {
      "epoch": 0.42557493050290623,
      "grad_norm": 2.1463863849639893,
      "learning_rate": 1.579746835443038e-05,
      "loss": 0.09922569990158081,
      "num_input_tokens_seen": 790578,
      "step": 421,
      "train_runtime": 1164.0445,
      "train_tokens_per_second": 679.165
    },
    {
      "epoch": 0.42658579732120294,
      "grad_norm": 1.575871467590332,
      "learning_rate": 1.57873417721519e-05,
      "loss": 0.10171511769294739,
      "num_input_tokens_seen": 792632,
      "step": 422,
      "train_runtime": 1166.3793,
      "train_tokens_per_second": 679.566
    },
    {
      "epoch": 0.4275966641394996,
      "grad_norm": 1.4514565467834473,
      "learning_rate": 1.577721518987342e-05,
      "loss": 0.0752696841955185,
      "num_input_tokens_seen": 794344,
      "step": 423,
      "train_runtime": 1168.567,
      "train_tokens_per_second": 679.759
    },
    {
      "epoch": 0.4286075309577963,
      "grad_norm": 1.4515169858932495,
      "learning_rate": 1.5767088607594938e-05,
      "loss": 0.0848160907626152,
      "num_input_tokens_seen": 796212,
      "step": 424,
      "train_runtime": 1170.7731,
      "train_tokens_per_second": 680.074
    },
    {
      "epoch": 0.429618397776093,
      "grad_norm": 1.5068243741989136,
      "learning_rate": 1.5756962025316457e-05,
      "loss": 0.10230260342359543,
      "num_input_tokens_seen": 798458,
      "step": 425,
      "train_runtime": 1173.0971,
      "train_tokens_per_second": 680.641
    },
    {
      "epoch": 0.4306292645943897,
      "grad_norm": 1.7472387552261353,
      "learning_rate": 1.5746835443037976e-05,
      "loss": 0.12000877410173416,
      "num_input_tokens_seen": 800190,
      "step": 426,
      "train_runtime": 1175.3301,
      "train_tokens_per_second": 680.821
    },
    {
      "epoch": 0.43164013141268637,
      "grad_norm": 1.6070133447647095,
      "learning_rate": 1.5736708860759495e-05,
      "loss": 0.08834629505872726,
      "num_input_tokens_seen": 801656,
      "step": 427,
      "train_runtime": 1177.4292,
      "train_tokens_per_second": 680.853
    },
    {
      "epoch": 0.4326509982309831,
      "grad_norm": 1.5826572179794312,
      "learning_rate": 1.5726582278481015e-05,
      "loss": 0.07955733686685562,
      "num_input_tokens_seen": 803426,
      "step": 428,
      "train_runtime": 1179.6092,
      "train_tokens_per_second": 681.095
    },
    {
      "epoch": 0.4336618650492798,
      "grad_norm": 1.5440285205841064,
      "learning_rate": 1.5716455696202534e-05,
      "loss": 0.0863671600818634,
      "num_input_tokens_seen": 805396,
      "step": 429,
      "train_runtime": 1181.7924,
      "train_tokens_per_second": 681.504
    },
    {
      "epoch": 0.43467273186757643,
      "grad_norm": 1.633183479309082,
      "learning_rate": 1.5706329113924053e-05,
      "loss": 0.11554452031850815,
      "num_input_tokens_seen": 807228,
      "step": 430,
      "train_runtime": 1183.9752,
      "train_tokens_per_second": 681.795
    },
    {
      "epoch": 0.43568359868587314,
      "grad_norm": 2.0593886375427246,
      "learning_rate": 1.5696202531645572e-05,
      "loss": 0.12381618469953537,
      "num_input_tokens_seen": 808880,
      "step": 431,
      "train_runtime": 1186.1214,
      "train_tokens_per_second": 681.954
    },
    {
      "epoch": 0.43669446550416985,
      "grad_norm": 1.516738772392273,
      "learning_rate": 1.568607594936709e-05,
      "loss": 0.10749244689941406,
      "num_input_tokens_seen": 811672,
      "step": 432,
      "train_runtime": 1188.561,
      "train_tokens_per_second": 682.903
    },
    {
      "epoch": 0.4377053323224665,
      "grad_norm": 1.5923534631729126,
      "learning_rate": 1.567594936708861e-05,
      "loss": 0.09094808995723724,
      "num_input_tokens_seen": 813512,
      "step": 433,
      "train_runtime": 1190.7375,
      "train_tokens_per_second": 683.2
    },
    {
      "epoch": 0.4387161991407632,
      "grad_norm": 1.855827808380127,
      "learning_rate": 1.566582278481013e-05,
      "loss": 0.10945839434862137,
      "num_input_tokens_seen": 815038,
      "step": 434,
      "train_runtime": 1192.8699,
      "train_tokens_per_second": 683.258
    },
    {
      "epoch": 0.4397270659590599,
      "grad_norm": 1.7897955179214478,
      "learning_rate": 1.5655696202531648e-05,
      "loss": 0.11381515115499496,
      "num_input_tokens_seen": 816976,
      "step": 435,
      "train_runtime": 1195.0553,
      "train_tokens_per_second": 683.63
    },
    {
      "epoch": 0.44073793277735657,
      "grad_norm": 1.8408046960830688,
      "learning_rate": 1.5645569620253164e-05,
      "loss": 0.09024758636951447,
      "num_input_tokens_seen": 818870,
      "step": 436,
      "train_runtime": 1197.2687,
      "train_tokens_per_second": 683.948
    },
    {
      "epoch": 0.4417487995956533,
      "grad_norm": 1.8731918334960938,
      "learning_rate": 1.5635443037974686e-05,
      "loss": 0.12296129763126373,
      "num_input_tokens_seen": 821004,
      "step": 437,
      "train_runtime": 1199.5552,
      "train_tokens_per_second": 684.424
    },
    {
      "epoch": 0.44275966641395,
      "grad_norm": 1.323952317237854,
      "learning_rate": 1.5625316455696205e-05,
      "loss": 0.080072320997715,
      "num_input_tokens_seen": 823306,
      "step": 438,
      "train_runtime": 1201.8619,
      "train_tokens_per_second": 685.025
    },
    {
      "epoch": 0.44377053323224663,
      "grad_norm": 1.9585412740707397,
      "learning_rate": 1.561518987341772e-05,
      "loss": 0.13973820209503174,
      "num_input_tokens_seen": 825712,
      "step": 439,
      "train_runtime": 1204.2047,
      "train_tokens_per_second": 685.691
    },
    {
      "epoch": 0.44478140005054334,
      "grad_norm": 1.9043835401535034,
      "learning_rate": 1.5605063291139244e-05,
      "loss": 0.10461394488811493,
      "num_input_tokens_seen": 827680,
      "step": 440,
      "train_runtime": 1206.4511,
      "train_tokens_per_second": 686.045
    },
    {
      "epoch": 0.44579226686884005,
      "grad_norm": 1.8181251287460327,
      "learning_rate": 1.5594936708860763e-05,
      "loss": 0.11697743833065033,
      "num_input_tokens_seen": 829328,
      "step": 441,
      "train_runtime": 1208.5951,
      "train_tokens_per_second": 686.192
    },
    {
      "epoch": 0.4468031336871367,
      "grad_norm": 1.7808847427368164,
      "learning_rate": 1.558481012658228e-05,
      "loss": 0.10102672129869461,
      "num_input_tokens_seen": 831254,
      "step": 442,
      "train_runtime": 1210.8,
      "train_tokens_per_second": 686.533
    },
    {
      "epoch": 0.4478140005054334,
      "grad_norm": 1.8276588916778564,
      "learning_rate": 1.55746835443038e-05,
      "loss": 0.12833836674690247,
      "num_input_tokens_seen": 833076,
      "step": 443,
      "train_runtime": 1213.0006,
      "train_tokens_per_second": 686.789
    },
    {
      "epoch": 0.4488248673237301,
      "grad_norm": 1.671140193939209,
      "learning_rate": 1.5564556962025317e-05,
      "loss": 0.11130937933921814,
      "num_input_tokens_seen": 835272,
      "step": 444,
      "train_runtime": 1215.3353,
      "train_tokens_per_second": 687.277
    },
    {
      "epoch": 0.44983573414202677,
      "grad_norm": 1.6901848316192627,
      "learning_rate": 1.5554430379746836e-05,
      "loss": 0.11136844754219055,
      "num_input_tokens_seen": 837422,
      "step": 445,
      "train_runtime": 1217.5852,
      "train_tokens_per_second": 687.773
    },
    {
      "epoch": 0.4508466009603235,
      "grad_norm": 1.4702390432357788,
      "learning_rate": 1.5544303797468358e-05,
      "loss": 0.08984880149364471,
      "num_input_tokens_seen": 839660,
      "step": 446,
      "train_runtime": 1219.848,
      "train_tokens_per_second": 688.332
    },
    {
      "epoch": 0.4518574677786202,
      "grad_norm": 2.1821863651275635,
      "learning_rate": 1.5534177215189874e-05,
      "loss": 0.14034874737262726,
      "num_input_tokens_seen": 841170,
      "step": 447,
      "train_runtime": 1221.9606,
      "train_tokens_per_second": 688.377
    },
    {
      "epoch": 0.45286833459691683,
      "grad_norm": 2.0638959407806396,
      "learning_rate": 1.5524050632911393e-05,
      "loss": 0.14524506032466888,
      "num_input_tokens_seen": 843086,
      "step": 448,
      "train_runtime": 1224.1944,
      "train_tokens_per_second": 688.686
    },
    {
      "epoch": 0.45387920141521354,
      "grad_norm": 1.34815514087677,
      "learning_rate": 1.5513924050632912e-05,
      "loss": 0.07662313431501389,
      "num_input_tokens_seen": 845328,
      "step": 449,
      "train_runtime": 1226.4751,
      "train_tokens_per_second": 689.234
    },
    {
      "epoch": 0.45489006823351025,
      "grad_norm": 2.104788064956665,
      "learning_rate": 1.550379746835443e-05,
      "loss": 0.12482360005378723,
      "num_input_tokens_seen": 847832,
      "step": 450,
      "train_runtime": 1228.9051,
      "train_tokens_per_second": 689.908
    },
    {
      "epoch": 0.4559009350518069,
      "grad_norm": 1.55012047290802,
      "learning_rate": 1.549367088607595e-05,
      "loss": 0.06931520998477936,
      "num_input_tokens_seen": 849910,
      "step": 451,
      "train_runtime": 1237.5726,
      "train_tokens_per_second": 686.756
    },
    {
      "epoch": 0.4569118018701036,
      "grad_norm": 1.5841842889785767,
      "learning_rate": 1.548354430379747e-05,
      "loss": 0.1137673556804657,
      "num_input_tokens_seen": 852294,
      "step": 452,
      "train_runtime": 1239.936,
      "train_tokens_per_second": 687.369
    },
    {
      "epoch": 0.4579226686884003,
      "grad_norm": 1.9309735298156738,
      "learning_rate": 1.5473417721518988e-05,
      "loss": 0.10654686391353607,
      "num_input_tokens_seen": 854262,
      "step": 453,
      "train_runtime": 1242.1737,
      "train_tokens_per_second": 687.715
    },
    {
      "epoch": 0.45893353550669697,
      "grad_norm": 2.4330663681030273,
      "learning_rate": 1.5463291139240507e-05,
      "loss": 0.1234653890132904,
      "num_input_tokens_seen": 856256,
      "step": 454,
      "train_runtime": 1244.3796,
      "train_tokens_per_second": 688.099
    },
    {
      "epoch": 0.4599444023249937,
      "grad_norm": 1.5877577066421509,
      "learning_rate": 1.5453164556962026e-05,
      "loss": 0.09403739124536514,
      "num_input_tokens_seen": 857960,
      "step": 455,
      "train_runtime": 1246.5842,
      "train_tokens_per_second": 688.249
    },
    {
      "epoch": 0.4609552691432904,
      "grad_norm": 1.748447060585022,
      "learning_rate": 1.5443037974683546e-05,
      "loss": 0.0888945460319519,
      "num_input_tokens_seen": 859844,
      "step": 456,
      "train_runtime": 1248.8123,
      "train_tokens_per_second": 688.529
    },
    {
      "epoch": 0.4619661359615871,
      "grad_norm": 1.9015161991119385,
      "learning_rate": 1.5432911392405065e-05,
      "loss": 0.13397936522960663,
      "num_input_tokens_seen": 862234,
      "step": 457,
      "train_runtime": 1251.1803,
      "train_tokens_per_second": 689.136
    },
    {
      "epoch": 0.46297700277988374,
      "grad_norm": 1.9944270849227905,
      "learning_rate": 1.5422784810126584e-05,
      "loss": 0.17752207815647125,
      "num_input_tokens_seen": 864526,
      "step": 458,
      "train_runtime": 1253.5349,
      "train_tokens_per_second": 689.67
    },
    {
      "epoch": 0.46398786959818045,
      "grad_norm": 1.6891884803771973,
      "learning_rate": 1.5412658227848103e-05,
      "loss": 0.0922621637582779,
      "num_input_tokens_seen": 866254,
      "step": 459,
      "train_runtime": 1255.7202,
      "train_tokens_per_second": 689.846
    },
    {
      "epoch": 0.46499873641647715,
      "grad_norm": 2.05324649810791,
      "learning_rate": 1.5402531645569622e-05,
      "loss": 0.12507976591587067,
      "num_input_tokens_seen": 868044,
      "step": 460,
      "train_runtime": 1257.9077,
      "train_tokens_per_second": 690.07
    },
    {
      "epoch": 0.4660096032347738,
      "grad_norm": 1.9538702964782715,
      "learning_rate": 1.539240506329114e-05,
      "loss": 0.13201306760311127,
      "num_input_tokens_seen": 869820,
      "step": 461,
      "train_runtime": 1260.0664,
      "train_tokens_per_second": 690.297
    },
    {
      "epoch": 0.4670204700530705,
      "grad_norm": 1.5424325466156006,
      "learning_rate": 1.538227848101266e-05,
      "loss": 0.07851071655750275,
      "num_input_tokens_seen": 871334,
      "step": 462,
      "train_runtime": 1262.1907,
      "train_tokens_per_second": 690.335
    },
    {
      "epoch": 0.4680313368713672,
      "grad_norm": 1.6234872341156006,
      "learning_rate": 1.537215189873418e-05,
      "loss": 0.11618975549936295,
      "num_input_tokens_seen": 873256,
      "step": 463,
      "train_runtime": 1264.4267,
      "train_tokens_per_second": 690.634
    },
    {
      "epoch": 0.4690422036896639,
      "grad_norm": 1.761287808418274,
      "learning_rate": 1.5362025316455698e-05,
      "loss": 0.11727702617645264,
      "num_input_tokens_seen": 875286,
      "step": 464,
      "train_runtime": 1266.6396,
      "train_tokens_per_second": 691.03
    },
    {
      "epoch": 0.4700530705079606,
      "grad_norm": 1.9724807739257812,
      "learning_rate": 1.5351898734177217e-05,
      "loss": 0.10613028705120087,
      "num_input_tokens_seen": 876870,
      "step": 465,
      "train_runtime": 1268.7845,
      "train_tokens_per_second": 691.11
    },
    {
      "epoch": 0.4710639373262573,
      "grad_norm": 1.6279855966567993,
      "learning_rate": 1.5341772151898736e-05,
      "loss": 0.1028311476111412,
      "num_input_tokens_seen": 878840,
      "step": 466,
      "train_runtime": 1271.0233,
      "train_tokens_per_second": 691.443
    },
    {
      "epoch": 0.47207480414455394,
      "grad_norm": 1.61843740940094,
      "learning_rate": 1.5331645569620255e-05,
      "loss": 0.08738434314727783,
      "num_input_tokens_seen": 880468,
      "step": 467,
      "train_runtime": 1273.1698,
      "train_tokens_per_second": 691.556
    },
    {
      "epoch": 0.47308567096285065,
      "grad_norm": 1.5544933080673218,
      "learning_rate": 1.532151898734177e-05,
      "loss": 0.09057053923606873,
      "num_input_tokens_seen": 882314,
      "step": 468,
      "train_runtime": 1275.3465,
      "train_tokens_per_second": 691.823
    },
    {
      "epoch": 0.47409653778114735,
      "grad_norm": 1.851883053779602,
      "learning_rate": 1.5311392405063294e-05,
      "loss": 0.10593780875205994,
      "num_input_tokens_seen": 884162,
      "step": 469,
      "train_runtime": 1277.5257,
      "train_tokens_per_second": 692.089
    },
    {
      "epoch": 0.475107404599444,
      "grad_norm": 1.8330367803573608,
      "learning_rate": 1.5301265822784813e-05,
      "loss": 0.09430311620235443,
      "num_input_tokens_seen": 885872,
      "step": 470,
      "train_runtime": 1279.6991,
      "train_tokens_per_second": 692.25
    },
    {
      "epoch": 0.4761182714177407,
      "grad_norm": 1.7879048585891724,
      "learning_rate": 1.529113924050633e-05,
      "loss": 0.10557078570127487,
      "num_input_tokens_seen": 887478,
      "step": 471,
      "train_runtime": 1281.8474,
      "train_tokens_per_second": 692.343
    },
    {
      "epoch": 0.4771291382360374,
      "grad_norm": 2.0346546173095703,
      "learning_rate": 1.528101265822785e-05,
      "loss": 0.11926902830600739,
      "num_input_tokens_seen": 889464,
      "step": 472,
      "train_runtime": 1284.0791,
      "train_tokens_per_second": 692.686
    },
    {
      "epoch": 0.47814000505433407,
      "grad_norm": 1.7213999032974243,
      "learning_rate": 1.527088607594937e-05,
      "loss": 0.09899803251028061,
      "num_input_tokens_seen": 890990,
      "step": 473,
      "train_runtime": 1286.1848,
      "train_tokens_per_second": 692.739
    },
    {
      "epoch": 0.4791508718726308,
      "grad_norm": 1.8353174924850464,
      "learning_rate": 1.5260759493670886e-05,
      "loss": 0.09149063378572464,
      "num_input_tokens_seen": 892672,
      "step": 474,
      "train_runtime": 1288.3919,
      "train_tokens_per_second": 692.857
    },
    {
      "epoch": 0.4801617386909275,
      "grad_norm": 1.7517122030258179,
      "learning_rate": 1.5250632911392406e-05,
      "loss": 0.10418742150068283,
      "num_input_tokens_seen": 894660,
      "step": 475,
      "train_runtime": 1290.6374,
      "train_tokens_per_second": 693.192
    },
    {
      "epoch": 0.48117260550922414,
      "grad_norm": 1.8344651460647583,
      "learning_rate": 1.5240506329113924e-05,
      "loss": 0.11346778273582458,
      "num_input_tokens_seen": 896764,
      "step": 476,
      "train_runtime": 1292.8751,
      "train_tokens_per_second": 693.62
    },
    {
      "epoch": 0.48218347232752085,
      "grad_norm": 2.2210395336151123,
      "learning_rate": 1.5230379746835445e-05,
      "loss": 0.10573138296604156,
      "num_input_tokens_seen": 898270,
      "step": 477,
      "train_runtime": 1295.0283,
      "train_tokens_per_second": 693.63
    },
    {
      "epoch": 0.48319433914581755,
      "grad_norm": 1.4662727117538452,
      "learning_rate": 1.5220253164556964e-05,
      "loss": 0.11092646420001984,
      "num_input_tokens_seen": 900776,
      "step": 478,
      "train_runtime": 1297.3784,
      "train_tokens_per_second": 694.305
    },
    {
      "epoch": 0.4842052059641142,
      "grad_norm": 2.1169073581695557,
      "learning_rate": 1.5210126582278481e-05,
      "loss": 0.09660357236862183,
      "num_input_tokens_seen": 902464,
      "step": 479,
      "train_runtime": 1299.5258,
      "train_tokens_per_second": 694.456
    },
    {
      "epoch": 0.4852160727824109,
      "grad_norm": 2.007495403289795,
      "learning_rate": 1.5200000000000002e-05,
      "loss": 0.11861640959978104,
      "num_input_tokens_seen": 904536,
      "step": 480,
      "train_runtime": 1301.7942,
      "train_tokens_per_second": 694.838
    },
    {
      "epoch": 0.4862269396007076,
      "grad_norm": 1.7178399562835693,
      "learning_rate": 1.5189873417721521e-05,
      "loss": 0.11039911210536957,
      "num_input_tokens_seen": 906342,
      "step": 481,
      "train_runtime": 1310.239,
      "train_tokens_per_second": 691.738
    },
    {
      "epoch": 0.48723780641900427,
      "grad_norm": 1.519370198249817,
      "learning_rate": 1.5179746835443038e-05,
      "loss": 0.08855210989713669,
      "num_input_tokens_seen": 908624,
      "step": 482,
      "train_runtime": 1312.565,
      "train_tokens_per_second": 692.251
    },
    {
      "epoch": 0.488248673237301,
      "grad_norm": 1.793309211730957,
      "learning_rate": 1.5169620253164559e-05,
      "loss": 0.09917984157800674,
      "num_input_tokens_seen": 910822,
      "step": 483,
      "train_runtime": 1314.84,
      "train_tokens_per_second": 692.725
    },
    {
      "epoch": 0.4892595400555977,
      "grad_norm": 2.017210006713867,
      "learning_rate": 1.5159493670886076e-05,
      "loss": 0.12656812369823456,
      "num_input_tokens_seen": 912538,
      "step": 484,
      "train_runtime": 1317.0247,
      "train_tokens_per_second": 692.878
    },
    {
      "epoch": 0.49027040687389434,
      "grad_norm": 1.6859157085418701,
      "learning_rate": 1.5149367088607596e-05,
      "loss": 0.08337490260601044,
      "num_input_tokens_seen": 914398,
      "step": 485,
      "train_runtime": 1319.2571,
      "train_tokens_per_second": 693.116
    },
    {
      "epoch": 0.49128127369219105,
      "grad_norm": 1.7310889959335327,
      "learning_rate": 1.5139240506329116e-05,
      "loss": 0.09235775470733643,
      "num_input_tokens_seen": 915784,
      "step": 486,
      "train_runtime": 1321.3648,
      "train_tokens_per_second": 693.059
    },
    {
      "epoch": 0.49229214051048775,
      "grad_norm": 1.7842894792556763,
      "learning_rate": 1.5129113924050634e-05,
      "loss": 0.1012832522392273,
      "num_input_tokens_seen": 917832,
      "step": 487,
      "train_runtime": 1323.5975,
      "train_tokens_per_second": 693.437
    },
    {
      "epoch": 0.49330300732878446,
      "grad_norm": 1.850504994392395,
      "learning_rate": 1.5118987341772153e-05,
      "loss": 0.14272227883338928,
      "num_input_tokens_seen": 919822,
      "step": 488,
      "train_runtime": 1325.8314,
      "train_tokens_per_second": 693.77
    },
    {
      "epoch": 0.4943138741470811,
      "grad_norm": 1.1433686017990112,
      "learning_rate": 1.5108860759493674e-05,
      "loss": 0.06848733127117157,
      "num_input_tokens_seen": 921798,
      "step": 489,
      "train_runtime": 1328.0818,
      "train_tokens_per_second": 694.082
    },
    {
      "epoch": 0.4953247409653778,
      "grad_norm": 2.013387441635132,
      "learning_rate": 1.5098734177215191e-05,
      "loss": 0.09337301552295685,
      "num_input_tokens_seen": 923560,
      "step": 490,
      "train_runtime": 1330.2725,
      "train_tokens_per_second": 694.264
    },
    {
      "epoch": 0.4963356077836745,
      "grad_norm": 2.0009877681732178,
      "learning_rate": 1.508860759493671e-05,
      "loss": 0.08730406314134598,
      "num_input_tokens_seen": 925226,
      "step": 491,
      "train_runtime": 1332.4696,
      "train_tokens_per_second": 694.369
    },
    {
      "epoch": 0.4973464746019712,
      "grad_norm": 1.698068618774414,
      "learning_rate": 1.5078481012658227e-05,
      "loss": 0.10106261819601059,
      "num_input_tokens_seen": 926804,
      "step": 492,
      "train_runtime": 1334.597,
      "train_tokens_per_second": 694.445
    },
    {
      "epoch": 0.4983573414202679,
      "grad_norm": 1.6992688179016113,
      "learning_rate": 1.5068354430379748e-05,
      "loss": 0.11859212070703506,
      "num_input_tokens_seen": 928740,
      "step": 493,
      "train_runtime": 1336.8252,
      "train_tokens_per_second": 694.736
    },
    {
      "epoch": 0.4993682082385646,
      "grad_norm": 2.515361785888672,
      "learning_rate": 1.5058227848101267e-05,
      "loss": 0.11421169340610504,
      "num_input_tokens_seen": 930264,
      "step": 494,
      "train_runtime": 1338.9348,
      "train_tokens_per_second": 694.779
    },
    {
      "epoch": 0.5003790750568613,
      "grad_norm": 2.014190196990967,
      "learning_rate": 1.5048101265822785e-05,
      "loss": 0.10539563000202179,
      "num_input_tokens_seen": 931754,
      "step": 495,
      "train_runtime": 1341.1158,
      "train_tokens_per_second": 694.76
    },
    {
      "epoch": 0.501389941875158,
      "grad_norm": 1.7498775720596313,
      "learning_rate": 1.5037974683544306e-05,
      "loss": 0.13124173879623413,
      "num_input_tokens_seen": 933608,
      "step": 496,
      "train_runtime": 1343.3599,
      "train_tokens_per_second": 694.98
    },
    {
      "epoch": 0.5024008086934546,
      "grad_norm": 1.7500488758087158,
      "learning_rate": 1.5027848101265825e-05,
      "loss": 0.11794276535511017,
      "num_input_tokens_seen": 935584,
      "step": 497,
      "train_runtime": 1345.5841,
      "train_tokens_per_second": 695.3
    },
    {
      "epoch": 0.5034116755117514,
      "grad_norm": 2.0071139335632324,
      "learning_rate": 1.5017721518987342e-05,
      "loss": 0.098157599568367,
      "num_input_tokens_seen": 937248,
      "step": 498,
      "train_runtime": 1347.7816,
      "train_tokens_per_second": 695.401
    },
    {
      "epoch": 0.504422542330048,
      "grad_norm": 1.5707734823226929,
      "learning_rate": 1.5007594936708863e-05,
      "loss": 0.09266889840364456,
      "num_input_tokens_seen": 939192,
      "step": 499,
      "train_runtime": 1350.0337,
      "train_tokens_per_second": 695.68
    },
    {
      "epoch": 0.5054334091483447,
      "grad_norm": 1.875820279121399,
      "learning_rate": 1.499746835443038e-05,
      "loss": 0.10333205759525299,
      "num_input_tokens_seen": 941142,
      "step": 500,
      "train_runtime": 1352.2952,
      "train_tokens_per_second": 695.959
    },
    {
      "epoch": 0.5064442759666414,
      "grad_norm": 1.783033013343811,
      "learning_rate": 1.49873417721519e-05,
      "loss": 0.0842694416642189,
      "num_input_tokens_seen": 942764,
      "step": 501,
      "train_runtime": 1354.5035,
      "train_tokens_per_second": 696.022
    },
    {
      "epoch": 0.5074551427849381,
      "grad_norm": 1.964139461517334,
      "learning_rate": 1.497721518987342e-05,
      "loss": 0.12120130658149719,
      "num_input_tokens_seen": 944456,
      "step": 502,
      "train_runtime": 1356.6572,
      "train_tokens_per_second": 696.164
    },
    {
      "epoch": 0.5084660096032347,
      "grad_norm": 1.8904763460159302,
      "learning_rate": 1.4967088607594937e-05,
      "loss": 0.1054663211107254,
      "num_input_tokens_seen": 946438,
      "step": 503,
      "train_runtime": 1358.8965,
      "train_tokens_per_second": 696.475
    },
    {
      "epoch": 0.5094768764215315,
      "grad_norm": 2.0818839073181152,
      "learning_rate": 1.4956962025316456e-05,
      "loss": 0.09397480636835098,
      "num_input_tokens_seen": 947942,
      "step": 504,
      "train_runtime": 1361.0449,
      "train_tokens_per_second": 696.481
    },
    {
      "epoch": 0.5104877432398282,
      "grad_norm": 2.1881535053253174,
      "learning_rate": 1.4946835443037977e-05,
      "loss": 0.08975251019001007,
      "num_input_tokens_seen": 950170,
      "step": 505,
      "train_runtime": 1363.3241,
      "train_tokens_per_second": 696.951
    },
    {
      "epoch": 0.5114986100581248,
      "grad_norm": 1.7329323291778564,
      "learning_rate": 1.4936708860759495e-05,
      "loss": 0.1096678152680397,
      "num_input_tokens_seen": 952068,
      "step": 506,
      "train_runtime": 1365.5013,
      "train_tokens_per_second": 697.23
    },
    {
      "epoch": 0.5125094768764216,
      "grad_norm": 1.7487717866897583,
      "learning_rate": 1.4926582278481014e-05,
      "loss": 0.08918820321559906,
      "num_input_tokens_seen": 954106,
      "step": 507,
      "train_runtime": 1367.7081,
      "train_tokens_per_second": 697.595
    },
    {
      "epoch": 0.5135203436947182,
      "grad_norm": 1.7221266031265259,
      "learning_rate": 1.4916455696202531e-05,
      "loss": 0.10318662226200104,
      "num_input_tokens_seen": 956278,
      "step": 508,
      "train_runtime": 1369.9429,
      "train_tokens_per_second": 698.042
    },
    {
      "epoch": 0.5145312105130149,
      "grad_norm": 1.8457298278808594,
      "learning_rate": 1.4906329113924052e-05,
      "loss": 0.13894438743591309,
      "num_input_tokens_seen": 957986,
      "step": 509,
      "train_runtime": 1372.1068,
      "train_tokens_per_second": 698.186
    },
    {
      "epoch": 0.5155420773313116,
      "grad_norm": 1.6781812906265259,
      "learning_rate": 1.4896202531645571e-05,
      "loss": 0.07831525057554245,
      "num_input_tokens_seen": 960114,
      "step": 510,
      "train_runtime": 1374.3639,
      "train_tokens_per_second": 698.588
    },
    {
      "epoch": 0.5165529441496083,
      "grad_norm": 1.417349100112915,
      "learning_rate": 1.4886075949367088e-05,
      "loss": 0.08226348459720612,
      "num_input_tokens_seen": 961990,
      "step": 511,
      "train_runtime": 1383.0206,
      "train_tokens_per_second": 695.572
    },
    {
      "epoch": 0.5175638109679049,
      "grad_norm": 1.7875341176986694,
      "learning_rate": 1.487594936708861e-05,
      "loss": 0.09959499537944794,
      "num_input_tokens_seen": 964296,
      "step": 512,
      "train_runtime": 1385.3328,
      "train_tokens_per_second": 696.075
    },
    {
      "epoch": 0.5185746777862017,
      "grad_norm": 1.9528840780258179,
      "learning_rate": 1.4865822784810128e-05,
      "loss": 0.1227521002292633,
      "num_input_tokens_seen": 965920,
      "step": 513,
      "train_runtime": 1387.5031,
      "train_tokens_per_second": 696.157
    },
    {
      "epoch": 0.5195855446044984,
      "grad_norm": 1.9798704385757446,
      "learning_rate": 1.4855696202531646e-05,
      "loss": 0.11006204038858414,
      "num_input_tokens_seen": 967894,
      "step": 514,
      "train_runtime": 1389.7419,
      "train_tokens_per_second": 696.456
    },
    {
      "epoch": 0.520596411422795,
      "grad_norm": 1.7312660217285156,
      "learning_rate": 1.4845569620253166e-05,
      "loss": 0.10807745158672333,
      "num_input_tokens_seen": 970006,
      "step": 515,
      "train_runtime": 1392.0082,
      "train_tokens_per_second": 696.839
    },
    {
      "epoch": 0.5216072782410918,
      "grad_norm": 1.6591758728027344,
      "learning_rate": 1.4835443037974684e-05,
      "loss": 0.09685654193162918,
      "num_input_tokens_seen": 971940,
      "step": 516,
      "train_runtime": 1394.222,
      "train_tokens_per_second": 697.12
    },
    {
      "epoch": 0.5226181450593884,
      "grad_norm": 1.5925546884536743,
      "learning_rate": 1.4825316455696203e-05,
      "loss": 0.10817191749811172,
      "num_input_tokens_seen": 973842,
      "step": 517,
      "train_runtime": 1396.4262,
      "train_tokens_per_second": 697.382
    },
    {
      "epoch": 0.5236290118776851,
      "grad_norm": 2.093906879425049,
      "learning_rate": 1.4815189873417724e-05,
      "loss": 0.14283540844917297,
      "num_input_tokens_seen": 975408,
      "step": 518,
      "train_runtime": 1398.548,
      "train_tokens_per_second": 697.443
    },
    {
      "epoch": 0.5246398786959818,
      "grad_norm": 2.173994779586792,
      "learning_rate": 1.4805063291139241e-05,
      "loss": 0.11630765348672867,
      "num_input_tokens_seen": 976992,
      "step": 519,
      "train_runtime": 1400.704,
      "train_tokens_per_second": 697.501
    },
    {
      "epoch": 0.5256507455142785,
      "grad_norm": 1.920044183731079,
      "learning_rate": 1.479493670886076e-05,
      "loss": 0.11933885514736176,
      "num_input_tokens_seen": 979234,
      "step": 520,
      "train_runtime": 1402.9703,
      "train_tokens_per_second": 697.972
    },
    {
      "epoch": 0.5266616123325751,
      "grad_norm": 2.08705997467041,
      "learning_rate": 1.4784810126582281e-05,
      "loss": 0.09583388268947601,
      "num_input_tokens_seen": 980844,
      "step": 521,
      "train_runtime": 1405.1419,
      "train_tokens_per_second": 698.039
    },
    {
      "epoch": 0.5276724791508719,
      "grad_norm": 1.824843406677246,
      "learning_rate": 1.4774683544303798e-05,
      "loss": 0.1340876817703247,
      "num_input_tokens_seen": 982764,
      "step": 522,
      "train_runtime": 1407.4225,
      "train_tokens_per_second": 698.272
    },
    {
      "epoch": 0.5286833459691685,
      "grad_norm": 1.854954481124878,
      "learning_rate": 1.4764556962025317e-05,
      "loss": 0.1237156018614769,
      "num_input_tokens_seen": 984420,
      "step": 523,
      "train_runtime": 1409.6078,
      "train_tokens_per_second": 698.364
    },
    {
      "epoch": 0.5296942127874652,
      "grad_norm": 2.0649797916412354,
      "learning_rate": 1.4754430379746836e-05,
      "loss": 0.13949503004550934,
      "num_input_tokens_seen": 986348,
      "step": 524,
      "train_runtime": 1411.8723,
      "train_tokens_per_second": 698.61
    },
    {
      "epoch": 0.530705079605762,
      "grad_norm": 1.5614497661590576,
      "learning_rate": 1.4744303797468356e-05,
      "loss": 0.09315375983715057,
      "num_input_tokens_seen": 988886,
      "step": 525,
      "train_runtime": 1414.2887,
      "train_tokens_per_second": 699.211
    },
    {
      "epoch": 0.5317159464240586,
      "grad_norm": 1.839135766029358,
      "learning_rate": 1.4734177215189875e-05,
      "loss": 0.1236603781580925,
      "num_input_tokens_seen": 991078,
      "step": 526,
      "train_runtime": 1416.5703,
      "train_tokens_per_second": 699.632
    },
    {
      "epoch": 0.5327268132423553,
      "grad_norm": 1.7902334928512573,
      "learning_rate": 1.4724050632911394e-05,
      "loss": 0.07974381744861603,
      "num_input_tokens_seen": 992786,
      "step": 527,
      "train_runtime": 1418.7559,
      "train_tokens_per_second": 699.758
    },
    {
      "epoch": 0.533737680060652,
      "grad_norm": 1.4455580711364746,
      "learning_rate": 1.4713924050632913e-05,
      "loss": 0.07908736169338226,
      "num_input_tokens_seen": 994364,
      "step": 528,
      "train_runtime": 1420.9037,
      "train_tokens_per_second": 699.811
    },
    {
      "epoch": 0.5347485468789487,
      "grad_norm": 1.8058191537857056,
      "learning_rate": 1.4703797468354432e-05,
      "loss": 0.08228744566440582,
      "num_input_tokens_seen": 996310,
      "step": 529,
      "train_runtime": 1423.126,
      "train_tokens_per_second": 700.086
    },
    {
      "epoch": 0.5357594136972453,
      "grad_norm": 1.8447226285934448,
      "learning_rate": 1.4693670886075951e-05,
      "loss": 0.11537911742925644,
      "num_input_tokens_seen": 998072,
      "step": 530,
      "train_runtime": 1425.3353,
      "train_tokens_per_second": 700.237
    },
    {
      "epoch": 0.5367702805155421,
      "grad_norm": 2.0417678356170654,
      "learning_rate": 1.468354430379747e-05,
      "loss": 0.09737853705883026,
      "num_input_tokens_seen": 999422,
      "step": 531,
      "train_runtime": 1427.434,
      "train_tokens_per_second": 700.153
    },
    {
      "epoch": 0.5377811473338387,
      "grad_norm": 2.273249864578247,
      "learning_rate": 1.4673417721518987e-05,
      "loss": 0.1221286803483963,
      "num_input_tokens_seen": 1001204,
      "step": 532,
      "train_runtime": 1429.6,
      "train_tokens_per_second": 700.339
    },
    {
      "epoch": 0.5387920141521354,
      "grad_norm": 2.3385322093963623,
      "learning_rate": 1.4663291139240508e-05,
      "loss": 0.09718027710914612,
      "num_input_tokens_seen": 1002712,
      "step": 533,
      "train_runtime": 1431.7801,
      "train_tokens_per_second": 700.325
    },
    {
      "epoch": 0.5398028809704322,
      "grad_norm": 2.52339506149292,
      "learning_rate": 1.4653164556962027e-05,
      "loss": 0.15465357899665833,
      "num_input_tokens_seen": 1004918,
      "step": 534,
      "train_runtime": 1434.1426,
      "train_tokens_per_second": 700.71
    },
    {
      "epoch": 0.5408137477887288,
      "grad_norm": 1.8176788091659546,
      "learning_rate": 1.4643037974683545e-05,
      "loss": 0.08890235424041748,
      "num_input_tokens_seen": 1006830,
      "step": 535,
      "train_runtime": 1436.3976,
      "train_tokens_per_second": 700.941
    },
    {
      "epoch": 0.5418246146070256,
      "grad_norm": 1.8820675611495972,
      "learning_rate": 1.4632911392405065e-05,
      "loss": 0.09948692470788956,
      "num_input_tokens_seen": 1008674,
      "step": 536,
      "train_runtime": 1438.6146,
      "train_tokens_per_second": 701.143
    },
    {
      "epoch": 0.5428354814253222,
      "grad_norm": 1.618376612663269,
      "learning_rate": 1.4622784810126585e-05,
      "loss": 0.10095593333244324,
      "num_input_tokens_seen": 1010454,
      "step": 537,
      "train_runtime": 1440.8478,
      "train_tokens_per_second": 701.291
    },
    {
      "epoch": 0.5438463482436189,
      "grad_norm": 1.6165224313735962,
      "learning_rate": 1.4612658227848102e-05,
      "loss": 0.09104248136281967,
      "num_input_tokens_seen": 1012450,
      "step": 538,
      "train_runtime": 1443.0792,
      "train_tokens_per_second": 701.59
    },
    {
      "epoch": 0.5448572150619156,
      "grad_norm": 1.7190626859664917,
      "learning_rate": 1.4602531645569623e-05,
      "loss": 0.10027013719081879,
      "num_input_tokens_seen": 1014482,
      "step": 539,
      "train_runtime": 1445.3367,
      "train_tokens_per_second": 701.9
    },
    {
      "epoch": 0.5458680818802123,
      "grad_norm": 1.9768813848495483,
      "learning_rate": 1.459240506329114e-05,
      "loss": 0.1081099808216095,
      "num_input_tokens_seen": 1016358,
      "step": 540,
      "train_runtime": 1447.5243,
      "train_tokens_per_second": 702.135
    },
    {
      "epoch": 0.546878948698509,
      "grad_norm": 1.793695092201233,
      "learning_rate": 1.458227848101266e-05,
      "loss": 0.10678578913211823,
      "num_input_tokens_seen": 1018128,
      "step": 541,
      "train_runtime": 1456.5568,
      "train_tokens_per_second": 698.996
    },
    {
      "epoch": 0.5478898155168057,
      "grad_norm": 1.36508309841156,
      "learning_rate": 1.457215189873418e-05,
      "loss": 0.09679611772298813,
      "num_input_tokens_seen": 1020734,
      "step": 542,
      "train_runtime": 1458.9916,
      "train_tokens_per_second": 699.616
    },
    {
      "epoch": 0.5489006823351024,
      "grad_norm": 2.0271501541137695,
      "learning_rate": 1.4562025316455697e-05,
      "loss": 0.1297168731689453,
      "num_input_tokens_seen": 1022694,
      "step": 543,
      "train_runtime": 1461.2267,
      "train_tokens_per_second": 699.887
    },
    {
      "epoch": 0.549911549153399,
      "grad_norm": 1.5494493246078491,
      "learning_rate": 1.4551898734177216e-05,
      "loss": 0.08901668339967728,
      "num_input_tokens_seen": 1024474,
      "step": 544,
      "train_runtime": 1463.401,
      "train_tokens_per_second": 700.064
    },
    {
      "epoch": 0.5509224159716958,
      "grad_norm": 1.9661540985107422,
      "learning_rate": 1.4541772151898734e-05,
      "loss": 0.12233540415763855,
      "num_input_tokens_seen": 1026564,
      "step": 545,
      "train_runtime": 1465.6521,
      "train_tokens_per_second": 700.414
    },
    {
      "epoch": 0.5519332827899924,
      "grad_norm": 1.566672921180725,
      "learning_rate": 1.4531645569620255e-05,
      "loss": 0.09095793217420578,
      "num_input_tokens_seen": 1028902,
      "step": 546,
      "train_runtime": 1467.9763,
      "train_tokens_per_second": 700.898
    },
    {
      "epoch": 0.5529441496082891,
      "grad_norm": 1.8163119554519653,
      "learning_rate": 1.4521518987341774e-05,
      "loss": 0.1103871688246727,
      "num_input_tokens_seen": 1030790,
      "step": 547,
      "train_runtime": 1470.162,
      "train_tokens_per_second": 701.14
    },
    {
      "epoch": 0.5539550164265858,
      "grad_norm": 2.0378549098968506,
      "learning_rate": 1.4511392405063291e-05,
      "loss": 0.1106257289648056,
      "num_input_tokens_seen": 1032664,
      "step": 548,
      "train_runtime": 1472.374,
      "train_tokens_per_second": 701.36
    },
    {
      "epoch": 0.5549658832448825,
      "grad_norm": 1.6189203262329102,
      "learning_rate": 1.4501265822784812e-05,
      "loss": 0.10852859914302826,
      "num_input_tokens_seen": 1035148,
      "step": 549,
      "train_runtime": 1474.6956,
      "train_tokens_per_second": 701.94
    },
    {
      "epoch": 0.5559767500631791,
      "grad_norm": 2.446418046951294,
      "learning_rate": 1.4491139240506331e-05,
      "loss": 0.1286158263683319,
      "num_input_tokens_seen": 1037144,
      "step": 550,
      "train_runtime": 1476.9385,
      "train_tokens_per_second": 702.226
    },
    {
      "epoch": 0.5569876168814759,
      "grad_norm": 1.8347103595733643,
      "learning_rate": 1.4481012658227848e-05,
      "loss": 0.09553838521242142,
      "num_input_tokens_seen": 1038914,
      "step": 551,
      "train_runtime": 1479.1227,
      "train_tokens_per_second": 702.385
    },
    {
      "epoch": 0.5579984836997726,
      "grad_norm": 1.836276888847351,
      "learning_rate": 1.4470886075949369e-05,
      "loss": 0.08587778359651566,
      "num_input_tokens_seen": 1040926,
      "step": 552,
      "train_runtime": 1481.4253,
      "train_tokens_per_second": 702.652
    },
    {
      "epoch": 0.5590093505180692,
      "grad_norm": 1.637007713317871,
      "learning_rate": 1.4460759493670887e-05,
      "loss": 0.08323580026626587,
      "num_input_tokens_seen": 1042626,
      "step": 553,
      "train_runtime": 1483.6596,
      "train_tokens_per_second": 702.739
    },
    {
      "epoch": 0.560020217336366,
      "grad_norm": 1.7831929922103882,
      "learning_rate": 1.4450632911392406e-05,
      "loss": 0.09477714449167252,
      "num_input_tokens_seen": 1044082,
      "step": 554,
      "train_runtime": 1485.7933,
      "train_tokens_per_second": 702.71
    },
    {
      "epoch": 0.5610310841546626,
      "grad_norm": 1.4510189294815063,
      "learning_rate": 1.4440506329113926e-05,
      "loss": 0.08061984181404114,
      "num_input_tokens_seen": 1045858,
      "step": 555,
      "train_runtime": 1487.9774,
      "train_tokens_per_second": 702.872
    },
    {
      "epoch": 0.5620419509729593,
      "grad_norm": 1.6569571495056152,
      "learning_rate": 1.4430379746835444e-05,
      "loss": 0.09726974368095398,
      "num_input_tokens_seen": 1047734,
      "step": 556,
      "train_runtime": 1490.1929,
      "train_tokens_per_second": 703.086
    },
    {
      "epoch": 0.563052817791256,
      "grad_norm": 1.685828447341919,
      "learning_rate": 1.4420253164556963e-05,
      "loss": 0.08759375661611557,
      "num_input_tokens_seen": 1049074,
      "step": 557,
      "train_runtime": 1492.2888,
      "train_tokens_per_second": 702.997
    },
    {
      "epoch": 0.5640636846095527,
      "grad_norm": 1.9097645282745361,
      "learning_rate": 1.4410126582278484e-05,
      "loss": 0.1268608123064041,
      "num_input_tokens_seen": 1050914,
      "step": 558,
      "train_runtime": 1494.4527,
      "train_tokens_per_second": 703.21
    },
    {
      "epoch": 0.5650745514278493,
      "grad_norm": 1.92098069190979,
      "learning_rate": 1.4400000000000001e-05,
      "loss": 0.11083795130252838,
      "num_input_tokens_seen": 1052764,
      "step": 559,
      "train_runtime": 1496.6575,
      "train_tokens_per_second": 703.41
    },
    {
      "epoch": 0.5660854182461461,
      "grad_norm": 1.5865694284439087,
      "learning_rate": 1.438987341772152e-05,
      "loss": 0.10960868746042252,
      "num_input_tokens_seen": 1054490,
      "step": 560,
      "train_runtime": 1498.858,
      "train_tokens_per_second": 703.529
    },
    {
      "epoch": 0.5670962850644428,
      "grad_norm": 1.6917788982391357,
      "learning_rate": 1.4379746835443038e-05,
      "loss": 0.09427762776613235,
      "num_input_tokens_seen": 1056460,
      "step": 561,
      "train_runtime": 1501.1064,
      "train_tokens_per_second": 703.788
    },
    {
      "epoch": 0.5681071518827394,
      "grad_norm": 1.9456087350845337,
      "learning_rate": 1.4369620253164558e-05,
      "loss": 0.12319564819335938,
      "num_input_tokens_seen": 1058084,
      "step": 562,
      "train_runtime": 1503.2658,
      "train_tokens_per_second": 703.857
    },
    {
      "epoch": 0.5691180187010362,
      "grad_norm": 1.8830398321151733,
      "learning_rate": 1.4359493670886077e-05,
      "loss": 0.11210659891366959,
      "num_input_tokens_seen": 1060308,
      "step": 563,
      "train_runtime": 1505.5587,
      "train_tokens_per_second": 704.262
    },
    {
      "epoch": 0.5701288855193328,
      "grad_norm": 1.7035287618637085,
      "learning_rate": 1.4349367088607595e-05,
      "loss": 0.09595833718776703,
      "num_input_tokens_seen": 1061998,
      "step": 564,
      "train_runtime": 1507.7226,
      "train_tokens_per_second": 704.372
    },
    {
      "epoch": 0.5711397523376295,
      "grad_norm": 2.0166547298431396,
      "learning_rate": 1.4339240506329116e-05,
      "loss": 0.12182815372943878,
      "num_input_tokens_seen": 1063624,
      "step": 565,
      "train_runtime": 1509.8871,
      "train_tokens_per_second": 704.439
    },
    {
      "epoch": 0.5721506191559262,
      "grad_norm": 1.8745771646499634,
      "learning_rate": 1.4329113924050635e-05,
      "loss": 0.0933096632361412,
      "num_input_tokens_seen": 1065408,
      "step": 566,
      "train_runtime": 1512.1608,
      "train_tokens_per_second": 704.56
    },
    {
      "epoch": 0.5731614859742229,
      "grad_norm": 1.516402006149292,
      "learning_rate": 1.4318987341772152e-05,
      "loss": 0.1060028076171875,
      "num_input_tokens_seen": 1067676,
      "step": 567,
      "train_runtime": 1514.5027,
      "train_tokens_per_second": 704.968
    },
    {
      "epoch": 0.5741723527925195,
      "grad_norm": 1.649111270904541,
      "learning_rate": 1.4308860759493673e-05,
      "loss": 0.09301664680242538,
      "num_input_tokens_seen": 1069738,
      "step": 568,
      "train_runtime": 1516.7405,
      "train_tokens_per_second": 705.287
    },
    {
      "epoch": 0.5751832196108163,
      "grad_norm": 2.101097822189331,
      "learning_rate": 1.429873417721519e-05,
      "loss": 0.09789851307868958,
      "num_input_tokens_seen": 1071286,
      "step": 569,
      "train_runtime": 1518.884,
      "train_tokens_per_second": 705.311
    },
    {
      "epoch": 0.576194086429113,
      "grad_norm": 1.9389387369155884,
      "learning_rate": 1.428860759493671e-05,
      "loss": 0.09537738561630249,
      "num_input_tokens_seen": 1073526,
      "step": 570,
      "train_runtime": 1521.1483,
      "train_tokens_per_second": 705.734
    },
    {
      "epoch": 0.5772049532474096,
      "grad_norm": 1.776051640510559,
      "learning_rate": 1.427848101265823e-05,
      "loss": 0.14023849368095398,
      "num_input_tokens_seen": 1076090,
      "step": 571,
      "train_runtime": 1531.0194,
      "train_tokens_per_second": 702.859
    },
    {
      "epoch": 0.5782158200657064,
      "grad_norm": 1.9601185321807861,
      "learning_rate": 1.4268354430379747e-05,
      "loss": 0.125863716006279,
      "num_input_tokens_seen": 1078008,
      "step": 572,
      "train_runtime": 1533.2106,
      "train_tokens_per_second": 703.105
    },
    {
      "epoch": 0.579226686884003,
      "grad_norm": 2.311521053314209,
      "learning_rate": 1.4258227848101267e-05,
      "loss": 0.1442338079214096,
      "num_input_tokens_seen": 1079674,
      "step": 573,
      "train_runtime": 1535.3805,
      "train_tokens_per_second": 703.196
    },
    {
      "epoch": 0.5802375537022997,
      "grad_norm": 1.456743597984314,
      "learning_rate": 1.4248101265822787e-05,
      "loss": 0.09219858050346375,
      "num_input_tokens_seen": 1081812,
      "step": 574,
      "train_runtime": 1537.6609,
      "train_tokens_per_second": 703.544
    },
    {
      "epoch": 0.5812484205205964,
      "grad_norm": 1.5369138717651367,
      "learning_rate": 1.4237974683544305e-05,
      "loss": 0.07453219592571259,
      "num_input_tokens_seen": 1084224,
      "step": 575,
      "train_runtime": 1539.9672,
      "train_tokens_per_second": 704.057
    },
    {
      "epoch": 0.5822592873388931,
      "grad_norm": 2.125049114227295,
      "learning_rate": 1.4227848101265824e-05,
      "loss": 0.10877341032028198,
      "num_input_tokens_seen": 1085852,
      "step": 576,
      "train_runtime": 1542.1179,
      "train_tokens_per_second": 704.13
    },
    {
      "epoch": 0.5832701541571897,
      "grad_norm": 1.5754179954528809,
      "learning_rate": 1.4217721518987341e-05,
      "loss": 0.09778103232383728,
      "num_input_tokens_seen": 1088162,
      "step": 577,
      "train_runtime": 1544.4524,
      "train_tokens_per_second": 704.562
    },
    {
      "epoch": 0.5842810209754865,
      "grad_norm": 1.6735082864761353,
      "learning_rate": 1.4207594936708862e-05,
      "loss": 0.08625233918428421,
      "num_input_tokens_seen": 1089830,
      "step": 578,
      "train_runtime": 1546.6006,
      "train_tokens_per_second": 704.662
    },
    {
      "epoch": 0.5852918877937832,
      "grad_norm": 1.7386170625686646,
      "learning_rate": 1.4197468354430381e-05,
      "loss": 0.099061518907547,
      "num_input_tokens_seen": 1091450,
      "step": 579,
      "train_runtime": 1548.7146,
      "train_tokens_per_second": 704.746
    },
    {
      "epoch": 0.5863027546120798,
      "grad_norm": 1.5631988048553467,
      "learning_rate": 1.4187341772151898e-05,
      "loss": 0.10082489997148514,
      "num_input_tokens_seen": 1093596,
      "step": 580,
      "train_runtime": 1550.9634,
      "train_tokens_per_second": 705.108
    },
    {
      "epoch": 0.5873136214303766,
      "grad_norm": 1.7483549118041992,
      "learning_rate": 1.417721518987342e-05,
      "loss": 0.11064396798610687,
      "num_input_tokens_seen": 1095482,
      "step": 581,
      "train_runtime": 1553.1398,
      "train_tokens_per_second": 705.334
    },
    {
      "epoch": 0.5883244882486732,
      "grad_norm": 1.8992518186569214,
      "learning_rate": 1.4167088607594938e-05,
      "loss": 0.11621041595935822,
      "num_input_tokens_seen": 1097222,
      "step": 582,
      "train_runtime": 1555.3131,
      "train_tokens_per_second": 705.467
    },
    {
      "epoch": 0.5893353550669699,
      "grad_norm": 1.825927734375,
      "learning_rate": 1.4156962025316456e-05,
      "loss": 0.10483191162347794,
      "num_input_tokens_seen": 1099366,
      "step": 583,
      "train_runtime": 1557.5525,
      "train_tokens_per_second": 705.829
    },
    {
      "epoch": 0.5903462218852666,
      "grad_norm": 1.885602355003357,
      "learning_rate": 1.4146835443037976e-05,
      "loss": 0.12778542935848236,
      "num_input_tokens_seen": 1101232,
      "step": 584,
      "train_runtime": 1559.7249,
      "train_tokens_per_second": 706.042
    },
    {
      "epoch": 0.5913570887035633,
      "grad_norm": 2.019209623336792,
      "learning_rate": 1.4136708860759494e-05,
      "loss": 0.14271020889282227,
      "num_input_tokens_seen": 1103144,
      "step": 585,
      "train_runtime": 1561.9296,
      "train_tokens_per_second": 706.27
    },
    {
      "epoch": 0.59236795552186,
      "grad_norm": 1.578254222869873,
      "learning_rate": 1.4126582278481013e-05,
      "loss": 0.099541574716568,
      "num_input_tokens_seen": 1105134,
      "step": 586,
      "train_runtime": 1564.1348,
      "train_tokens_per_second": 706.547
    },
    {
      "epoch": 0.5933788223401567,
      "grad_norm": 1.7057363986968994,
      "learning_rate": 1.4116455696202534e-05,
      "loss": 0.11436079442501068,
      "num_input_tokens_seen": 1106752,
      "step": 587,
      "train_runtime": 1566.3434,
      "train_tokens_per_second": 706.583
    },
    {
      "epoch": 0.5943896891584534,
      "grad_norm": 2.0448598861694336,
      "learning_rate": 1.4106329113924051e-05,
      "loss": 0.09593261778354645,
      "num_input_tokens_seen": 1108152,
      "step": 588,
      "train_runtime": 1568.4306,
      "train_tokens_per_second": 706.536
    },
    {
      "epoch": 0.59540055597675,
      "grad_norm": 1.9079177379608154,
      "learning_rate": 1.409620253164557e-05,
      "loss": 0.09666799008846283,
      "num_input_tokens_seen": 1109994,
      "step": 589,
      "train_runtime": 1570.6073,
      "train_tokens_per_second": 706.729
    },
    {
      "epoch": 0.5964114227950468,
      "grad_norm": 1.7336835861206055,
      "learning_rate": 1.4086075949367091e-05,
      "loss": 0.10393904149532318,
      "num_input_tokens_seen": 1112366,
      "step": 590,
      "train_runtime": 1572.9235,
      "train_tokens_per_second": 707.197
    },
    {
      "epoch": 0.5974222896133434,
      "grad_norm": 1.9441266059875488,
      "learning_rate": 1.4075949367088608e-05,
      "loss": 0.11883354932069778,
      "num_input_tokens_seen": 1114230,
      "step": 591,
      "train_runtime": 1575.1086,
      "train_tokens_per_second": 707.399
    },
    {
      "epoch": 0.5984331564316401,
      "grad_norm": 2.1221354007720947,
      "learning_rate": 1.4065822784810127e-05,
      "loss": 0.11868761479854584,
      "num_input_tokens_seen": 1115886,
      "step": 592,
      "train_runtime": 1577.2972,
      "train_tokens_per_second": 707.467
    },
    {
      "epoch": 0.5994440232499368,
      "grad_norm": 1.576833724975586,
      "learning_rate": 1.4055696202531647e-05,
      "loss": 0.08950021117925644,
      "num_input_tokens_seen": 1117622,
      "step": 593,
      "train_runtime": 1579.4803,
      "train_tokens_per_second": 707.588
    },
    {
      "epoch": 0.6004548900682335,
      "grad_norm": 1.7010703086853027,
      "learning_rate": 1.4045569620253166e-05,
      "loss": 0.0935027077794075,
      "num_input_tokens_seen": 1119962,
      "step": 594,
      "train_runtime": 1581.7795,
      "train_tokens_per_second": 708.039
    },
    {
      "epoch": 0.6004548900682335,
      "eval_loss": 0.40808728337287903,
      "eval_runtime": 51.7638,
      "eval_samples_per_second": 16.981,
      "eval_steps_per_second": 8.5,
      "num_input_tokens_seen": 1119962,
      "step": 594
    },
    {
      "epoch": 0.6014657568865301,
      "grad_norm": 2.1104953289031982,
      "learning_rate": 1.4035443037974685e-05,
      "loss": 0.1295463591814041,
      "num_input_tokens_seen": 1121502,
      "step": 595,
      "train_runtime": 1635.6536,
      "train_tokens_per_second": 685.66
    },
    {
      "epoch": 0.6024766237048269,
      "grad_norm": 1.6768834590911865,
      "learning_rate": 1.4025316455696202e-05,
      "loss": 0.09007760137319565,
      "num_input_tokens_seen": 1123712,
      "step": 596,
      "train_runtime": 1637.8906,
      "train_tokens_per_second": 686.073
    },
    {
      "epoch": 0.6034874905231236,
      "grad_norm": 1.860482931137085,
      "learning_rate": 1.4015189873417723e-05,
      "loss": 0.09720242768526077,
      "num_input_tokens_seen": 1125700,
      "step": 597,
      "train_runtime": 1640.1269,
      "train_tokens_per_second": 686.349
    },
    {
      "epoch": 0.6044983573414203,
      "grad_norm": 1.9982037544250488,
      "learning_rate": 1.4005063291139242e-05,
      "loss": 0.13862156867980957,
      "num_input_tokens_seen": 1127428,
      "step": 598,
      "train_runtime": 1642.2882,
      "train_tokens_per_second": 686.498
    },
    {
      "epoch": 0.605509224159717,
      "grad_norm": 3.5887558460235596,
      "learning_rate": 1.399493670886076e-05,
      "loss": 0.11241520941257477,
      "num_input_tokens_seen": 1128956,
      "step": 599,
      "train_runtime": 1644.4132,
      "train_tokens_per_second": 686.54
    },
    {
      "epoch": 0.6065200909780136,
      "grad_norm": 1.8000441789627075,
      "learning_rate": 1.398481012658228e-05,
      "loss": 0.1194303035736084,
      "num_input_tokens_seen": 1130624,
      "step": 600,
      "train_runtime": 1646.5699,
      "train_tokens_per_second": 686.654
    },
    {
      "epoch": 0.6075309577963104,
      "grad_norm": 1.8652498722076416,
      "learning_rate": 1.3974683544303798e-05,
      "loss": 0.11491431295871735,
      "num_input_tokens_seen": 1132692,
      "step": 601,
      "train_runtime": 1656.0443,
      "train_tokens_per_second": 683.974
    },
    {
      "epoch": 0.608541824614607,
      "grad_norm": 2.5798568725585938,
      "learning_rate": 1.3964556962025317e-05,
      "loss": 0.1649254560470581,
      "num_input_tokens_seen": 1134506,
      "step": 602,
      "train_runtime": 1658.4002,
      "train_tokens_per_second": 684.097
    },
    {
      "epoch": 0.6095526914329037,
      "grad_norm": 1.4901211261749268,
      "learning_rate": 1.3954430379746837e-05,
      "loss": 0.08390946686267853,
      "num_input_tokens_seen": 1136336,
      "step": 603,
      "train_runtime": 1660.5878,
      "train_tokens_per_second": 684.297
    },
    {
      "epoch": 0.6105635582512005,
      "grad_norm": 1.8414229154586792,
      "learning_rate": 1.3944303797468355e-05,
      "loss": 0.08612199127674103,
      "num_input_tokens_seen": 1138190,
      "step": 604,
      "train_runtime": 1662.8355,
      "train_tokens_per_second": 684.487
    },
    {
      "epoch": 0.6115744250694971,
      "grad_norm": 2.048098087310791,
      "learning_rate": 1.3934177215189874e-05,
      "loss": 0.1272742748260498,
      "num_input_tokens_seen": 1140056,
      "step": 605,
      "train_runtime": 1665.0136,
      "train_tokens_per_second": 684.713
    },
    {
      "epoch": 0.6125852918877938,
      "grad_norm": 1.5180420875549316,
      "learning_rate": 1.3924050632911395e-05,
      "loss": 0.0807880088686943,
      "num_input_tokens_seen": 1141870,
      "step": 606,
      "train_runtime": 1667.2241,
      "train_tokens_per_second": 684.893
    },
    {
      "epoch": 0.6135961587060905,
      "grad_norm": 1.9689815044403076,
      "learning_rate": 1.3913924050632912e-05,
      "loss": 0.11715269088745117,
      "num_input_tokens_seen": 1143626,
      "step": 607,
      "train_runtime": 1669.4342,
      "train_tokens_per_second": 685.038
    },
    {
      "epoch": 0.6146070255243872,
      "grad_norm": 1.6542195081710815,
      "learning_rate": 1.3903797468354431e-05,
      "loss": 0.1243257075548172,
      "num_input_tokens_seen": 1145798,
      "step": 608,
      "train_runtime": 1671.6778,
      "train_tokens_per_second": 685.418
    },
    {
      "epoch": 0.6156178923426838,
      "grad_norm": 1.7163242101669312,
      "learning_rate": 1.389367088607595e-05,
      "loss": 0.09878333657979965,
      "num_input_tokens_seen": 1147482,
      "step": 609,
      "train_runtime": 1673.8422,
      "train_tokens_per_second": 685.538
    },
    {
      "epoch": 0.6166287591609806,
      "grad_norm": 1.6032116413116455,
      "learning_rate": 1.388354430379747e-05,
      "loss": 0.09439636766910553,
      "num_input_tokens_seen": 1149396,
      "step": 610,
      "train_runtime": 1676.0515,
      "train_tokens_per_second": 685.776
    },
    {
      "epoch": 0.6176396259792772,
      "grad_norm": 1.5943632125854492,
      "learning_rate": 1.3873417721518988e-05,
      "loss": 0.07847783714532852,
      "num_input_tokens_seen": 1150896,
      "step": 611,
      "train_runtime": 1678.1657,
      "train_tokens_per_second": 685.806
    },
    {
      "epoch": 0.6186504927975739,
      "grad_norm": 1.8938099145889282,
      "learning_rate": 1.3863291139240507e-05,
      "loss": 0.1550203561782837,
      "num_input_tokens_seen": 1153316,
      "step": 612,
      "train_runtime": 1680.4594,
      "train_tokens_per_second": 686.31
    },
    {
      "epoch": 0.6196613596158707,
      "grad_norm": 2.2747018337249756,
      "learning_rate": 1.3853164556962027e-05,
      "loss": 0.10700055956840515,
      "num_input_tokens_seen": 1155122,
      "step": 613,
      "train_runtime": 1682.6449,
      "train_tokens_per_second": 686.492
    },
    {
      "epoch": 0.6206722264341673,
      "grad_norm": 1.7282172441482544,
      "learning_rate": 1.3843037974683546e-05,
      "loss": 0.08904734998941422,
      "num_input_tokens_seen": 1156720,
      "step": 614,
      "train_runtime": 1684.7665,
      "train_tokens_per_second": 686.576
    },
    {
      "epoch": 0.621683093252464,
      "grad_norm": 1.474847435951233,
      "learning_rate": 1.3832911392405065e-05,
      "loss": 0.08788709342479706,
      "num_input_tokens_seen": 1158748,
      "step": 615,
      "train_runtime": 1686.9879,
      "train_tokens_per_second": 686.874
    },
    {
      "epoch": 0.6226939600707607,
      "grad_norm": 1.943629503250122,
      "learning_rate": 1.3822784810126584e-05,
      "loss": 0.13403654098510742,
      "num_input_tokens_seen": 1160814,
      "step": 616,
      "train_runtime": 1689.1839,
      "train_tokens_per_second": 687.204
    },
    {
      "epoch": 0.6237048268890574,
      "grad_norm": 1.8858280181884766,
      "learning_rate": 1.3812658227848101e-05,
      "loss": 0.14593029022216797,
      "num_input_tokens_seen": 1162850,
      "step": 617,
      "train_runtime": 1691.4292,
      "train_tokens_per_second": 687.496
    },
    {
      "epoch": 0.624715693707354,
      "grad_norm": 1.5736969709396362,
      "learning_rate": 1.3802531645569622e-05,
      "loss": 0.098531574010849,
      "num_input_tokens_seen": 1165406,
      "step": 618,
      "train_runtime": 1693.7998,
      "train_tokens_per_second": 688.042
    },
    {
      "epoch": 0.6257265605256508,
      "grad_norm": 2.322218894958496,
      "learning_rate": 1.3792405063291141e-05,
      "loss": 0.12939922511577606,
      "num_input_tokens_seen": 1167150,
      "step": 619,
      "train_runtime": 1695.9924,
      "train_tokens_per_second": 688.181
    },
    {
      "epoch": 0.6267374273439474,
      "grad_norm": 1.5616765022277832,
      "learning_rate": 1.3782278481012658e-05,
      "loss": 0.11167999356985092,
      "num_input_tokens_seen": 1169178,
      "step": 620,
      "train_runtime": 1698.2538,
      "train_tokens_per_second": 688.459
    },
    {
      "epoch": 0.6277482941622441,
      "grad_norm": 1.7037088871002197,
      "learning_rate": 1.377215189873418e-05,
      "loss": 0.10070261359214783,
      "num_input_tokens_seen": 1170776,
      "step": 621,
      "train_runtime": 1700.4086,
      "train_tokens_per_second": 688.526
    },
    {
      "epoch": 0.6287591609805409,
      "grad_norm": 1.8399248123168945,
      "learning_rate": 1.3762025316455698e-05,
      "loss": 0.08328185230493546,
      "num_input_tokens_seen": 1172508,
      "step": 622,
      "train_runtime": 1702.5938,
      "train_tokens_per_second": 688.66
    },
    {
      "epoch": 0.6297700277988375,
      "grad_norm": 1.5935235023498535,
      "learning_rate": 1.3751898734177216e-05,
      "loss": 0.10739567875862122,
      "num_input_tokens_seen": 1174694,
      "step": 623,
      "train_runtime": 1704.893,
      "train_tokens_per_second": 689.013
    },
    {
      "epoch": 0.6307808946171342,
      "grad_norm": 1.9569668769836426,
      "learning_rate": 1.3741772151898736e-05,
      "loss": 0.09974449872970581,
      "num_input_tokens_seen": 1176490,
      "step": 624,
      "train_runtime": 1707.1196,
      "train_tokens_per_second": 689.167
    },
    {
      "epoch": 0.6317917614354309,
      "grad_norm": 1.612789511680603,
      "learning_rate": 1.3731645569620254e-05,
      "loss": 0.08718269318342209,
      "num_input_tokens_seen": 1178784,
      "step": 625,
      "train_runtime": 1709.3876,
      "train_tokens_per_second": 689.594
    },
    {
      "epoch": 0.6328026282537276,
      "grad_norm": 1.868043065071106,
      "learning_rate": 1.3721518987341773e-05,
      "loss": 0.10843443125486374,
      "num_input_tokens_seen": 1180080,
      "step": 626,
      "train_runtime": 1711.5052,
      "train_tokens_per_second": 689.498
    },
    {
      "epoch": 0.6338134950720242,
      "grad_norm": 1.8122128248214722,
      "learning_rate": 1.3711392405063294e-05,
      "loss": 0.08479487895965576,
      "num_input_tokens_seen": 1181322,
      "step": 627,
      "train_runtime": 1713.538,
      "train_tokens_per_second": 689.405
    },
    {
      "epoch": 0.634824361890321,
      "grad_norm": 1.8465759754180908,
      "learning_rate": 1.3701265822784811e-05,
      "loss": 0.14208438992500305,
      "num_input_tokens_seen": 1183382,
      "step": 628,
      "train_runtime": 1715.7791,
      "train_tokens_per_second": 689.705
    },
    {
      "epoch": 0.6358352287086176,
      "grad_norm": 1.786992073059082,
      "learning_rate": 1.369113924050633e-05,
      "loss": 0.11269376426935196,
      "num_input_tokens_seen": 1185392,
      "step": 629,
      "train_runtime": 1717.9893,
      "train_tokens_per_second": 689.988
    },
    {
      "epoch": 0.6368460955269143,
      "grad_norm": 1.9114515781402588,
      "learning_rate": 1.3681012658227851e-05,
      "loss": 0.13633716106414795,
      "num_input_tokens_seen": 1187446,
      "step": 630,
      "train_runtime": 1720.2564,
      "train_tokens_per_second": 690.273
    },
    {
      "epoch": 0.637856962345211,
      "grad_norm": 1.664825201034546,
      "learning_rate": 1.3670886075949368e-05,
      "loss": 0.1060963124036789,
      "num_input_tokens_seen": 1189472,
      "step": 631,
      "train_runtime": 1729.7407,
      "train_tokens_per_second": 687.659
    },
    {
      "epoch": 0.6388678291635077,
      "grad_norm": 1.6935018301010132,
      "learning_rate": 1.3660759493670887e-05,
      "loss": 0.11799652874469757,
      "num_input_tokens_seen": 1191462,
      "step": 632,
      "train_runtime": 1731.9926,
      "train_tokens_per_second": 687.914
    },
    {
      "epoch": 0.6398786959818044,
      "grad_norm": 1.6820141077041626,
      "learning_rate": 1.3650632911392405e-05,
      "loss": 0.11976519227027893,
      "num_input_tokens_seen": 1193630,
      "step": 633,
      "train_runtime": 1734.3499,
      "train_tokens_per_second": 688.229
    },
    {
      "epoch": 0.6408895628001011,
      "grad_norm": 1.840476155281067,
      "learning_rate": 1.3640506329113926e-05,
      "loss": 0.10967463999986649,
      "num_input_tokens_seen": 1195248,
      "step": 634,
      "train_runtime": 1736.5188,
      "train_tokens_per_second": 688.301
    },
    {
      "epoch": 0.6419004296183978,
      "grad_norm": 1.674445629119873,
      "learning_rate": 1.3630379746835445e-05,
      "loss": 0.1125248521566391,
      "num_input_tokens_seen": 1197508,
      "step": 635,
      "train_runtime": 1738.8871,
      "train_tokens_per_second": 688.663
    },
    {
      "epoch": 0.6429112964366944,
      "grad_norm": 1.5510694980621338,
      "learning_rate": 1.3620253164556962e-05,
      "loss": 0.10820190608501434,
      "num_input_tokens_seen": 1199538,
      "step": 636,
      "train_runtime": 1741.1192,
      "train_tokens_per_second": 688.947
    },
    {
      "epoch": 0.6439221632549912,
      "grad_norm": 1.8009198904037476,
      "learning_rate": 1.3610126582278483e-05,
      "loss": 0.11245355755090714,
      "num_input_tokens_seen": 1201214,
      "step": 637,
      "train_runtime": 1743.3326,
      "train_tokens_per_second": 689.033
    },
    {
      "epoch": 0.6449330300732878,
      "grad_norm": 1.7082339525222778,
      "learning_rate": 1.3600000000000002e-05,
      "loss": 0.10139073431491852,
      "num_input_tokens_seen": 1202928,
      "step": 638,
      "train_runtime": 1745.5632,
      "train_tokens_per_second": 689.135
    },
    {
      "epoch": 0.6459438968915845,
      "grad_norm": 1.6657874584197998,
      "learning_rate": 1.358987341772152e-05,
      "loss": 0.09136809408664703,
      "num_input_tokens_seen": 1205536,
      "step": 639,
      "train_runtime": 1747.9084,
      "train_tokens_per_second": 689.702
    },
    {
      "epoch": 0.6469547637098813,
      "grad_norm": 1.4045131206512451,
      "learning_rate": 1.357974683544304e-05,
      "loss": 0.10315302014350891,
      "num_input_tokens_seen": 1208210,
      "step": 640,
      "train_runtime": 1750.2476,
      "train_tokens_per_second": 690.308
    },
    {
      "epoch": 0.6479656305281779,
      "grad_norm": 1.7283861637115479,
      "learning_rate": 1.3569620253164557e-05,
      "loss": 0.10815608501434326,
      "num_input_tokens_seen": 1209926,
      "step": 641,
      "train_runtime": 1752.433,
      "train_tokens_per_second": 690.426
    },
    {
      "epoch": 0.6489764973464746,
      "grad_norm": 2.3188953399658203,
      "learning_rate": 1.3559493670886077e-05,
      "loss": 0.1732199639081955,
      "num_input_tokens_seen": 1211520,
      "step": 642,
      "train_runtime": 1754.5886,
      "train_tokens_per_second": 690.487
    },
    {
      "epoch": 0.6499873641647713,
      "grad_norm": 1.921383023262024,
      "learning_rate": 1.3549367088607597e-05,
      "loss": 0.10076595842838287,
      "num_input_tokens_seen": 1213840,
      "step": 643,
      "train_runtime": 1756.9758,
      "train_tokens_per_second": 690.869
    },
    {
      "epoch": 0.650998230983068,
      "grad_norm": 2.255113363265991,
      "learning_rate": 1.3539240506329115e-05,
      "loss": 0.15450282394886017,
      "num_input_tokens_seen": 1215450,
      "step": 644,
      "train_runtime": 1759.1032,
      "train_tokens_per_second": 690.949
    },
    {
      "epoch": 0.6520090978013646,
      "grad_norm": 1.5902783870697021,
      "learning_rate": 1.3529113924050634e-05,
      "loss": 0.11672405898571014,
      "num_input_tokens_seen": 1217814,
      "step": 645,
      "train_runtime": 1761.4052,
      "train_tokens_per_second": 691.388
    },
    {
      "epoch": 0.6530199646196614,
      "grad_norm": 1.8295010328292847,
      "learning_rate": 1.3518987341772155e-05,
      "loss": 0.12819403409957886,
      "num_input_tokens_seen": 1220000,
      "step": 646,
      "train_runtime": 1763.6971,
      "train_tokens_per_second": 691.729
    },
    {
      "epoch": 0.654030831437958,
      "grad_norm": 1.8442647457122803,
      "learning_rate": 1.3508860759493672e-05,
      "loss": 0.10710816830396652,
      "num_input_tokens_seen": 1221386,
      "step": 647,
      "train_runtime": 1765.8024,
      "train_tokens_per_second": 691.689
    },
    {
      "epoch": 0.6550416982562547,
      "grad_norm": 1.8489795923233032,
      "learning_rate": 1.3498734177215191e-05,
      "loss": 0.09648443013429642,
      "num_input_tokens_seen": 1222972,
      "step": 648,
      "train_runtime": 1767.9622,
      "train_tokens_per_second": 691.741
    },
    {
      "epoch": 0.6560525650745515,
      "grad_norm": 1.423836350440979,
      "learning_rate": 1.3488607594936708e-05,
      "loss": 0.08564922958612442,
      "num_input_tokens_seen": 1224776,
      "step": 649,
      "train_runtime": 1770.1922,
      "train_tokens_per_second": 691.889
    },
    {
      "epoch": 0.6570634318928481,
      "grad_norm": 1.779313087463379,
      "learning_rate": 1.347848101265823e-05,
      "loss": 0.11435481905937195,
      "num_input_tokens_seen": 1226784,
      "step": 650,
      "train_runtime": 1772.4521,
      "train_tokens_per_second": 692.139
    },
    {
      "epoch": 0.6580742987111448,
      "grad_norm": 1.8369637727737427,
      "learning_rate": 1.3468354430379748e-05,
      "loss": 0.11493565142154694,
      "num_input_tokens_seen": 1228484,
      "step": 651,
      "train_runtime": 1774.7013,
      "train_tokens_per_second": 692.22
    },
    {
      "epoch": 0.6590851655294415,
      "grad_norm": 1.5541417598724365,
      "learning_rate": 1.3458227848101266e-05,
      "loss": 0.09260427206754684,
      "num_input_tokens_seen": 1230210,
      "step": 652,
      "train_runtime": 1776.8983,
      "train_tokens_per_second": 692.336
    },
    {
      "epoch": 0.6600960323477382,
      "grad_norm": 2.0325305461883545,
      "learning_rate": 1.3448101265822786e-05,
      "loss": 0.15358027815818787,
      "num_input_tokens_seen": 1232078,
      "step": 653,
      "train_runtime": 1779.1357,
      "train_tokens_per_second": 692.515
    },
    {
      "epoch": 0.6611068991660348,
      "grad_norm": 1.5942155122756958,
      "learning_rate": 1.3437974683544306e-05,
      "loss": 0.104779914021492,
      "num_input_tokens_seen": 1233970,
      "step": 654,
      "train_runtime": 1781.3051,
      "train_tokens_per_second": 692.734
    },
    {
      "epoch": 0.6621177659843316,
      "grad_norm": 1.6773817539215088,
      "learning_rate": 1.3427848101265823e-05,
      "loss": 0.1102675199508667,
      "num_input_tokens_seen": 1235770,
      "step": 655,
      "train_runtime": 1783.5556,
      "train_tokens_per_second": 692.869
    },
    {
      "epoch": 0.6631286328026282,
      "grad_norm": 1.7004425525665283,
      "learning_rate": 1.3417721518987344e-05,
      "loss": 0.09976387023925781,
      "num_input_tokens_seen": 1237816,
      "step": 656,
      "train_runtime": 1785.807,
      "train_tokens_per_second": 693.141
    },
    {
      "epoch": 0.6641394996209249,
      "grad_norm": 1.7719749212265015,
      "learning_rate": 1.3407594936708861e-05,
      "loss": 0.09984132647514343,
      "num_input_tokens_seen": 1240466,
      "step": 657,
      "train_runtime": 1788.2125,
      "train_tokens_per_second": 693.69
    },
    {
      "epoch": 0.6651503664392217,
      "grad_norm": 1.3475958108901978,
      "learning_rate": 1.339746835443038e-05,
      "loss": 0.09303415566682816,
      "num_input_tokens_seen": 1242766,
      "step": 658,
      "train_runtime": 1790.5161,
      "train_tokens_per_second": 694.083
    },
    {
      "epoch": 0.6661612332575183,
      "grad_norm": 1.8140196800231934,
      "learning_rate": 1.3387341772151901e-05,
      "loss": 0.0961821973323822,
      "num_input_tokens_seen": 1244288,
      "step": 659,
      "train_runtime": 1792.6296,
      "train_tokens_per_second": 694.113
    },
    {
      "epoch": 0.6671721000758151,
      "grad_norm": 1.5088071823120117,
      "learning_rate": 1.3377215189873418e-05,
      "loss": 0.07277616858482361,
      "num_input_tokens_seen": 1245836,
      "step": 660,
      "train_runtime": 1794.7355,
      "train_tokens_per_second": 694.161
    },
    {
      "epoch": 0.6681829668941117,
      "grad_norm": 1.5696074962615967,
      "learning_rate": 1.3367088607594937e-05,
      "loss": 0.08693429082632065,
      "num_input_tokens_seen": 1247956,
      "step": 661,
      "train_runtime": 1804.2071,
      "train_tokens_per_second": 691.692
    },
    {
      "epoch": 0.6691938337124084,
      "grad_norm": 1.8871132135391235,
      "learning_rate": 1.3356962025316458e-05,
      "loss": 0.10091319680213928,
      "num_input_tokens_seen": 1249858,
      "step": 662,
      "train_runtime": 1806.4483,
      "train_tokens_per_second": 691.887
    },
    {
      "epoch": 0.6702047005307051,
      "grad_norm": 1.6133759021759033,
      "learning_rate": 1.3346835443037976e-05,
      "loss": 0.11935020238161087,
      "num_input_tokens_seen": 1252000,
      "step": 663,
      "train_runtime": 1808.7267,
      "train_tokens_per_second": 692.2
    },
    {
      "epoch": 0.6712155673490018,
      "grad_norm": 1.6783758401870728,
      "learning_rate": 1.3336708860759495e-05,
      "loss": 0.07670894265174866,
      "num_input_tokens_seen": 1253574,
      "step": 664,
      "train_runtime": 1810.8889,
      "train_tokens_per_second": 692.242
    },
    {
      "epoch": 0.6722264341672984,
      "grad_norm": 1.9245914220809937,
      "learning_rate": 1.3326582278481012e-05,
      "loss": 0.08715176582336426,
      "num_input_tokens_seen": 1255260,
      "step": 665,
      "train_runtime": 1813.0689,
      "train_tokens_per_second": 692.34
    },
    {
      "epoch": 0.6732373009855952,
      "grad_norm": 1.4565939903259277,
      "learning_rate": 1.3316455696202533e-05,
      "loss": 0.11130739748477936,
      "num_input_tokens_seen": 1257338,
      "step": 666,
      "train_runtime": 1815.3233,
      "train_tokens_per_second": 692.625
    },
    {
      "epoch": 0.6742481678038919,
      "grad_norm": 2.1480889320373535,
      "learning_rate": 1.3306329113924052e-05,
      "loss": 0.11984491348266602,
      "num_input_tokens_seen": 1259314,
      "step": 667,
      "train_runtime": 1817.5696,
      "train_tokens_per_second": 692.856
    },
    {
      "epoch": 0.6752590346221885,
      "grad_norm": 1.7935190200805664,
      "learning_rate": 1.329620253164557e-05,
      "loss": 0.09027568250894547,
      "num_input_tokens_seen": 1261426,
      "step": 668,
      "train_runtime": 1819.7888,
      "train_tokens_per_second": 693.172
    },
    {
      "epoch": 0.6762699014404853,
      "grad_norm": 1.55329167842865,
      "learning_rate": 1.328607594936709e-05,
      "loss": 0.09494169056415558,
      "num_input_tokens_seen": 1263376,
      "step": 669,
      "train_runtime": 1822.0588,
      "train_tokens_per_second": 693.378
    },
    {
      "epoch": 0.6772807682587819,
      "grad_norm": 1.898910641670227,
      "learning_rate": 1.327594936708861e-05,
      "loss": 0.10689137876033783,
      "num_input_tokens_seen": 1265248,
      "step": 670,
      "train_runtime": 1824.2729,
      "train_tokens_per_second": 693.563
    },
    {
      "epoch": 0.6782916350770786,
      "grad_norm": 2.039963483810425,
      "learning_rate": 1.3265822784810127e-05,
      "loss": 0.08548976480960846,
      "num_input_tokens_seen": 1267244,
      "step": 671,
      "train_runtime": 1826.4885,
      "train_tokens_per_second": 693.814
    },
    {
      "epoch": 0.6793025018953753,
      "grad_norm": 1.8944624662399292,
      "learning_rate": 1.3255696202531647e-05,
      "loss": 0.11550843715667725,
      "num_input_tokens_seen": 1269154,
      "step": 672,
      "train_runtime": 1828.6718,
      "train_tokens_per_second": 694.031
    },
    {
      "epoch": 0.680313368713672,
      "grad_norm": 1.735344648361206,
      "learning_rate": 1.3245569620253165e-05,
      "loss": 0.06827060878276825,
      "num_input_tokens_seen": 1270498,
      "step": 673,
      "train_runtime": 1830.7549,
      "train_tokens_per_second": 693.975
    },
    {
      "epoch": 0.6813242355319686,
      "grad_norm": 2.0431125164031982,
      "learning_rate": 1.3235443037974684e-05,
      "loss": 0.1385970413684845,
      "num_input_tokens_seen": 1272808,
      "step": 674,
      "train_runtime": 1833.0853,
      "train_tokens_per_second": 694.353
    },
    {
      "epoch": 0.6823351023502654,
      "grad_norm": 1.8650518655776978,
      "learning_rate": 1.3225316455696205e-05,
      "loss": 0.09701419621706009,
      "num_input_tokens_seen": 1274488,
      "step": 675,
      "train_runtime": 1835.2786,
      "train_tokens_per_second": 694.438
    },
    {
      "epoch": 0.683345969168562,
      "grad_norm": 2.133262872695923,
      "learning_rate": 1.3215189873417722e-05,
      "loss": 0.10070274770259857,
      "num_input_tokens_seen": 1275734,
      "step": 676,
      "train_runtime": 1837.312,
      "train_tokens_per_second": 694.348
    },
    {
      "epoch": 0.6843568359868587,
      "grad_norm": 1.8701945543289185,
      "learning_rate": 1.3205063291139241e-05,
      "loss": 0.10083901882171631,
      "num_input_tokens_seen": 1278096,
      "step": 677,
      "train_runtime": 1839.63,
      "train_tokens_per_second": 694.757
    },
    {
      "epoch": 0.6853677028051555,
      "grad_norm": 2.1649763584136963,
      "learning_rate": 1.3194936708860762e-05,
      "loss": 0.10348096489906311,
      "num_input_tokens_seen": 1279612,
      "step": 678,
      "train_runtime": 1841.7237,
      "train_tokens_per_second": 694.79
    },
    {
      "epoch": 0.6863785696234521,
      "grad_norm": 2.2240641117095947,
      "learning_rate": 1.318481012658228e-05,
      "loss": 0.14250193536281586,
      "num_input_tokens_seen": 1281712,
      "step": 679,
      "train_runtime": 1843.9929,
      "train_tokens_per_second": 695.074
    },
    {
      "epoch": 0.6873894364417488,
      "grad_norm": 1.875397801399231,
      "learning_rate": 1.3174683544303798e-05,
      "loss": 0.12258376181125641,
      "num_input_tokens_seen": 1284012,
      "step": 680,
      "train_runtime": 1846.2983,
      "train_tokens_per_second": 695.452
    },
    {
      "epoch": 0.6884003032600455,
      "grad_norm": 1.5993077754974365,
      "learning_rate": 1.3164556962025317e-05,
      "loss": 0.07303463667631149,
      "num_input_tokens_seen": 1285832,
      "step": 681,
      "train_runtime": 1848.5111,
      "train_tokens_per_second": 695.604
    },
    {
      "epoch": 0.6894111700783422,
      "grad_norm": 1.8077025413513184,
      "learning_rate": 1.3154430379746837e-05,
      "loss": 0.12092280387878418,
      "num_input_tokens_seen": 1287984,
      "step": 682,
      "train_runtime": 1850.78,
      "train_tokens_per_second": 695.914
    },
    {
      "epoch": 0.6904220368966388,
      "grad_norm": 2.380758285522461,
      "learning_rate": 1.3144303797468356e-05,
      "loss": 0.15384668111801147,
      "num_input_tokens_seen": 1289848,
      "step": 683,
      "train_runtime": 1852.9968,
      "train_tokens_per_second": 696.088
    },
    {
      "epoch": 0.6914329037149356,
      "grad_norm": 2.1788816452026367,
      "learning_rate": 1.3134177215189875e-05,
      "loss": 0.12366777658462524,
      "num_input_tokens_seen": 1291486,
      "step": 684,
      "train_runtime": 1855.2049,
      "train_tokens_per_second": 696.142
    },
    {
      "epoch": 0.6924437705332323,
      "grad_norm": 1.907159447669983,
      "learning_rate": 1.3124050632911394e-05,
      "loss": 0.12734639644622803,
      "num_input_tokens_seen": 1293226,
      "step": 685,
      "train_runtime": 1857.4282,
      "train_tokens_per_second": 696.245
    },
    {
      "epoch": 0.6934546373515289,
      "grad_norm": 1.6104480028152466,
      "learning_rate": 1.3113924050632913e-05,
      "loss": 0.0874997153878212,
      "num_input_tokens_seen": 1294980,
      "step": 686,
      "train_runtime": 1859.6324,
      "train_tokens_per_second": 696.363
    },
    {
      "epoch": 0.6944655041698257,
      "grad_norm": 1.7767722606658936,
      "learning_rate": 1.310379746835443e-05,
      "loss": 0.08478476107120514,
      "num_input_tokens_seen": 1296604,
      "step": 687,
      "train_runtime": 1861.77,
      "train_tokens_per_second": 696.436
    },
    {
      "epoch": 0.6954763709881223,
      "grad_norm": 1.8747044801712036,
      "learning_rate": 1.3093670886075951e-05,
      "loss": 0.09321515262126923,
      "num_input_tokens_seen": 1298266,
      "step": 688,
      "train_runtime": 1863.8942,
      "train_tokens_per_second": 696.534
    },
    {
      "epoch": 0.696487237806419,
      "grad_norm": 1.5795440673828125,
      "learning_rate": 1.3083544303797468e-05,
      "loss": 0.0884488895535469,
      "num_input_tokens_seen": 1300102,
      "step": 689,
      "train_runtime": 1866.1114,
      "train_tokens_per_second": 696.69
    },
    {
      "epoch": 0.6974981046247157,
      "grad_norm": 1.7722744941711426,
      "learning_rate": 1.3073417721518988e-05,
      "loss": 0.11583846807479858,
      "num_input_tokens_seen": 1301776,
      "step": 690,
      "train_runtime": 1868.2577,
      "train_tokens_per_second": 696.786
    },
    {
      "epoch": 0.6985089714430124,
      "grad_norm": 2.364607810974121,
      "learning_rate": 1.3063291139240508e-05,
      "loss": 0.08460243046283722,
      "num_input_tokens_seen": 1303442,
      "step": 691,
      "train_runtime": 1877.5728,
      "train_tokens_per_second": 694.216
    },
    {
      "epoch": 0.699519838261309,
      "grad_norm": 2.0333120822906494,
      "learning_rate": 1.3053164556962026e-05,
      "loss": 0.12729576230049133,
      "num_input_tokens_seen": 1305156,
      "step": 692,
      "train_runtime": 1879.7593,
      "train_tokens_per_second": 694.321
    },
    {
      "epoch": 0.7005307050796058,
      "grad_norm": 1.848425030708313,
      "learning_rate": 1.3043037974683545e-05,
      "loss": 0.1383374184370041,
      "num_input_tokens_seen": 1306902,
      "step": 693,
      "train_runtime": 1881.9951,
      "train_tokens_per_second": 694.424
    },
    {
      "epoch": 0.7015415718979024,
      "grad_norm": 1.8356151580810547,
      "learning_rate": 1.3032911392405066e-05,
      "loss": 0.10371893644332886,
      "num_input_tokens_seen": 1308664,
      "step": 694,
      "train_runtime": 1884.1682,
      "train_tokens_per_second": 694.558
    },
    {
      "epoch": 0.7025524387161991,
      "grad_norm": 1.4701520204544067,
      "learning_rate": 1.3022784810126583e-05,
      "loss": 0.10076653957366943,
      "num_input_tokens_seen": 1310814,
      "step": 695,
      "train_runtime": 1886.4556,
      "train_tokens_per_second": 694.855
    },
    {
      "epoch": 0.7035633055344959,
      "grad_norm": 2.1058008670806885,
      "learning_rate": 1.3012658227848102e-05,
      "loss": 0.11140087246894836,
      "num_input_tokens_seen": 1312786,
      "step": 696,
      "train_runtime": 1888.7239,
      "train_tokens_per_second": 695.065
    },
    {
      "epoch": 0.7045741723527925,
      "grad_norm": 1.3815808296203613,
      "learning_rate": 1.3002531645569621e-05,
      "loss": 0.07784861326217651,
      "num_input_tokens_seen": 1315162,
      "step": 697,
      "train_runtime": 1891.0169,
      "train_tokens_per_second": 695.479
    },
    {
      "epoch": 0.7055850391710892,
      "grad_norm": 1.7595512866973877,
      "learning_rate": 1.299240506329114e-05,
      "loss": 0.1133749857544899,
      "num_input_tokens_seen": 1316704,
      "step": 698,
      "train_runtime": 1893.1302,
      "train_tokens_per_second": 695.517
    },
    {
      "epoch": 0.7065959059893859,
      "grad_norm": 2.2063496112823486,
      "learning_rate": 1.298227848101266e-05,
      "loss": 0.10239047557115555,
      "num_input_tokens_seen": 1318362,
      "step": 699,
      "train_runtime": 1895.2615,
      "train_tokens_per_second": 695.61
    },
    {
      "epoch": 0.7076067728076826,
      "grad_norm": 2.041888952255249,
      "learning_rate": 1.2972151898734178e-05,
      "loss": 0.12548968195915222,
      "num_input_tokens_seen": 1319922,
      "step": 700,
      "train_runtime": 1897.3866,
      "train_tokens_per_second": 695.653
    },
    {
      "epoch": 0.7086176396259792,
      "grad_norm": 1.497320294380188,
      "learning_rate": 1.2962025316455697e-05,
      "loss": 0.0973314642906189,
      "num_input_tokens_seen": 1322010,
      "step": 701,
      "train_runtime": 1899.6352,
      "train_tokens_per_second": 695.928
    },
    {
      "epoch": 0.709628506444276,
      "grad_norm": 1.729615330696106,
      "learning_rate": 1.2951898734177217e-05,
      "loss": 0.12024053931236267,
      "num_input_tokens_seen": 1324172,
      "step": 702,
      "train_runtime": 1901.9102,
      "train_tokens_per_second": 696.233
    },
    {
      "epoch": 0.7106393732625726,
      "grad_norm": 1.5231678485870361,
      "learning_rate": 1.2941772151898736e-05,
      "loss": 0.07961850613355637,
      "num_input_tokens_seen": 1325838,
      "step": 703,
      "train_runtime": 1904.0819,
      "train_tokens_per_second": 696.314
    },
    {
      "epoch": 0.7116502400808693,
      "grad_norm": 2.207820415496826,
      "learning_rate": 1.2931645569620255e-05,
      "loss": 0.11461731791496277,
      "num_input_tokens_seen": 1327436,
      "step": 704,
      "train_runtime": 1906.1964,
      "train_tokens_per_second": 696.379
    },
    {
      "epoch": 0.7126611068991661,
      "grad_norm": 1.5595227479934692,
      "learning_rate": 1.2921518987341772e-05,
      "loss": 0.08717923611402512,
      "num_input_tokens_seen": 1329294,
      "step": 705,
      "train_runtime": 1908.4102,
      "train_tokens_per_second": 696.545
    },
    {
      "epoch": 0.7136719737174627,
      "grad_norm": 1.8853257894515991,
      "learning_rate": 1.2911392405063293e-05,
      "loss": 0.09264266490936279,
      "num_input_tokens_seen": 1331232,
      "step": 706,
      "train_runtime": 1910.6381,
      "train_tokens_per_second": 696.747
    },
    {
      "epoch": 0.7146828405357594,
      "grad_norm": 1.9685697555541992,
      "learning_rate": 1.2901265822784812e-05,
      "loss": 0.13390126824378967,
      "num_input_tokens_seen": 1333056,
      "step": 707,
      "train_runtime": 1912.8921,
      "train_tokens_per_second": 696.88
    },
    {
      "epoch": 0.7156937073540561,
      "grad_norm": 1.809499979019165,
      "learning_rate": 1.289113924050633e-05,
      "loss": 0.12011405825614929,
      "num_input_tokens_seen": 1334582,
      "step": 708,
      "train_runtime": 1915.0501,
      "train_tokens_per_second": 696.891
    },
    {
      "epoch": 0.7167045741723528,
      "grad_norm": 1.5666016340255737,
      "learning_rate": 1.288101265822785e-05,
      "loss": 0.08794930577278137,
      "num_input_tokens_seen": 1336404,
      "step": 709,
      "train_runtime": 1917.3004,
      "train_tokens_per_second": 697.024
    },
    {
      "epoch": 0.7177154409906494,
      "grad_norm": 1.8679513931274414,
      "learning_rate": 1.287088607594937e-05,
      "loss": 0.1294182687997818,
      "num_input_tokens_seen": 1338188,
      "step": 710,
      "train_runtime": 1919.4672,
      "train_tokens_per_second": 697.166
    },
    {
      "epoch": 0.7187263078089462,
      "grad_norm": 1.5189850330352783,
      "learning_rate": 1.2860759493670887e-05,
      "loss": 0.09942683577537537,
      "num_input_tokens_seen": 1340470,
      "step": 711,
      "train_runtime": 1921.7736,
      "train_tokens_per_second": 697.517
    },
    {
      "epoch": 0.7197371746272428,
      "grad_norm": 2.1296935081481934,
      "learning_rate": 1.2850632911392407e-05,
      "loss": 0.11390013992786407,
      "num_input_tokens_seen": 1342374,
      "step": 712,
      "train_runtime": 1923.9525,
      "train_tokens_per_second": 697.717
    },
    {
      "epoch": 0.7207480414455395,
      "grad_norm": 1.7226743698120117,
      "learning_rate": 1.2840506329113925e-05,
      "loss": 0.08570823073387146,
      "num_input_tokens_seen": 1344294,
      "step": 713,
      "train_runtime": 1926.1609,
      "train_tokens_per_second": 697.914
    },
    {
      "epoch": 0.7217589082638363,
      "grad_norm": 1.4276052713394165,
      "learning_rate": 1.2830379746835444e-05,
      "loss": 0.08571764081716537,
      "num_input_tokens_seen": 1346268,
      "step": 714,
      "train_runtime": 1928.3676,
      "train_tokens_per_second": 698.139
    },
    {
      "epoch": 0.7227697750821329,
      "grad_norm": 1.7173999547958374,
      "learning_rate": 1.2820253164556965e-05,
      "loss": 0.09007744491100311,
      "num_input_tokens_seen": 1347826,
      "step": 715,
      "train_runtime": 1930.5239,
      "train_tokens_per_second": 698.166
    },
    {
      "epoch": 0.7237806419004296,
      "grad_norm": 1.850311279296875,
      "learning_rate": 1.2810126582278482e-05,
      "loss": 0.10102897882461548,
      "num_input_tokens_seen": 1349612,
      "step": 716,
      "train_runtime": 1932.706,
      "train_tokens_per_second": 698.302
    },
    {
      "epoch": 0.7247915087187263,
      "grad_norm": 1.758142113685608,
      "learning_rate": 1.2800000000000001e-05,
      "loss": 0.10980450361967087,
      "num_input_tokens_seen": 1351630,
      "step": 717,
      "train_runtime": 1934.9169,
      "train_tokens_per_second": 698.547
    },
    {
      "epoch": 0.725802375537023,
      "grad_norm": 1.403580904006958,
      "learning_rate": 1.2789873417721522e-05,
      "loss": 0.09110099077224731,
      "num_input_tokens_seen": 1353802,
      "step": 718,
      "train_runtime": 1937.221,
      "train_tokens_per_second": 698.837
    },
    {
      "epoch": 0.7268132423553196,
      "grad_norm": 1.5719237327575684,
      "learning_rate": 1.277974683544304e-05,
      "loss": 0.08539742976427078,
      "num_input_tokens_seen": 1355654,
      "step": 719,
      "train_runtime": 1939.4297,
      "train_tokens_per_second": 698.996
    },
    {
      "epoch": 0.7278241091736164,
      "grad_norm": 2.1756863594055176,
      "learning_rate": 1.2769620253164558e-05,
      "loss": 0.15095317363739014,
      "num_input_tokens_seen": 1357500,
      "step": 720,
      "train_runtime": 1941.654,
      "train_tokens_per_second": 699.146
    },
    {
      "epoch": 0.728834975991913,
      "grad_norm": 1.7486422061920166,
      "learning_rate": 1.2759493670886076e-05,
      "loss": 0.10526642203330994,
      "num_input_tokens_seen": 1359508,
      "step": 721,
      "train_runtime": 1951.0673,
      "train_tokens_per_second": 696.802
    },
    {
      "epoch": 0.7298458428102097,
      "grad_norm": 1.548586130142212,
      "learning_rate": 1.2749367088607597e-05,
      "loss": 0.09513578563928604,
      "num_input_tokens_seen": 1361332,
      "step": 722,
      "train_runtime": 1953.3023,
      "train_tokens_per_second": 696.939
    },
    {
      "epoch": 0.7308567096285065,
      "grad_norm": 1.6462899446487427,
      "learning_rate": 1.2739240506329116e-05,
      "loss": 0.10609134286642075,
      "num_input_tokens_seen": 1363258,
      "step": 723,
      "train_runtime": 1955.588,
      "train_tokens_per_second": 697.109
    },
    {
      "epoch": 0.7318675764468031,
      "grad_norm": 1.4377866983413696,
      "learning_rate": 1.2729113924050633e-05,
      "loss": 0.07567666471004486,
      "num_input_tokens_seen": 1365702,
      "step": 724,
      "train_runtime": 1957.9287,
      "train_tokens_per_second": 697.524
    },
    {
      "epoch": 0.7328784432650999,
      "grad_norm": 1.8099792003631592,
      "learning_rate": 1.2718987341772154e-05,
      "loss": 0.08297598361968994,
      "num_input_tokens_seen": 1367666,
      "step": 725,
      "train_runtime": 1960.1522,
      "train_tokens_per_second": 697.735
    },
    {
      "epoch": 0.7338893100833965,
      "grad_norm": 1.7828843593597412,
      "learning_rate": 1.2708860759493671e-05,
      "loss": 0.09248098731040955,
      "num_input_tokens_seen": 1369268,
      "step": 726,
      "train_runtime": 1962.2784,
      "train_tokens_per_second": 697.795
    },
    {
      "epoch": 0.7349001769016932,
      "grad_norm": 2.007511615753174,
      "learning_rate": 1.269873417721519e-05,
      "loss": 0.09541449695825577,
      "num_input_tokens_seen": 1370694,
      "step": 727,
      "train_runtime": 1964.3707,
      "train_tokens_per_second": 697.778
    },
    {
      "epoch": 0.7359110437199899,
      "grad_norm": 1.7251880168914795,
      "learning_rate": 1.2688607594936711e-05,
      "loss": 0.11526394635438919,
      "num_input_tokens_seen": 1372434,
      "step": 728,
      "train_runtime": 1966.5293,
      "train_tokens_per_second": 697.897
    },
    {
      "epoch": 0.7369219105382866,
      "grad_norm": 1.4944250583648682,
      "learning_rate": 1.2678481012658228e-05,
      "loss": 0.0895615816116333,
      "num_input_tokens_seen": 1374256,
      "step": 729,
      "train_runtime": 1968.7071,
      "train_tokens_per_second": 698.05
    },
    {
      "epoch": 0.7379327773565832,
      "grad_norm": 1.6431430578231812,
      "learning_rate": 1.2668354430379748e-05,
      "loss": 0.09937413781881332,
      "num_input_tokens_seen": 1376118,
      "step": 730,
      "train_runtime": 1970.9224,
      "train_tokens_per_second": 698.21
    },
    {
      "epoch": 0.73894364417488,
      "grad_norm": 1.801094889640808,
      "learning_rate": 1.2658227848101268e-05,
      "loss": 0.09845241904258728,
      "num_input_tokens_seen": 1377884,
      "step": 731,
      "train_runtime": 1973.077,
      "train_tokens_per_second": 698.343
    },
    {
      "epoch": 0.7399545109931767,
      "grad_norm": 1.6009243726730347,
      "learning_rate": 1.2648101265822786e-05,
      "loss": 0.09818520396947861,
      "num_input_tokens_seen": 1379496,
      "step": 732,
      "train_runtime": 1975.2218,
      "train_tokens_per_second": 698.401
    },
    {
      "epoch": 0.7409653778114733,
      "grad_norm": 1.6257437467575073,
      "learning_rate": 1.2637974683544305e-05,
      "loss": 0.08480114489793777,
      "num_input_tokens_seen": 1381244,
      "step": 733,
      "train_runtime": 1977.4021,
      "train_tokens_per_second": 698.514
    },
    {
      "epoch": 0.7419762446297701,
      "grad_norm": 1.7684394121170044,
      "learning_rate": 1.2627848101265822e-05,
      "loss": 0.10317234694957733,
      "num_input_tokens_seen": 1383306,
      "step": 734,
      "train_runtime": 1979.6733,
      "train_tokens_per_second": 698.755
    },
    {
      "epoch": 0.7429871114480667,
      "grad_norm": 1.490909457206726,
      "learning_rate": 1.2617721518987343e-05,
      "loss": 0.0988856852054596,
      "num_input_tokens_seen": 1385492,
      "step": 735,
      "train_runtime": 1981.9417,
      "train_tokens_per_second": 699.058
    },
    {
      "epoch": 0.7439979782663634,
      "grad_norm": 2.0628840923309326,
      "learning_rate": 1.2607594936708862e-05,
      "loss": 0.12445077300071716,
      "num_input_tokens_seen": 1387608,
      "step": 736,
      "train_runtime": 1984.1985,
      "train_tokens_per_second": 699.329
    },
    {
      "epoch": 0.7450088450846601,
      "grad_norm": 1.647877812385559,
      "learning_rate": 1.259746835443038e-05,
      "loss": 0.11410157382488251,
      "num_input_tokens_seen": 1389814,
      "step": 737,
      "train_runtime": 1986.5101,
      "train_tokens_per_second": 699.626
    },
    {
      "epoch": 0.7460197119029568,
      "grad_norm": 1.7574702501296997,
      "learning_rate": 1.25873417721519e-05,
      "loss": 0.12652677297592163,
      "num_input_tokens_seen": 1391662,
      "step": 738,
      "train_runtime": 1988.7121,
      "train_tokens_per_second": 699.781
    },
    {
      "epoch": 0.7470305787212534,
      "grad_norm": 1.6824933290481567,
      "learning_rate": 1.257721518987342e-05,
      "loss": 0.10845467448234558,
      "num_input_tokens_seen": 1393230,
      "step": 739,
      "train_runtime": 1990.8369,
      "train_tokens_per_second": 699.821
    },
    {
      "epoch": 0.7480414455395502,
      "grad_norm": 1.4367671012878418,
      "learning_rate": 1.2567088607594937e-05,
      "loss": 0.07859230786561966,
      "num_input_tokens_seen": 1395328,
      "step": 740,
      "train_runtime": 1993.0943,
      "train_tokens_per_second": 700.081
    },
    {
      "epoch": 0.7490523123578469,
      "grad_norm": 1.9502897262573242,
      "learning_rate": 1.2556962025316457e-05,
      "loss": 0.13023243844509125,
      "num_input_tokens_seen": 1397110,
      "step": 741,
      "train_runtime": 1995.2831,
      "train_tokens_per_second": 700.206
    },
    {
      "epoch": 0.7500631791761435,
      "grad_norm": 2.0860912799835205,
      "learning_rate": 1.2546835443037975e-05,
      "loss": 0.11763650923967361,
      "num_input_tokens_seen": 1399078,
      "step": 742,
      "train_runtime": 1997.5185,
      "train_tokens_per_second": 700.408
    },
    {
      "epoch": 0.7510740459944403,
      "grad_norm": 1.8446928262710571,
      "learning_rate": 1.2536708860759494e-05,
      "loss": 0.08820986747741699,
      "num_input_tokens_seen": 1400736,
      "step": 743,
      "train_runtime": 1999.6882,
      "train_tokens_per_second": 700.477
    },
    {
      "epoch": 0.7520849128127369,
      "grad_norm": 1.6179382801055908,
      "learning_rate": 1.2526582278481015e-05,
      "loss": 0.10503236204385757,
      "num_input_tokens_seen": 1403160,
      "step": 744,
      "train_runtime": 2002.0174,
      "train_tokens_per_second": 700.873
    },
    {
      "epoch": 0.7530957796310336,
      "grad_norm": 1.6173311471939087,
      "learning_rate": 1.2516455696202532e-05,
      "loss": 0.09356211125850677,
      "num_input_tokens_seen": 1405066,
      "step": 745,
      "train_runtime": 2004.19,
      "train_tokens_per_second": 701.064
    },
    {
      "epoch": 0.7541066464493303,
      "grad_norm": 1.9140892028808594,
      "learning_rate": 1.2506329113924051e-05,
      "loss": 0.1226038932800293,
      "num_input_tokens_seen": 1406536,
      "step": 746,
      "train_runtime": 2006.2952,
      "train_tokens_per_second": 701.061
    },
    {
      "epoch": 0.755117513267627,
      "grad_norm": 1.3898205757141113,
      "learning_rate": 1.2496202531645572e-05,
      "loss": 0.07656963169574738,
      "num_input_tokens_seen": 1408360,
      "step": 747,
      "train_runtime": 2008.4951,
      "train_tokens_per_second": 701.202
    },
    {
      "epoch": 0.7561283800859236,
      "grad_norm": 1.4517537355422974,
      "learning_rate": 1.248607594936709e-05,
      "loss": 0.08344630151987076,
      "num_input_tokens_seen": 1410444,
      "step": 748,
      "train_runtime": 2010.7305,
      "train_tokens_per_second": 701.459
    },
    {
      "epoch": 0.7571392469042204,
      "grad_norm": 1.534968376159668,
      "learning_rate": 1.2475949367088608e-05,
      "loss": 0.08581095188856125,
      "num_input_tokens_seen": 1412882,
      "step": 749,
      "train_runtime": 2013.0799,
      "train_tokens_per_second": 701.851
    },
    {
      "epoch": 0.7581501137225171,
      "grad_norm": 1.670307993888855,
      "learning_rate": 1.2465822784810126e-05,
      "loss": 0.10169275104999542,
      "num_input_tokens_seen": 1415136,
      "step": 750,
      "train_runtime": 2015.3603,
      "train_tokens_per_second": 702.175
    },
    {
      "epoch": 0.7591609805408137,
      "grad_norm": 2.2500455379486084,
      "learning_rate": 1.2455696202531647e-05,
      "loss": 0.09628786146640778,
      "num_input_tokens_seen": 1416954,
      "step": 751,
      "train_runtime": 2024.8329,
      "train_tokens_per_second": 699.788
    },
    {
      "epoch": 0.7601718473591105,
      "grad_norm": 1.824182391166687,
      "learning_rate": 1.2445569620253166e-05,
      "loss": 0.09869734197854996,
      "num_input_tokens_seen": 1419028,
      "step": 752,
      "train_runtime": 2027.1807,
      "train_tokens_per_second": 700.001
    },
    {
      "epoch": 0.7611827141774071,
      "grad_norm": 1.7404752969741821,
      "learning_rate": 1.2435443037974683e-05,
      "loss": 0.13448429107666016,
      "num_input_tokens_seen": 1421050,
      "step": 753,
      "train_runtime": 2029.4385,
      "train_tokens_per_second": 700.218
    },
    {
      "epoch": 0.7621935809957038,
      "grad_norm": 1.6820043325424194,
      "learning_rate": 1.2425316455696204e-05,
      "loss": 0.08861427009105682,
      "num_input_tokens_seen": 1422814,
      "step": 754,
      "train_runtime": 2031.6095,
      "train_tokens_per_second": 700.338
    },
    {
      "epoch": 0.7632044478140005,
      "grad_norm": 1.8669121265411377,
      "learning_rate": 1.2415189873417723e-05,
      "loss": 0.10285475850105286,
      "num_input_tokens_seen": 1424398,
      "step": 755,
      "train_runtime": 2033.7218,
      "train_tokens_per_second": 700.39
    },
    {
      "epoch": 0.7642153146322972,
      "grad_norm": 1.5383599996566772,
      "learning_rate": 1.240506329113924e-05,
      "loss": 0.09902352839708328,
      "num_input_tokens_seen": 1426688,
      "step": 756,
      "train_runtime": 2036.033,
      "train_tokens_per_second": 700.719
    },
    {
      "epoch": 0.7652261814505938,
      "grad_norm": 1.8109301328659058,
      "learning_rate": 1.2394936708860761e-05,
      "loss": 0.1470569372177124,
      "num_input_tokens_seen": 1428828,
      "step": 757,
      "train_runtime": 2038.3818,
      "train_tokens_per_second": 700.962
    },
    {
      "epoch": 0.7662370482688906,
      "grad_norm": 1.5089824199676514,
      "learning_rate": 1.2384810126582278e-05,
      "loss": 0.0822841078042984,
      "num_input_tokens_seen": 1430676,
      "step": 758,
      "train_runtime": 2040.5503,
      "train_tokens_per_second": 701.123
    },
    {
      "epoch": 0.7672479150871873,
      "grad_norm": 1.8709169626235962,
      "learning_rate": 1.2374683544303798e-05,
      "loss": 0.1347963511943817,
      "num_input_tokens_seen": 1432730,
      "step": 759,
      "train_runtime": 2042.7856,
      "train_tokens_per_second": 701.361
    },
    {
      "epoch": 0.7682587819054839,
      "grad_norm": 1.6669872999191284,
      "learning_rate": 1.2364556962025318e-05,
      "loss": 0.10360300540924072,
      "num_input_tokens_seen": 1434948,
      "step": 760,
      "train_runtime": 2045.0728,
      "train_tokens_per_second": 701.661
    },
    {
      "epoch": 0.7692696487237807,
      "grad_norm": 1.7485250234603882,
      "learning_rate": 1.2354430379746836e-05,
      "loss": 0.10558325052261353,
      "num_input_tokens_seen": 1436938,
      "step": 761,
      "train_runtime": 2047.2884,
      "train_tokens_per_second": 701.874
    },
    {
      "epoch": 0.7702805155420773,
      "grad_norm": 1.4305222034454346,
      "learning_rate": 1.2344303797468355e-05,
      "loss": 0.07555848360061646,
      "num_input_tokens_seen": 1439072,
      "step": 762,
      "train_runtime": 2049.5636,
      "train_tokens_per_second": 702.136
    },
    {
      "epoch": 0.771291382360374,
      "grad_norm": 1.8085486888885498,
      "learning_rate": 1.2334177215189876e-05,
      "loss": 0.09838170558214188,
      "num_input_tokens_seen": 1440696,
      "step": 763,
      "train_runtime": 2051.7249,
      "train_tokens_per_second": 702.188
    },
    {
      "epoch": 0.7723022491786707,
      "grad_norm": 1.6283233165740967,
      "learning_rate": 1.2324050632911393e-05,
      "loss": 0.09942428767681122,
      "num_input_tokens_seen": 1442820,
      "step": 764,
      "train_runtime": 2054.0258,
      "train_tokens_per_second": 702.435
    },
    {
      "epoch": 0.7733131159969674,
      "grad_norm": 1.8893804550170898,
      "learning_rate": 1.2313924050632912e-05,
      "loss": 0.10178426653146744,
      "num_input_tokens_seen": 1445002,
      "step": 765,
      "train_runtime": 2056.3714,
      "train_tokens_per_second": 702.695
    },
    {
      "epoch": 0.774323982815264,
      "grad_norm": 1.7042813301086426,
      "learning_rate": 1.2303797468354431e-05,
      "loss": 0.09114382416009903,
      "num_input_tokens_seen": 1446852,
      "step": 766,
      "train_runtime": 2058.7022,
      "train_tokens_per_second": 702.798
    },
    {
      "epoch": 0.7753348496335608,
      "grad_norm": 1.401929259300232,
      "learning_rate": 1.229367088607595e-05,
      "loss": 0.058845121413469315,
      "num_input_tokens_seen": 1448588,
      "step": 767,
      "train_runtime": 2060.9668,
      "train_tokens_per_second": 702.868
    },
    {
      "epoch": 0.7763457164518575,
      "grad_norm": 1.7265045642852783,
      "learning_rate": 1.228354430379747e-05,
      "loss": 0.11025261878967285,
      "num_input_tokens_seen": 1450580,
      "step": 768,
      "train_runtime": 2063.1969,
      "train_tokens_per_second": 703.074
    },
    {
      "epoch": 0.7773565832701541,
      "grad_norm": 1.5884586572647095,
      "learning_rate": 1.2273417721518988e-05,
      "loss": 0.09003075212240219,
      "num_input_tokens_seen": 1452580,
      "step": 769,
      "train_runtime": 2065.5002,
      "train_tokens_per_second": 703.258
    },
    {
      "epoch": 0.7783674500884509,
      "grad_norm": 1.795130729675293,
      "learning_rate": 1.2263291139240508e-05,
      "loss": 0.13106182217597961,
      "num_input_tokens_seen": 1454246,
      "step": 770,
      "train_runtime": 2067.7271,
      "train_tokens_per_second": 703.307
    },
    {
      "epoch": 0.7793783169067475,
      "grad_norm": 1.8099521398544312,
      "learning_rate": 1.2253164556962027e-05,
      "loss": 0.10356763750314713,
      "num_input_tokens_seen": 1456172,
      "step": 771,
      "train_runtime": 2069.9755,
      "train_tokens_per_second": 703.473
    },
    {
      "epoch": 0.7803891837250442,
      "grad_norm": 1.536196231842041,
      "learning_rate": 1.2243037974683546e-05,
      "loss": 0.10875052213668823,
      "num_input_tokens_seen": 1458110,
      "step": 772,
      "train_runtime": 2072.1982,
      "train_tokens_per_second": 703.654
    },
    {
      "epoch": 0.7814000505433409,
      "grad_norm": 1.8153913021087646,
      "learning_rate": 1.2232911392405065e-05,
      "loss": 0.12295549362897873,
      "num_input_tokens_seen": 1459878,
      "step": 773,
      "train_runtime": 2074.3954,
      "train_tokens_per_second": 703.761
    },
    {
      "epoch": 0.7824109173616376,
      "grad_norm": 1.9409207105636597,
      "learning_rate": 1.2222784810126582e-05,
      "loss": 0.10511298477649689,
      "num_input_tokens_seen": 1461484,
      "step": 774,
      "train_runtime": 2076.601,
      "train_tokens_per_second": 703.787
    },
    {
      "epoch": 0.7834217841799342,
      "grad_norm": 1.710295557975769,
      "learning_rate": 1.2212658227848101e-05,
      "loss": 0.10304154455661774,
      "num_input_tokens_seen": 1463336,
      "step": 775,
      "train_runtime": 2078.8189,
      "train_tokens_per_second": 703.927
    },
    {
      "epoch": 0.784432650998231,
      "grad_norm": 1.9755862951278687,
      "learning_rate": 1.2202531645569622e-05,
      "loss": 0.08260329812765121,
      "num_input_tokens_seen": 1464984,
      "step": 776,
      "train_runtime": 2080.9981,
      "train_tokens_per_second": 703.981
    },
    {
      "epoch": 0.7854435178165277,
      "grad_norm": 2.2708795070648193,
      "learning_rate": 1.219240506329114e-05,
      "loss": 0.1592435985803604,
      "num_input_tokens_seen": 1466894,
      "step": 777,
      "train_runtime": 2083.1873,
      "train_tokens_per_second": 704.158
    },
    {
      "epoch": 0.7864543846348243,
      "grad_norm": 1.8243274688720703,
      "learning_rate": 1.2182278481012658e-05,
      "loss": 0.10652625560760498,
      "num_input_tokens_seen": 1468682,
      "step": 778,
      "train_runtime": 2085.3698,
      "train_tokens_per_second": 704.279
    },
    {
      "epoch": 0.7874652514531211,
      "grad_norm": 1.468176007270813,
      "learning_rate": 1.217215189873418e-05,
      "loss": 0.08889875560998917,
      "num_input_tokens_seen": 1470236,
      "step": 779,
      "train_runtime": 2087.492,
      "train_tokens_per_second": 704.307
    },
    {
      "epoch": 0.7884761182714177,
      "grad_norm": 1.6787092685699463,
      "learning_rate": 1.2162025316455697e-05,
      "loss": 0.08062463998794556,
      "num_input_tokens_seen": 1472040,
      "step": 780,
      "train_runtime": 2089.691,
      "train_tokens_per_second": 704.43
    },
    {
      "epoch": 0.7894869850897144,
      "grad_norm": 1.847848653793335,
      "learning_rate": 1.2151898734177216e-05,
      "loss": 0.0998370349407196,
      "num_input_tokens_seen": 1474404,
      "step": 781,
      "train_runtime": 2099.3086,
      "train_tokens_per_second": 702.328
    },
    {
      "epoch": 0.7904978519080111,
      "grad_norm": 1.9789323806762695,
      "learning_rate": 1.2141772151898735e-05,
      "loss": 0.13471730053424835,
      "num_input_tokens_seen": 1476702,
      "step": 782,
      "train_runtime": 2101.6403,
      "train_tokens_per_second": 702.643
    },
    {
      "epoch": 0.7915087187263078,
      "grad_norm": 1.543714165687561,
      "learning_rate": 1.2131645569620254e-05,
      "loss": 0.10380326211452484,
      "num_input_tokens_seen": 1478698,
      "step": 783,
      "train_runtime": 2103.8482,
      "train_tokens_per_second": 702.854
    },
    {
      "epoch": 0.7925195855446044,
      "grad_norm": 1.6039838790893555,
      "learning_rate": 1.2121518987341773e-05,
      "loss": 0.10127082467079163,
      "num_input_tokens_seen": 1480804,
      "step": 784,
      "train_runtime": 2106.0975,
      "train_tokens_per_second": 703.103
    },
    {
      "epoch": 0.7935304523629012,
      "grad_norm": 1.8809661865234375,
      "learning_rate": 1.2111392405063292e-05,
      "loss": 0.14595478773117065,
      "num_input_tokens_seen": 1482470,
      "step": 785,
      "train_runtime": 2108.248,
      "train_tokens_per_second": 703.176
    },
    {
      "epoch": 0.7945413191811979,
      "grad_norm": 1.524720549583435,
      "learning_rate": 1.2101265822784811e-05,
      "loss": 0.09433682262897491,
      "num_input_tokens_seen": 1484010,
      "step": 786,
      "train_runtime": 2110.3496,
      "train_tokens_per_second": 703.206
    },
    {
      "epoch": 0.7955521859994946,
      "grad_norm": 1.9089356660842896,
      "learning_rate": 1.209113924050633e-05,
      "loss": 0.12183887511491776,
      "num_input_tokens_seen": 1485590,
      "step": 787,
      "train_runtime": 2112.484,
      "train_tokens_per_second": 703.243
    },
    {
      "epoch": 0.7965630528177913,
      "grad_norm": 1.9138861894607544,
      "learning_rate": 1.208101265822785e-05,
      "loss": 0.11920693516731262,
      "num_input_tokens_seen": 1487662,
      "step": 788,
      "train_runtime": 2114.7519,
      "train_tokens_per_second": 703.469
    },
    {
      "epoch": 0.7975739196360879,
      "grad_norm": 1.7757669687271118,
      "learning_rate": 1.2070886075949368e-05,
      "loss": 0.15390071272850037,
      "num_input_tokens_seen": 1490132,
      "step": 789,
      "train_runtime": 2117.0738,
      "train_tokens_per_second": 703.864
    },
    {
      "epoch": 0.7985847864543847,
      "grad_norm": 1.7771114110946655,
      "learning_rate": 1.2060759493670886e-05,
      "loss": 0.09804435074329376,
      "num_input_tokens_seen": 1491952,
      "step": 790,
      "train_runtime": 2119.2554,
      "train_tokens_per_second": 703.998
    },
    {
      "epoch": 0.7995956532726813,
      "grad_norm": 1.7257702350616455,
      "learning_rate": 1.2050632911392407e-05,
      "loss": 0.09261702746152878,
      "num_input_tokens_seen": 1493938,
      "step": 791,
      "train_runtime": 2121.4735,
      "train_tokens_per_second": 704.198
    },
    {
      "epoch": 0.800606520090978,
      "grad_norm": 2.043962001800537,
      "learning_rate": 1.2040506329113926e-05,
      "loss": 0.11694172024726868,
      "num_input_tokens_seen": 1495656,
      "step": 792,
      "train_runtime": 2123.7914,
      "train_tokens_per_second": 704.239
    },
    {
      "epoch": 0.800606520090978,
      "eval_loss": 0.3946111798286438,
      "eval_runtime": 52.1824,
      "eval_samples_per_second": 16.845,
      "eval_steps_per_second": 8.432,
      "num_input_tokens_seen": 1495656,
      "step": 792
    },
    {
      "epoch": 0.8016173869092748,
      "grad_norm": 2.063417673110962,
      "learning_rate": 1.2030379746835443e-05,
      "loss": 0.12162243574857712,
      "num_input_tokens_seen": 1497184,
      "step": 793,
      "train_runtime": 2178.1274,
      "train_tokens_per_second": 687.372
    },
    {
      "epoch": 0.8026282537275714,
      "grad_norm": 1.5715523958206177,
      "learning_rate": 1.2020253164556964e-05,
      "loss": 0.06807722896337509,
      "num_input_tokens_seen": 1498650,
      "step": 794,
      "train_runtime": 2180.2449,
      "train_tokens_per_second": 687.377
    },
    {
      "epoch": 0.8036391205458681,
      "grad_norm": 1.8881858587265015,
      "learning_rate": 1.2010126582278483e-05,
      "loss": 0.1074509546160698,
      "num_input_tokens_seen": 1500418,
      "step": 795,
      "train_runtime": 2182.4027,
      "train_tokens_per_second": 687.507
    },
    {
      "epoch": 0.8046499873641648,
      "grad_norm": 1.501157283782959,
      "learning_rate": 1.2e-05,
      "loss": 0.09269465506076813,
      "num_input_tokens_seen": 1502404,
      "step": 796,
      "train_runtime": 2184.6473,
      "train_tokens_per_second": 687.71
    },
    {
      "epoch": 0.8056608541824615,
      "grad_norm": 1.39267098903656,
      "learning_rate": 1.1989873417721521e-05,
      "loss": 0.08595585823059082,
      "num_input_tokens_seen": 1504506,
      "step": 797,
      "train_runtime": 2186.885,
      "train_tokens_per_second": 687.968
    },
    {
      "epoch": 0.8066717210007581,
      "grad_norm": 1.6402462720870972,
      "learning_rate": 1.1979746835443038e-05,
      "loss": 0.09339673072099686,
      "num_input_tokens_seen": 1506334,
      "step": 798,
      "train_runtime": 2189.0674,
      "train_tokens_per_second": 688.117
    },
    {
      "epoch": 0.8076825878190549,
      "grad_norm": 1.8428109884262085,
      "learning_rate": 1.1969620253164558e-05,
      "loss": 0.1302780658006668,
      "num_input_tokens_seen": 1508208,
      "step": 799,
      "train_runtime": 2191.2815,
      "train_tokens_per_second": 688.277
    },
    {
      "epoch": 0.8086934546373515,
      "grad_norm": 1.6349430084228516,
      "learning_rate": 1.1959493670886078e-05,
      "loss": 0.09174898266792297,
      "num_input_tokens_seen": 1509974,
      "step": 800,
      "train_runtime": 2193.483,
      "train_tokens_per_second": 688.391
    },
    {
      "epoch": 0.8097043214556482,
      "grad_norm": 2.0324203968048096,
      "learning_rate": 1.1949367088607596e-05,
      "loss": 0.1389024704694748,
      "num_input_tokens_seen": 1511792,
      "step": 801,
      "train_runtime": 2195.6667,
      "train_tokens_per_second": 688.534
    },
    {
      "epoch": 0.810715188273945,
      "grad_norm": 1.9356026649475098,
      "learning_rate": 1.1939240506329115e-05,
      "loss": 0.11095651239156723,
      "num_input_tokens_seen": 1513666,
      "step": 802,
      "train_runtime": 2197.8771,
      "train_tokens_per_second": 688.695
    },
    {
      "epoch": 0.8117260550922416,
      "grad_norm": 1.6426819562911987,
      "learning_rate": 1.1929113924050636e-05,
      "loss": 0.10243912041187286,
      "num_input_tokens_seen": 1516078,
      "step": 803,
      "train_runtime": 2200.1985,
      "train_tokens_per_second": 689.064
    },
    {
      "epoch": 0.8127369219105383,
      "grad_norm": 1.7166860103607178,
      "learning_rate": 1.1918987341772153e-05,
      "loss": 0.08783088624477386,
      "num_input_tokens_seen": 1517678,
      "step": 804,
      "train_runtime": 2202.3235,
      "train_tokens_per_second": 689.126
    },
    {
      "epoch": 0.813747788728835,
      "grad_norm": 1.8186951875686646,
      "learning_rate": 1.1908860759493672e-05,
      "loss": 0.11066818237304688,
      "num_input_tokens_seen": 1519652,
      "step": 805,
      "train_runtime": 2204.5375,
      "train_tokens_per_second": 689.329
    },
    {
      "epoch": 0.8147586555471317,
      "grad_norm": 1.9489704370498657,
      "learning_rate": 1.189873417721519e-05,
      "loss": 0.10720035433769226,
      "num_input_tokens_seen": 1521208,
      "step": 806,
      "train_runtime": 2206.6554,
      "train_tokens_per_second": 689.373
    },
    {
      "epoch": 0.8157695223654283,
      "grad_norm": 1.7648284435272217,
      "learning_rate": 1.188860759493671e-05,
      "loss": 0.07367865741252899,
      "num_input_tokens_seen": 1523202,
      "step": 807,
      "train_runtime": 2208.87,
      "train_tokens_per_second": 689.584
    },
    {
      "epoch": 0.8167803891837251,
      "grad_norm": 2.136040449142456,
      "learning_rate": 1.187848101265823e-05,
      "loss": 0.13560248911380768,
      "num_input_tokens_seen": 1524950,
      "step": 808,
      "train_runtime": 2211.0494,
      "train_tokens_per_second": 689.695
    },
    {
      "epoch": 0.8177912560020217,
      "grad_norm": 1.6972389221191406,
      "learning_rate": 1.1868354430379747e-05,
      "loss": 0.08793585002422333,
      "num_input_tokens_seen": 1526390,
      "step": 809,
      "train_runtime": 2213.1484,
      "train_tokens_per_second": 689.692
    },
    {
      "epoch": 0.8188021228203184,
      "grad_norm": 1.5497407913208008,
      "learning_rate": 1.1858227848101267e-05,
      "loss": 0.10550501942634583,
      "num_input_tokens_seen": 1528232,
      "step": 810,
      "train_runtime": 2215.4065,
      "train_tokens_per_second": 689.82
    },
    {
      "epoch": 0.8198129896386152,
      "grad_norm": 1.8824611902236938,
      "learning_rate": 1.1848101265822787e-05,
      "loss": 0.11529868841171265,
      "num_input_tokens_seen": 1530056,
      "step": 811,
      "train_runtime": 2224.8161,
      "train_tokens_per_second": 687.722
    },
    {
      "epoch": 0.8208238564569118,
      "grad_norm": 1.964501976966858,
      "learning_rate": 1.1837974683544304e-05,
      "loss": 0.13124974071979523,
      "num_input_tokens_seen": 1532250,
      "step": 812,
      "train_runtime": 2227.1965,
      "train_tokens_per_second": 687.973
    },
    {
      "epoch": 0.8218347232752085,
      "grad_norm": 1.4976140260696411,
      "learning_rate": 1.1827848101265825e-05,
      "loss": 0.07861673831939697,
      "num_input_tokens_seen": 1534102,
      "step": 813,
      "train_runtime": 2229.4008,
      "train_tokens_per_second": 688.123
    },
    {
      "epoch": 0.8228455900935052,
      "grad_norm": 1.904380202293396,
      "learning_rate": 1.1817721518987342e-05,
      "loss": 0.09402893483638763,
      "num_input_tokens_seen": 1535648,
      "step": 814,
      "train_runtime": 2231.5509,
      "train_tokens_per_second": 688.153
    },
    {
      "epoch": 0.8238564569118019,
      "grad_norm": 1.507312297821045,
      "learning_rate": 1.1807594936708861e-05,
      "loss": 0.09126953780651093,
      "num_input_tokens_seen": 1537408,
      "step": 815,
      "train_runtime": 2233.8027,
      "train_tokens_per_second": 688.247
    },
    {
      "epoch": 0.8248673237300985,
      "grad_norm": 1.3340409994125366,
      "learning_rate": 1.1797468354430382e-05,
      "loss": 0.06178020313382149,
      "num_input_tokens_seen": 1539398,
      "step": 816,
      "train_runtime": 2236.0572,
      "train_tokens_per_second": 688.443
    },
    {
      "epoch": 0.8258781905483953,
      "grad_norm": 1.8956761360168457,
      "learning_rate": 1.17873417721519e-05,
      "loss": 0.10389361530542374,
      "num_input_tokens_seen": 1541054,
      "step": 817,
      "train_runtime": 2238.2117,
      "train_tokens_per_second": 688.52
    },
    {
      "epoch": 0.8268890573666919,
      "grad_norm": 1.2940924167633057,
      "learning_rate": 1.1777215189873418e-05,
      "loss": 0.0874934196472168,
      "num_input_tokens_seen": 1543138,
      "step": 818,
      "train_runtime": 2240.4517,
      "train_tokens_per_second": 688.762
    },
    {
      "epoch": 0.8278999241849886,
      "grad_norm": 1.4896349906921387,
      "learning_rate": 1.176708860759494e-05,
      "loss": 0.08060645312070847,
      "num_input_tokens_seen": 1544674,
      "step": 819,
      "train_runtime": 2242.603,
      "train_tokens_per_second": 688.786
    },
    {
      "epoch": 0.8289107910032854,
      "grad_norm": 1.979724407196045,
      "learning_rate": 1.1756962025316457e-05,
      "loss": 0.1410706639289856,
      "num_input_tokens_seen": 1546470,
      "step": 820,
      "train_runtime": 2244.7967,
      "train_tokens_per_second": 688.913
    },
    {
      "epoch": 0.829921657821582,
      "grad_norm": 1.7215502262115479,
      "learning_rate": 1.1746835443037976e-05,
      "loss": 0.11855530738830566,
      "num_input_tokens_seen": 1548530,
      "step": 821,
      "train_runtime": 2247.0656,
      "train_tokens_per_second": 689.134
    },
    {
      "epoch": 0.8309325246398787,
      "grad_norm": 1.9731192588806152,
      "learning_rate": 1.1736708860759493e-05,
      "loss": 0.1129259392619133,
      "num_input_tokens_seen": 1549984,
      "step": 822,
      "train_runtime": 2249.1843,
      "train_tokens_per_second": 689.132
    },
    {
      "epoch": 0.8319433914581754,
      "grad_norm": 1.5840198993682861,
      "learning_rate": 1.1726582278481014e-05,
      "loss": 0.07799205183982849,
      "num_input_tokens_seen": 1552118,
      "step": 823,
      "train_runtime": 2251.4513,
      "train_tokens_per_second": 689.386
    },
    {
      "epoch": 0.8329542582764721,
      "grad_norm": 1.7128299474716187,
      "learning_rate": 1.1716455696202533e-05,
      "loss": 0.12098036706447601,
      "num_input_tokens_seen": 1554850,
      "step": 824,
      "train_runtime": 2253.8871,
      "train_tokens_per_second": 689.853
    },
    {
      "epoch": 0.8339651250947687,
      "grad_norm": 1.6522421836853027,
      "learning_rate": 1.170632911392405e-05,
      "loss": 0.10111503303050995,
      "num_input_tokens_seen": 1556788,
      "step": 825,
      "train_runtime": 2256.1549,
      "train_tokens_per_second": 690.018
    },
    {
      "epoch": 0.8349759919130655,
      "grad_norm": 1.9402137994766235,
      "learning_rate": 1.1696202531645571e-05,
      "loss": 0.13037458062171936,
      "num_input_tokens_seen": 1558446,
      "step": 826,
      "train_runtime": 2258.3077,
      "train_tokens_per_second": 690.095
    },
    {
      "epoch": 0.8359868587313621,
      "grad_norm": 1.9426116943359375,
      "learning_rate": 1.168607594936709e-05,
      "loss": 0.12384985387325287,
      "num_input_tokens_seen": 1560354,
      "step": 827,
      "train_runtime": 2260.5189,
      "train_tokens_per_second": 690.264
    },
    {
      "epoch": 0.8369977255496588,
      "grad_norm": 2.1715128421783447,
      "learning_rate": 1.1675949367088608e-05,
      "loss": 0.15766561031341553,
      "num_input_tokens_seen": 1562322,
      "step": 828,
      "train_runtime": 2262.7235,
      "train_tokens_per_second": 690.461
    },
    {
      "epoch": 0.8380085923679556,
      "grad_norm": 2.079145908355713,
      "learning_rate": 1.1665822784810128e-05,
      "loss": 0.10856711119413376,
      "num_input_tokens_seen": 1564184,
      "step": 829,
      "train_runtime": 2264.9465,
      "train_tokens_per_second": 690.605
    },
    {
      "epoch": 0.8390194591862522,
      "grad_norm": 2.0513410568237305,
      "learning_rate": 1.1655696202531646e-05,
      "loss": 0.12570130825042725,
      "num_input_tokens_seen": 1565994,
      "step": 830,
      "train_runtime": 2267.1186,
      "train_tokens_per_second": 690.742
    },
    {
      "epoch": 0.8400303260045489,
      "grad_norm": 1.5659273862838745,
      "learning_rate": 1.1645569620253165e-05,
      "loss": 0.09665076434612274,
      "num_input_tokens_seen": 1567916,
      "step": 831,
      "train_runtime": 2269.2946,
      "train_tokens_per_second": 690.927
    },
    {
      "epoch": 0.8410411928228456,
      "grad_norm": 1.7730907201766968,
      "learning_rate": 1.1635443037974686e-05,
      "loss": 0.10670483112335205,
      "num_input_tokens_seen": 1569890,
      "step": 832,
      "train_runtime": 2271.5691,
      "train_tokens_per_second": 691.104
    },
    {
      "epoch": 0.8420520596411423,
      "grad_norm": 1.9787683486938477,
      "learning_rate": 1.1625316455696203e-05,
      "loss": 0.12965239584445953,
      "num_input_tokens_seen": 1572012,
      "step": 833,
      "train_runtime": 2273.8198,
      "train_tokens_per_second": 691.353
    },
    {
      "epoch": 0.8430629264594389,
      "grad_norm": 1.4029730558395386,
      "learning_rate": 1.1615189873417722e-05,
      "loss": 0.06591518968343735,
      "num_input_tokens_seen": 1573832,
      "step": 834,
      "train_runtime": 2276.0364,
      "train_tokens_per_second": 691.479
    },
    {
      "epoch": 0.8440737932777357,
      "grad_norm": 1.5094635486602783,
      "learning_rate": 1.1605063291139243e-05,
      "loss": 0.09049294143915176,
      "num_input_tokens_seen": 1575922,
      "step": 835,
      "train_runtime": 2278.2906,
      "train_tokens_per_second": 691.712
    },
    {
      "epoch": 0.8450846600960323,
      "grad_norm": 1.7904154062271118,
      "learning_rate": 1.159493670886076e-05,
      "loss": 0.12147239595651627,
      "num_input_tokens_seen": 1577630,
      "step": 836,
      "train_runtime": 2280.4466,
      "train_tokens_per_second": 691.807
    },
    {
      "epoch": 0.846095526914329,
      "grad_norm": 1.8628566265106201,
      "learning_rate": 1.158481012658228e-05,
      "loss": 0.08792897313833237,
      "num_input_tokens_seen": 1579326,
      "step": 837,
      "train_runtime": 2282.6114,
      "train_tokens_per_second": 691.894
    },
    {
      "epoch": 0.8471063937326258,
      "grad_norm": 1.6908202171325684,
      "learning_rate": 1.1574683544303797e-05,
      "loss": 0.09818118065595627,
      "num_input_tokens_seen": 1581406,
      "step": 838,
      "train_runtime": 2284.8464,
      "train_tokens_per_second": 692.128
    },
    {
      "epoch": 0.8481172605509224,
      "grad_norm": 1.7988734245300293,
      "learning_rate": 1.1564556962025318e-05,
      "loss": 0.10540547966957092,
      "num_input_tokens_seen": 1583126,
      "step": 839,
      "train_runtime": 2286.9908,
      "train_tokens_per_second": 692.231
    },
    {
      "epoch": 0.849128127369219,
      "grad_norm": 1.7324339151382446,
      "learning_rate": 1.1554430379746837e-05,
      "loss": 0.11427002400159836,
      "num_input_tokens_seen": 1585318,
      "step": 840,
      "train_runtime": 2289.309,
      "train_tokens_per_second": 692.488
    },
    {
      "epoch": 0.8501389941875158,
      "grad_norm": 1.4927492141723633,
      "learning_rate": 1.1544303797468354e-05,
      "loss": 0.06348370015621185,
      "num_input_tokens_seen": 1586846,
      "step": 841,
      "train_runtime": 2298.5765,
      "train_tokens_per_second": 690.36
    },
    {
      "epoch": 0.8511498610058125,
      "grad_norm": 2.1162431240081787,
      "learning_rate": 1.1534177215189875e-05,
      "loss": 0.09501577168703079,
      "num_input_tokens_seen": 1588624,
      "step": 842,
      "train_runtime": 2300.7384,
      "train_tokens_per_second": 690.484
    },
    {
      "epoch": 0.8521607278241091,
      "grad_norm": 1.9396066665649414,
      "learning_rate": 1.1524050632911394e-05,
      "loss": 0.10647747665643692,
      "num_input_tokens_seen": 1590574,
      "step": 843,
      "train_runtime": 2303.0593,
      "train_tokens_per_second": 690.635
    },
    {
      "epoch": 0.8531715946424059,
      "grad_norm": 2.0144662857055664,
      "learning_rate": 1.1513924050632911e-05,
      "loss": 0.09945391118526459,
      "num_input_tokens_seen": 1592676,
      "step": 844,
      "train_runtime": 2305.2867,
      "train_tokens_per_second": 690.88
    },
    {
      "epoch": 0.8541824614607025,
      "grad_norm": 1.7529017925262451,
      "learning_rate": 1.1503797468354432e-05,
      "loss": 0.09879845380783081,
      "num_input_tokens_seen": 1594634,
      "step": 845,
      "train_runtime": 2307.5112,
      "train_tokens_per_second": 691.062
    },
    {
      "epoch": 0.8551933282789992,
      "grad_norm": 2.058156967163086,
      "learning_rate": 1.149367088607595e-05,
      "loss": 0.0892823114991188,
      "num_input_tokens_seen": 1596328,
      "step": 846,
      "train_runtime": 2309.6959,
      "train_tokens_per_second": 691.142
    },
    {
      "epoch": 0.856204195097296,
      "grad_norm": 1.6276295185089111,
      "learning_rate": 1.1483544303797469e-05,
      "loss": 0.12782657146453857,
      "num_input_tokens_seen": 1598918,
      "step": 847,
      "train_runtime": 2312.0539,
      "train_tokens_per_second": 691.557
    },
    {
      "epoch": 0.8572150619155926,
      "grad_norm": 1.9061850309371948,
      "learning_rate": 1.147341772151899e-05,
      "loss": 0.13704827427864075,
      "num_input_tokens_seen": 1600810,
      "step": 848,
      "train_runtime": 2314.2804,
      "train_tokens_per_second": 691.71
    },
    {
      "epoch": 0.8582259287338894,
      "grad_norm": 1.7898938655853271,
      "learning_rate": 1.1463291139240507e-05,
      "loss": 0.1129920706152916,
      "num_input_tokens_seen": 1602868,
      "step": 849,
      "train_runtime": 2316.4976,
      "train_tokens_per_second": 691.936
    },
    {
      "epoch": 0.859236795552186,
      "grad_norm": 1.886339783668518,
      "learning_rate": 1.1453164556962026e-05,
      "loss": 0.09292490035295486,
      "num_input_tokens_seen": 1604334,
      "step": 850,
      "train_runtime": 2318.6179,
      "train_tokens_per_second": 691.935
    },
    {
      "epoch": 0.8602476623704827,
      "grad_norm": 2.0642142295837402,
      "learning_rate": 1.1443037974683547e-05,
      "loss": 0.13143715262413025,
      "num_input_tokens_seen": 1606456,
      "step": 851,
      "train_runtime": 2320.8556,
      "train_tokens_per_second": 692.183
    },
    {
      "epoch": 0.8612585291887794,
      "grad_norm": 1.727893352508545,
      "learning_rate": 1.1432911392405064e-05,
      "loss": 0.10075180232524872,
      "num_input_tokens_seen": 1608280,
      "step": 852,
      "train_runtime": 2323.0335,
      "train_tokens_per_second": 692.319
    },
    {
      "epoch": 0.8622693960070761,
      "grad_norm": 1.4375556707382202,
      "learning_rate": 1.1422784810126583e-05,
      "loss": 0.0873575434088707,
      "num_input_tokens_seen": 1609990,
      "step": 853,
      "train_runtime": 2325.2215,
      "train_tokens_per_second": 692.403
    },
    {
      "epoch": 0.8632802628253727,
      "grad_norm": 1.7615019083023071,
      "learning_rate": 1.1412658227848102e-05,
      "loss": 0.1255250871181488,
      "num_input_tokens_seen": 1611726,
      "step": 854,
      "train_runtime": 2327.4046,
      "train_tokens_per_second": 692.499
    },
    {
      "epoch": 0.8642911296436695,
      "grad_norm": 1.795871615409851,
      "learning_rate": 1.1402531645569621e-05,
      "loss": 0.11938530951738358,
      "num_input_tokens_seen": 1613962,
      "step": 855,
      "train_runtime": 2329.7183,
      "train_tokens_per_second": 692.771
    },
    {
      "epoch": 0.8653019964619661,
      "grad_norm": 1.8448328971862793,
      "learning_rate": 1.139240506329114e-05,
      "loss": 0.09347003698348999,
      "num_input_tokens_seen": 1615632,
      "step": 856,
      "train_runtime": 2331.8939,
      "train_tokens_per_second": 692.841
    },
    {
      "epoch": 0.8663128632802628,
      "grad_norm": 2.698441505432129,
      "learning_rate": 1.138227848101266e-05,
      "loss": 0.09510399401187897,
      "num_input_tokens_seen": 1617550,
      "step": 857,
      "train_runtime": 2334.1389,
      "train_tokens_per_second": 692.996
    },
    {
      "epoch": 0.8673237300985596,
      "grad_norm": 1.5858604907989502,
      "learning_rate": 1.1372151898734178e-05,
      "loss": 0.10066479444503784,
      "num_input_tokens_seen": 1619282,
      "step": 858,
      "train_runtime": 2336.3435,
      "train_tokens_per_second": 693.084
    },
    {
      "epoch": 0.8683345969168562,
      "grad_norm": 1.9796504974365234,
      "learning_rate": 1.1362025316455698e-05,
      "loss": 0.11774136871099472,
      "num_input_tokens_seen": 1620816,
      "step": 859,
      "train_runtime": 2338.5267,
      "train_tokens_per_second": 693.093
    },
    {
      "epoch": 0.8693454637351529,
      "grad_norm": 1.4962421655654907,
      "learning_rate": 1.1351898734177217e-05,
      "loss": 0.10000722110271454,
      "num_input_tokens_seen": 1622940,
      "step": 860,
      "train_runtime": 2340.8249,
      "train_tokens_per_second": 693.32
    },
    {
      "epoch": 0.8703563305534496,
      "grad_norm": 1.5053681135177612,
      "learning_rate": 1.1341772151898736e-05,
      "loss": 0.09694317728281021,
      "num_input_tokens_seen": 1624928,
      "step": 861,
      "train_runtime": 2343.0777,
      "train_tokens_per_second": 693.502
    },
    {
      "epoch": 0.8713671973717463,
      "grad_norm": 1.6079481840133667,
      "learning_rate": 1.1331645569620253e-05,
      "loss": 0.1144719272851944,
      "num_input_tokens_seen": 1626842,
      "step": 862,
      "train_runtime": 2345.2893,
      "train_tokens_per_second": 693.664
    },
    {
      "epoch": 0.8723780641900429,
      "grad_norm": 1.7395708560943604,
      "learning_rate": 1.1321518987341772e-05,
      "loss": 0.10291267931461334,
      "num_input_tokens_seen": 1628466,
      "step": 863,
      "train_runtime": 2347.6038,
      "train_tokens_per_second": 693.672
    },
    {
      "epoch": 0.8733889310083397,
      "grad_norm": 1.8408080339431763,
      "learning_rate": 1.1311392405063293e-05,
      "loss": 0.09824563562870026,
      "num_input_tokens_seen": 1630404,
      "step": 864,
      "train_runtime": 2349.8853,
      "train_tokens_per_second": 693.823
    },
    {
      "epoch": 0.8743997978266363,
      "grad_norm": 1.6320384740829468,
      "learning_rate": 1.130126582278481e-05,
      "loss": 0.10983906686306,
      "num_input_tokens_seen": 1632534,
      "step": 865,
      "train_runtime": 2352.2028,
      "train_tokens_per_second": 694.045
    },
    {
      "epoch": 0.875410664644933,
      "grad_norm": 2.0745017528533936,
      "learning_rate": 1.129113924050633e-05,
      "loss": 0.09887329488992691,
      "num_input_tokens_seen": 1634250,
      "step": 866,
      "train_runtime": 2354.4418,
      "train_tokens_per_second": 694.114
    },
    {
      "epoch": 0.8764215314632298,
      "grad_norm": 1.4929437637329102,
      "learning_rate": 1.128101265822785e-05,
      "loss": 0.09026893973350525,
      "num_input_tokens_seen": 1636256,
      "step": 867,
      "train_runtime": 2356.6751,
      "train_tokens_per_second": 694.307
    },
    {
      "epoch": 0.8774323982815264,
      "grad_norm": 1.5981075763702393,
      "learning_rate": 1.1270886075949368e-05,
      "loss": 0.11353608220815659,
      "num_input_tokens_seen": 1638308,
      "step": 868,
      "train_runtime": 2358.9372,
      "train_tokens_per_second": 694.511
    },
    {
      "epoch": 0.8784432650998231,
      "grad_norm": 1.4297696352005005,
      "learning_rate": 1.1260759493670887e-05,
      "loss": 0.06749273836612701,
      "num_input_tokens_seen": 1639966,
      "step": 869,
      "train_runtime": 2361.116,
      "train_tokens_per_second": 694.572
    },
    {
      "epoch": 0.8794541319181198,
      "grad_norm": 1.5817461013793945,
      "learning_rate": 1.1250632911392406e-05,
      "loss": 0.10096173733472824,
      "num_input_tokens_seen": 1642326,
      "step": 870,
      "train_runtime": 2363.4636,
      "train_tokens_per_second": 694.881
    },
    {
      "epoch": 0.8804649987364165,
      "grad_norm": 1.586337924003601,
      "learning_rate": 1.1240506329113925e-05,
      "loss": 0.07885732501745224,
      "num_input_tokens_seen": 1643814,
      "step": 871,
      "train_runtime": 2372.7965,
      "train_tokens_per_second": 692.775
    },
    {
      "epoch": 0.8814758655547131,
      "grad_norm": 1.8464782238006592,
      "learning_rate": 1.1230379746835444e-05,
      "loss": 0.09361685067415237,
      "num_input_tokens_seen": 1645608,
      "step": 872,
      "train_runtime": 2375.1096,
      "train_tokens_per_second": 692.856
    },
    {
      "epoch": 0.8824867323730099,
      "grad_norm": 2.068006992340088,
      "learning_rate": 1.1220253164556963e-05,
      "loss": 0.13546811044216156,
      "num_input_tokens_seen": 1647372,
      "step": 873,
      "train_runtime": 2377.3765,
      "train_tokens_per_second": 692.937
    },
    {
      "epoch": 0.8834975991913065,
      "grad_norm": 2.653832197189331,
      "learning_rate": 1.1210126582278482e-05,
      "loss": 0.13507585227489471,
      "num_input_tokens_seen": 1649480,
      "step": 874,
      "train_runtime": 2379.7096,
      "train_tokens_per_second": 693.143
    },
    {
      "epoch": 0.8845084660096032,
      "grad_norm": 1.9615648984909058,
      "learning_rate": 1.1200000000000001e-05,
      "loss": 0.1325993537902832,
      "num_input_tokens_seen": 1651222,
      "step": 875,
      "train_runtime": 2381.9222,
      "train_tokens_per_second": 693.231
    },
    {
      "epoch": 0.8855193328279,
      "grad_norm": 1.3614580631256104,
      "learning_rate": 1.118987341772152e-05,
      "loss": 0.08537785708904266,
      "num_input_tokens_seen": 1653380,
      "step": 876,
      "train_runtime": 2384.2121,
      "train_tokens_per_second": 693.47
    },
    {
      "epoch": 0.8865301996461966,
      "grad_norm": 2.209545612335205,
      "learning_rate": 1.117974683544304e-05,
      "loss": 0.12375745922327042,
      "num_input_tokens_seen": 1654968,
      "step": 877,
      "train_runtime": 2386.3402,
      "train_tokens_per_second": 693.517
    },
    {
      "epoch": 0.8875410664644933,
      "grad_norm": 1.984891414642334,
      "learning_rate": 1.1169620253164557e-05,
      "loss": 0.10398299992084503,
      "num_input_tokens_seen": 1656582,
      "step": 878,
      "train_runtime": 2388.4894,
      "train_tokens_per_second": 693.569
    },
    {
      "epoch": 0.88855193328279,
      "grad_norm": 1.7964671850204468,
      "learning_rate": 1.1159493670886078e-05,
      "loss": 0.09898118674755096,
      "num_input_tokens_seen": 1658438,
      "step": 879,
      "train_runtime": 2390.7287,
      "train_tokens_per_second": 693.696
    },
    {
      "epoch": 0.8895628001010867,
      "grad_norm": 1.7326607704162598,
      "learning_rate": 1.1149367088607597e-05,
      "loss": 0.12106341868638992,
      "num_input_tokens_seen": 1660068,
      "step": 880,
      "train_runtime": 2392.871,
      "train_tokens_per_second": 693.756
    },
    {
      "epoch": 0.8905736669193833,
      "grad_norm": 1.5902682542800903,
      "learning_rate": 1.1139240506329114e-05,
      "loss": 0.09547918289899826,
      "num_input_tokens_seen": 1661494,
      "step": 881,
      "train_runtime": 2395.0434,
      "train_tokens_per_second": 693.722
    },
    {
      "epoch": 0.8915845337376801,
      "grad_norm": 1.4557594060897827,
      "learning_rate": 1.1129113924050635e-05,
      "loss": 0.05885666236281395,
      "num_input_tokens_seen": 1663150,
      "step": 882,
      "train_runtime": 2397.2304,
      "train_tokens_per_second": 693.78
    },
    {
      "epoch": 0.8925954005559767,
      "grad_norm": 1.624995470046997,
      "learning_rate": 1.1118987341772154e-05,
      "loss": 0.089319609105587,
      "num_input_tokens_seen": 1664702,
      "step": 883,
      "train_runtime": 2399.3751,
      "train_tokens_per_second": 693.806
    },
    {
      "epoch": 0.8936062673742734,
      "grad_norm": 1.4931955337524414,
      "learning_rate": 1.1108860759493671e-05,
      "loss": 0.09750652313232422,
      "num_input_tokens_seen": 1667500,
      "step": 884,
      "train_runtime": 2401.7873,
      "train_tokens_per_second": 694.275
    },
    {
      "epoch": 0.8946171341925702,
      "grad_norm": 1.4132208824157715,
      "learning_rate": 1.1098734177215192e-05,
      "loss": 0.08524797856807709,
      "num_input_tokens_seen": 1669594,
      "step": 885,
      "train_runtime": 2404.0569,
      "train_tokens_per_second": 694.49
    },
    {
      "epoch": 0.8956280010108668,
      "grad_norm": 2.098870038986206,
      "learning_rate": 1.108860759493671e-05,
      "loss": 0.10673464834690094,
      "num_input_tokens_seen": 1671040,
      "step": 886,
      "train_runtime": 2406.1485,
      "train_tokens_per_second": 694.487
    },
    {
      "epoch": 0.8966388678291635,
      "grad_norm": 1.7068244218826294,
      "learning_rate": 1.1078481012658229e-05,
      "loss": 0.09890075027942657,
      "num_input_tokens_seen": 1672500,
      "step": 887,
      "train_runtime": 2408.3028,
      "train_tokens_per_second": 694.472
    },
    {
      "epoch": 0.8976497346474602,
      "grad_norm": 1.5627840757369995,
      "learning_rate": 1.106835443037975e-05,
      "loss": 0.08412830531597137,
      "num_input_tokens_seen": 1674008,
      "step": 888,
      "train_runtime": 2410.4956,
      "train_tokens_per_second": 694.466
    },
    {
      "epoch": 0.8986606014657569,
      "grad_norm": 1.4462611675262451,
      "learning_rate": 1.1058227848101267e-05,
      "loss": 0.07267887145280838,
      "num_input_tokens_seen": 1676148,
      "step": 889,
      "train_runtime": 2412.8538,
      "train_tokens_per_second": 694.674
    },
    {
      "epoch": 0.8996714682840535,
      "grad_norm": 2.0984222888946533,
      "learning_rate": 1.1048101265822786e-05,
      "loss": 0.14878182113170624,
      "num_input_tokens_seen": 1677864,
      "step": 890,
      "train_runtime": 2415.0363,
      "train_tokens_per_second": 694.757
    },
    {
      "epoch": 0.9006823351023503,
      "grad_norm": 1.4134807586669922,
      "learning_rate": 1.1037974683544307e-05,
      "loss": 0.06635203212499619,
      "num_input_tokens_seen": 1679694,
      "step": 891,
      "train_runtime": 2417.2867,
      "train_tokens_per_second": 694.868
    },
    {
      "epoch": 0.901693201920647,
      "grad_norm": 1.820233702659607,
      "learning_rate": 1.1027848101265824e-05,
      "loss": 0.12284374982118607,
      "num_input_tokens_seen": 1681590,
      "step": 892,
      "train_runtime": 2419.5393,
      "train_tokens_per_second": 695.004
    },
    {
      "epoch": 0.9027040687389436,
      "grad_norm": 1.402327299118042,
      "learning_rate": 1.1017721518987343e-05,
      "loss": 0.06556926667690277,
      "num_input_tokens_seen": 1683350,
      "step": 893,
      "train_runtime": 2421.7277,
      "train_tokens_per_second": 695.103
    },
    {
      "epoch": 0.9037149355572404,
      "grad_norm": 1.325697422027588,
      "learning_rate": 1.100759493670886e-05,
      "loss": 0.06861738860607147,
      "num_input_tokens_seen": 1685174,
      "step": 894,
      "train_runtime": 2423.919,
      "train_tokens_per_second": 695.227
    },
    {
      "epoch": 0.904725802375537,
      "grad_norm": 1.6240726709365845,
      "learning_rate": 1.0997468354430381e-05,
      "loss": 0.10145942866802216,
      "num_input_tokens_seen": 1687326,
      "step": 895,
      "train_runtime": 2426.1683,
      "train_tokens_per_second": 695.469
    },
    {
      "epoch": 0.9057366691938337,
      "grad_norm": 1.628271222114563,
      "learning_rate": 1.09873417721519e-05,
      "loss": 0.09534229338169098,
      "num_input_tokens_seen": 1689160,
      "step": 896,
      "train_runtime": 2428.3708,
      "train_tokens_per_second": 695.594
    },
    {
      "epoch": 0.9067475360121304,
      "grad_norm": 2.222989797592163,
      "learning_rate": 1.0977215189873418e-05,
      "loss": 0.08482712507247925,
      "num_input_tokens_seen": 1690894,
      "step": 897,
      "train_runtime": 2430.5452,
      "train_tokens_per_second": 695.685
    },
    {
      "epoch": 0.9077584028304271,
      "grad_norm": 1.6820749044418335,
      "learning_rate": 1.0967088607594938e-05,
      "loss": 0.09884992986917496,
      "num_input_tokens_seen": 1692400,
      "step": 898,
      "train_runtime": 2432.691,
      "train_tokens_per_second": 695.69
    },
    {
      "epoch": 0.9087692696487237,
      "grad_norm": 1.7538076639175415,
      "learning_rate": 1.0956962025316458e-05,
      "loss": 0.10634830594062805,
      "num_input_tokens_seen": 1694112,
      "step": 899,
      "train_runtime": 2434.8546,
      "train_tokens_per_second": 695.775
    },
    {
      "epoch": 0.9097801364670205,
      "grad_norm": 1.9143922328948975,
      "learning_rate": 1.0946835443037975e-05,
      "loss": 0.13174308836460114,
      "num_input_tokens_seen": 1695600,
      "step": 900,
      "train_runtime": 2436.9476,
      "train_tokens_per_second": 695.788
    },
    {
      "epoch": 0.9107910032853171,
      "grad_norm": 1.4580113887786865,
      "learning_rate": 1.0936708860759496e-05,
      "loss": 0.10258884727954865,
      "num_input_tokens_seen": 1697734,
      "step": 901,
      "train_runtime": 2446.3717,
      "train_tokens_per_second": 693.98
    },
    {
      "epoch": 0.9118018701036138,
      "grad_norm": 1.4541841745376587,
      "learning_rate": 1.0926582278481013e-05,
      "loss": 0.08640728145837784,
      "num_input_tokens_seen": 1699566,
      "step": 902,
      "train_runtime": 2448.5989,
      "train_tokens_per_second": 694.097
    },
    {
      "epoch": 0.9128127369219106,
      "grad_norm": 1.6778219938278198,
      "learning_rate": 1.0916455696202532e-05,
      "loss": 0.08899082988500595,
      "num_input_tokens_seen": 1701632,
      "step": 903,
      "train_runtime": 2450.8212,
      "train_tokens_per_second": 694.311
    },
    {
      "epoch": 0.9138236037402072,
      "grad_norm": 1.7936662435531616,
      "learning_rate": 1.0906329113924053e-05,
      "loss": 0.09717955440282822,
      "num_input_tokens_seen": 1703610,
      "step": 904,
      "train_runtime": 2453.057,
      "train_tokens_per_second": 694.484
    },
    {
      "epoch": 0.9148344705585039,
      "grad_norm": 1.6489940881729126,
      "learning_rate": 1.089620253164557e-05,
      "loss": 0.09587058424949646,
      "num_input_tokens_seen": 1705760,
      "step": 905,
      "train_runtime": 2455.2887,
      "train_tokens_per_second": 694.729
    },
    {
      "epoch": 0.9158453373768006,
      "grad_norm": 1.867157220840454,
      "learning_rate": 1.088607594936709e-05,
      "loss": 0.1219547688961029,
      "num_input_tokens_seen": 1708032,
      "step": 906,
      "train_runtime": 2457.5743,
      "train_tokens_per_second": 695.007
    },
    {
      "epoch": 0.9168562041950973,
      "grad_norm": 2.6823201179504395,
      "learning_rate": 1.0875949367088607e-05,
      "loss": 0.15664678812026978,
      "num_input_tokens_seen": 1709656,
      "step": 907,
      "train_runtime": 2459.7658,
      "train_tokens_per_second": 695.048
    },
    {
      "epoch": 0.9178670710133939,
      "grad_norm": 1.6702855825424194,
      "learning_rate": 1.0865822784810128e-05,
      "loss": 0.09947992861270905,
      "num_input_tokens_seen": 1711488,
      "step": 908,
      "train_runtime": 2461.9467,
      "train_tokens_per_second": 695.177
    },
    {
      "epoch": 0.9188779378316907,
      "grad_norm": 1.5098493099212646,
      "learning_rate": 1.0855696202531647e-05,
      "loss": 0.07601018249988556,
      "num_input_tokens_seen": 1713238,
      "step": 909,
      "train_runtime": 2464.1187,
      "train_tokens_per_second": 695.274
    },
    {
      "epoch": 0.9198888046499873,
      "grad_norm": 1.5049681663513184,
      "learning_rate": 1.0845569620253164e-05,
      "loss": 0.08372560143470764,
      "num_input_tokens_seen": 1714968,
      "step": 910,
      "train_runtime": 2466.2875,
      "train_tokens_per_second": 695.364
    },
    {
      "epoch": 0.9208996714682841,
      "grad_norm": 1.6285735368728638,
      "learning_rate": 1.0835443037974685e-05,
      "loss": 0.07884393632411957,
      "num_input_tokens_seen": 1716480,
      "step": 911,
      "train_runtime": 2468.4271,
      "train_tokens_per_second": 695.374
    },
    {
      "epoch": 0.9219105382865808,
      "grad_norm": 1.430666208267212,
      "learning_rate": 1.0825316455696204e-05,
      "loss": 0.06769224256277084,
      "num_input_tokens_seen": 1718392,
      "step": 912,
      "train_runtime": 2470.7027,
      "train_tokens_per_second": 695.507
    },
    {
      "epoch": 0.9229214051048774,
      "grad_norm": 1.9251612424850464,
      "learning_rate": 1.0815189873417721e-05,
      "loss": 0.11051812767982483,
      "num_input_tokens_seen": 1720378,
      "step": 913,
      "train_runtime": 2473.1359,
      "train_tokens_per_second": 695.626
    },
    {
      "epoch": 0.9239322719231742,
      "grad_norm": 1.7929903268814087,
      "learning_rate": 1.0805063291139242e-05,
      "loss": 0.12665323913097382,
      "num_input_tokens_seen": 1722602,
      "step": 914,
      "train_runtime": 2475.4055,
      "train_tokens_per_second": 695.887
    },
    {
      "epoch": 0.9249431387414708,
      "grad_norm": 1.718106985092163,
      "learning_rate": 1.079493670886076e-05,
      "loss": 0.0947490707039833,
      "num_input_tokens_seen": 1724174,
      "step": 915,
      "train_runtime": 2477.5849,
      "train_tokens_per_second": 695.909
    },
    {
      "epoch": 0.9259540055597675,
      "grad_norm": 1.8282264471054077,
      "learning_rate": 1.0784810126582279e-05,
      "loss": 0.11499021202325821,
      "num_input_tokens_seen": 1726094,
      "step": 916,
      "train_runtime": 2479.8114,
      "train_tokens_per_second": 696.059
    },
    {
      "epoch": 0.9269648723780642,
      "grad_norm": 1.7916219234466553,
      "learning_rate": 1.07746835443038e-05,
      "loss": 0.12264146655797958,
      "num_input_tokens_seen": 1728142,
      "step": 917,
      "train_runtime": 2482.0531,
      "train_tokens_per_second": 696.255
    },
    {
      "epoch": 0.9279757391963609,
      "grad_norm": 1.693138599395752,
      "learning_rate": 1.0764556962025317e-05,
      "loss": 0.09038832783699036,
      "num_input_tokens_seen": 1729944,
      "step": 918,
      "train_runtime": 2484.2402,
      "train_tokens_per_second": 696.367
    },
    {
      "epoch": 0.9289866060146575,
      "grad_norm": 1.6037993431091309,
      "learning_rate": 1.0754430379746836e-05,
      "loss": 0.08162438869476318,
      "num_input_tokens_seen": 1731584,
      "step": 919,
      "train_runtime": 2486.4928,
      "train_tokens_per_second": 696.396
    },
    {
      "epoch": 0.9299974728329543,
      "grad_norm": 1.5550346374511719,
      "learning_rate": 1.0744303797468357e-05,
      "loss": 0.08870391547679901,
      "num_input_tokens_seen": 1733392,
      "step": 920,
      "train_runtime": 2488.711,
      "train_tokens_per_second": 696.502
    },
    {
      "epoch": 0.931008339651251,
      "grad_norm": 1.3631558418273926,
      "learning_rate": 1.0734177215189874e-05,
      "loss": 0.07918770611286163,
      "num_input_tokens_seen": 1735478,
      "step": 921,
      "train_runtime": 2491.0273,
      "train_tokens_per_second": 696.692
    },
    {
      "epoch": 0.9320192064695476,
      "grad_norm": 1.8613494634628296,
      "learning_rate": 1.0724050632911393e-05,
      "loss": 0.11431507766246796,
      "num_input_tokens_seen": 1737024,
      "step": 922,
      "train_runtime": 2493.1602,
      "train_tokens_per_second": 696.716
    },
    {
      "epoch": 0.9330300732878444,
      "grad_norm": 1.7264175415039062,
      "learning_rate": 1.071392405063291e-05,
      "loss": 0.0920799970626831,
      "num_input_tokens_seen": 1738466,
      "step": 923,
      "train_runtime": 2495.2574,
      "train_tokens_per_second": 696.708
    },
    {
      "epoch": 0.934040940106141,
      "grad_norm": 1.6748762130737305,
      "learning_rate": 1.0703797468354431e-05,
      "loss": 0.09041954576969147,
      "num_input_tokens_seen": 1740182,
      "step": 924,
      "train_runtime": 2497.4465,
      "train_tokens_per_second": 696.785
    },
    {
      "epoch": 0.9350518069244377,
      "grad_norm": 1.7656514644622803,
      "learning_rate": 1.069367088607595e-05,
      "loss": 0.09037000685930252,
      "num_input_tokens_seen": 1742024,
      "step": 925,
      "train_runtime": 2499.6331,
      "train_tokens_per_second": 696.912
    },
    {
      "epoch": 0.9360626737427344,
      "grad_norm": 1.6852999925613403,
      "learning_rate": 1.0683544303797468e-05,
      "loss": 0.08460486680269241,
      "num_input_tokens_seen": 1744238,
      "step": 926,
      "train_runtime": 2501.912,
      "train_tokens_per_second": 697.162
    },
    {
      "epoch": 0.9370735405610311,
      "grad_norm": 2.154926300048828,
      "learning_rate": 1.0673417721518988e-05,
      "loss": 0.11571793258190155,
      "num_input_tokens_seen": 1746118,
      "step": 927,
      "train_runtime": 2504.0921,
      "train_tokens_per_second": 697.306
    },
    {
      "epoch": 0.9380844073793277,
      "grad_norm": 1.6266175508499146,
      "learning_rate": 1.0663291139240508e-05,
      "loss": 0.0903712585568428,
      "num_input_tokens_seen": 1748128,
      "step": 928,
      "train_runtime": 2506.3025,
      "train_tokens_per_second": 697.493
    },
    {
      "epoch": 0.9390952741976245,
      "grad_norm": 1.5624279975891113,
      "learning_rate": 1.0653164556962025e-05,
      "loss": 0.08536509424448013,
      "num_input_tokens_seen": 1749964,
      "step": 929,
      "train_runtime": 2508.4779,
      "train_tokens_per_second": 697.62
    },
    {
      "epoch": 0.9401061410159212,
      "grad_norm": 2.820284605026245,
      "learning_rate": 1.0643037974683546e-05,
      "loss": 0.1169944703578949,
      "num_input_tokens_seen": 1752108,
      "step": 930,
      "train_runtime": 2510.7292,
      "train_tokens_per_second": 697.848
    },
    {
      "epoch": 0.9411170078342178,
      "grad_norm": 1.3601627349853516,
      "learning_rate": 1.0632911392405063e-05,
      "loss": 0.06507647782564163,
      "num_input_tokens_seen": 1753674,
      "step": 931,
      "train_runtime": 2520.0221,
      "train_tokens_per_second": 695.896
    },
    {
      "epoch": 0.9421278746525146,
      "grad_norm": 2.0614519119262695,
      "learning_rate": 1.0622784810126582e-05,
      "loss": 0.14873051643371582,
      "num_input_tokens_seen": 1755348,
      "step": 932,
      "train_runtime": 2522.189,
      "train_tokens_per_second": 695.962
    },
    {
      "epoch": 0.9431387414708112,
      "grad_norm": 1.4850870370864868,
      "learning_rate": 1.0612658227848103e-05,
      "loss": 0.07789217680692673,
      "num_input_tokens_seen": 1757138,
      "step": 933,
      "train_runtime": 2524.3834,
      "train_tokens_per_second": 696.066
    },
    {
      "epoch": 0.9441496082891079,
      "grad_norm": 1.842126488685608,
      "learning_rate": 1.060253164556962e-05,
      "loss": 0.12018195539712906,
      "num_input_tokens_seen": 1759246,
      "step": 934,
      "train_runtime": 2526.6456,
      "train_tokens_per_second": 696.277
    },
    {
      "epoch": 0.9451604751074046,
      "grad_norm": 1.6177433729171753,
      "learning_rate": 1.059240506329114e-05,
      "loss": 0.10012046992778778,
      "num_input_tokens_seen": 1760924,
      "step": 935,
      "train_runtime": 2528.8352,
      "train_tokens_per_second": 696.338
    },
    {
      "epoch": 0.9461713419257013,
      "grad_norm": 1.530370831489563,
      "learning_rate": 1.058227848101266e-05,
      "loss": 0.10050593316555023,
      "num_input_tokens_seen": 1762612,
      "step": 936,
      "train_runtime": 2531.0299,
      "train_tokens_per_second": 696.401
    },
    {
      "epoch": 0.9471822087439979,
      "grad_norm": 2.1758036613464355,
      "learning_rate": 1.0572151898734178e-05,
      "loss": 0.10238852351903915,
      "num_input_tokens_seen": 1764110,
      "step": 937,
      "train_runtime": 2533.1454,
      "train_tokens_per_second": 696.411
    },
    {
      "epoch": 0.9481930755622947,
      "grad_norm": 1.4356797933578491,
      "learning_rate": 1.0562025316455697e-05,
      "loss": 0.07690879702568054,
      "num_input_tokens_seen": 1765750,
      "step": 938,
      "train_runtime": 2535.2657,
      "train_tokens_per_second": 696.475
    },
    {
      "epoch": 0.9492039423805914,
      "grad_norm": 1.4490258693695068,
      "learning_rate": 1.0551898734177216e-05,
      "loss": 0.0861949697136879,
      "num_input_tokens_seen": 1768042,
      "step": 939,
      "train_runtime": 2537.5346,
      "train_tokens_per_second": 696.756
    },
    {
      "epoch": 0.950214809198888,
      "grad_norm": 1.7231247425079346,
      "learning_rate": 1.0541772151898735e-05,
      "loss": 0.09950204193592072,
      "num_input_tokens_seen": 1769708,
      "step": 940,
      "train_runtime": 2539.653,
      "train_tokens_per_second": 696.831
    },
    {
      "epoch": 0.9512256760171848,
      "grad_norm": 1.4435205459594727,
      "learning_rate": 1.0531645569620254e-05,
      "loss": 0.07849620282649994,
      "num_input_tokens_seen": 1771470,
      "step": 941,
      "train_runtime": 2541.8041,
      "train_tokens_per_second": 696.934
    },
    {
      "epoch": 0.9522365428354814,
      "grad_norm": 1.6984418630599976,
      "learning_rate": 1.0521518987341773e-05,
      "loss": 0.11402212083339691,
      "num_input_tokens_seen": 1773556,
      "step": 942,
      "train_runtime": 2544.0973,
      "train_tokens_per_second": 697.126
    },
    {
      "epoch": 0.9532474096537781,
      "grad_norm": 1.6391994953155518,
      "learning_rate": 1.0511392405063292e-05,
      "loss": 0.10636426508426666,
      "num_input_tokens_seen": 1775544,
      "step": 943,
      "train_runtime": 2546.3665,
      "train_tokens_per_second": 697.285
    },
    {
      "epoch": 0.9542582764720748,
      "grad_norm": 1.7203863859176636,
      "learning_rate": 1.0501265822784811e-05,
      "loss": 0.09291623532772064,
      "num_input_tokens_seen": 1777788,
      "step": 944,
      "train_runtime": 2548.6243,
      "train_tokens_per_second": 697.548
    },
    {
      "epoch": 0.9552691432903715,
      "grad_norm": 2.124809980392456,
      "learning_rate": 1.049113924050633e-05,
      "loss": 0.125258207321167,
      "num_input_tokens_seen": 1779460,
      "step": 945,
      "train_runtime": 2550.7651,
      "train_tokens_per_second": 697.618
    },
    {
      "epoch": 0.9562800101086681,
      "grad_norm": 1.8077716827392578,
      "learning_rate": 1.048101265822785e-05,
      "loss": 0.11999374628067017,
      "num_input_tokens_seen": 1781276,
      "step": 946,
      "train_runtime": 2552.975,
      "train_tokens_per_second": 697.726
    },
    {
      "epoch": 0.9572908769269649,
      "grad_norm": 1.5265414714813232,
      "learning_rate": 1.0470886075949367e-05,
      "loss": 0.08729196339845657,
      "num_input_tokens_seen": 1783112,
      "step": 947,
      "train_runtime": 2555.1893,
      "train_tokens_per_second": 697.839
    },
    {
      "epoch": 0.9583017437452616,
      "grad_norm": 1.4505257606506348,
      "learning_rate": 1.0460759493670888e-05,
      "loss": 0.07770172506570816,
      "num_input_tokens_seen": 1784986,
      "step": 948,
      "train_runtime": 2557.4548,
      "train_tokens_per_second": 697.954
    },
    {
      "epoch": 0.9593126105635582,
      "grad_norm": 1.8700900077819824,
      "learning_rate": 1.0450632911392407e-05,
      "loss": 0.13766318559646606,
      "num_input_tokens_seen": 1787244,
      "step": 949,
      "train_runtime": 2559.7514,
      "train_tokens_per_second": 698.21
    },
    {
      "epoch": 0.960323477381855,
      "grad_norm": 1.7047700881958008,
      "learning_rate": 1.0440506329113924e-05,
      "loss": 0.09592680633068085,
      "num_input_tokens_seen": 1788888,
      "step": 950,
      "train_runtime": 2561.9143,
      "train_tokens_per_second": 698.262
    },
    {
      "epoch": 0.9613343442001516,
      "grad_norm": 2.1990907192230225,
      "learning_rate": 1.0430379746835443e-05,
      "loss": 0.09640653431415558,
      "num_input_tokens_seen": 1790766,
      "step": 951,
      "train_runtime": 2564.2404,
      "train_tokens_per_second": 698.361
    },
    {
      "epoch": 0.9623452110184483,
      "grad_norm": 1.5744006633758545,
      "learning_rate": 1.0420253164556964e-05,
      "loss": 0.09867499768733978,
      "num_input_tokens_seen": 1792828,
      "step": 952,
      "train_runtime": 2566.5341,
      "train_tokens_per_second": 698.54
    },
    {
      "epoch": 0.963356077836745,
      "grad_norm": 2.293447971343994,
      "learning_rate": 1.0410126582278481e-05,
      "loss": 0.10601398348808289,
      "num_input_tokens_seen": 1794592,
      "step": 953,
      "train_runtime": 2568.7134,
      "train_tokens_per_second": 698.635
    },
    {
      "epoch": 0.9643669446550417,
      "grad_norm": 1.9407451152801514,
      "learning_rate": 1.04e-05,
      "loss": 0.10499546676874161,
      "num_input_tokens_seen": 1796582,
      "step": 954,
      "train_runtime": 2570.8862,
      "train_tokens_per_second": 698.818
    },
    {
      "epoch": 0.9653778114733383,
      "grad_norm": 1.9761666059494019,
      "learning_rate": 1.038987341772152e-05,
      "loss": 0.10242647677659988,
      "num_input_tokens_seen": 1798762,
      "step": 955,
      "train_runtime": 2573.1185,
      "train_tokens_per_second": 699.059
    },
    {
      "epoch": 0.9663886782916351,
      "grad_norm": 2.1975674629211426,
      "learning_rate": 1.0379746835443039e-05,
      "loss": 0.10388092696666718,
      "num_input_tokens_seen": 1800326,
      "step": 956,
      "train_runtime": 2575.2121,
      "train_tokens_per_second": 699.098
    },
    {
      "epoch": 0.9673995451099318,
      "grad_norm": 1.1983399391174316,
      "learning_rate": 1.0369620253164558e-05,
      "loss": 0.06608706712722778,
      "num_input_tokens_seen": 1802300,
      "step": 957,
      "train_runtime": 2577.4185,
      "train_tokens_per_second": 699.266
    },
    {
      "epoch": 0.9684104119282284,
      "grad_norm": 1.9207825660705566,
      "learning_rate": 1.0359493670886077e-05,
      "loss": 0.11018344014883041,
      "num_input_tokens_seen": 1804308,
      "step": 958,
      "train_runtime": 2579.651,
      "train_tokens_per_second": 699.439
    },
    {
      "epoch": 0.9694212787465252,
      "grad_norm": 1.5126341581344604,
      "learning_rate": 1.0349367088607596e-05,
      "loss": 0.09055756777524948,
      "num_input_tokens_seen": 1806076,
      "step": 959,
      "train_runtime": 2581.8815,
      "train_tokens_per_second": 699.519
    },
    {
      "epoch": 0.9704321455648218,
      "grad_norm": 1.6882377862930298,
      "learning_rate": 1.0339240506329115e-05,
      "loss": 0.07870013266801834,
      "num_input_tokens_seen": 1807730,
      "step": 960,
      "train_runtime": 2584.0348,
      "train_tokens_per_second": 699.577
    },
    {
      "epoch": 0.9714430123831185,
      "grad_norm": 1.9891482591629028,
      "learning_rate": 1.0329113924050634e-05,
      "loss": 0.10766452550888062,
      "num_input_tokens_seen": 1809348,
      "step": 961,
      "train_runtime": 2593.2073,
      "train_tokens_per_second": 697.726
    },
    {
      "epoch": 0.9724538792014152,
      "grad_norm": 2.0645711421966553,
      "learning_rate": 1.0318987341772153e-05,
      "loss": 0.11320405453443527,
      "num_input_tokens_seen": 1811064,
      "step": 962,
      "train_runtime": 2595.4138,
      "train_tokens_per_second": 697.794
    },
    {
      "epoch": 0.9734647460197119,
      "grad_norm": 1.7952691316604614,
      "learning_rate": 1.030886075949367e-05,
      "loss": 0.1397959589958191,
      "num_input_tokens_seen": 1813668,
      "step": 963,
      "train_runtime": 2597.7892,
      "train_tokens_per_second": 698.158
    },
    {
      "epoch": 0.9744756128380085,
      "grad_norm": 1.5612289905548096,
      "learning_rate": 1.0298734177215191e-05,
      "loss": 0.08474574983119965,
      "num_input_tokens_seen": 1815248,
      "step": 964,
      "train_runtime": 2599.8867,
      "train_tokens_per_second": 698.203
    },
    {
      "epoch": 0.9754864796563053,
      "grad_norm": 1.7819595336914062,
      "learning_rate": 1.028860759493671e-05,
      "loss": 0.08624096214771271,
      "num_input_tokens_seen": 1816818,
      "step": 965,
      "train_runtime": 2602.011,
      "train_tokens_per_second": 698.236
    },
    {
      "epoch": 0.976497346474602,
      "grad_norm": 1.9061187505722046,
      "learning_rate": 1.0278481012658228e-05,
      "loss": 0.09254920482635498,
      "num_input_tokens_seen": 1818406,
      "step": 966,
      "train_runtime": 2604.1729,
      "train_tokens_per_second": 698.266
    },
    {
      "epoch": 0.9775082132928986,
      "grad_norm": 1.5857383012771606,
      "learning_rate": 1.0268354430379748e-05,
      "loss": 0.09439010918140411,
      "num_input_tokens_seen": 1820820,
      "step": 967,
      "train_runtime": 2606.4961,
      "train_tokens_per_second": 698.57
    },
    {
      "epoch": 0.9785190801111954,
      "grad_norm": 2.2102553844451904,
      "learning_rate": 1.0258227848101268e-05,
      "loss": 0.10842249542474747,
      "num_input_tokens_seen": 1822530,
      "step": 968,
      "train_runtime": 2608.6957,
      "train_tokens_per_second": 698.636
    },
    {
      "epoch": 0.979529946929492,
      "grad_norm": 1.7259248495101929,
      "learning_rate": 1.0248101265822785e-05,
      "loss": 0.07752262055873871,
      "num_input_tokens_seen": 1824318,
      "step": 969,
      "train_runtime": 2610.8761,
      "train_tokens_per_second": 698.738
    },
    {
      "epoch": 0.9805408137477887,
      "grad_norm": 1.5594508647918701,
      "learning_rate": 1.0237974683544306e-05,
      "loss": 0.08729353547096252,
      "num_input_tokens_seen": 1826020,
      "step": 970,
      "train_runtime": 2613.0343,
      "train_tokens_per_second": 698.812
    },
    {
      "epoch": 0.9815516805660854,
      "grad_norm": 1.7379978895187378,
      "learning_rate": 1.0227848101265823e-05,
      "loss": 0.11745233088731766,
      "num_input_tokens_seen": 1827850,
      "step": 971,
      "train_runtime": 2615.2119,
      "train_tokens_per_second": 698.93
    },
    {
      "epoch": 0.9825625473843821,
      "grad_norm": 1.4099138975143433,
      "learning_rate": 1.0217721518987342e-05,
      "loss": 0.091756671667099,
      "num_input_tokens_seen": 1829772,
      "step": 972,
      "train_runtime": 2617.4338,
      "train_tokens_per_second": 699.071
    },
    {
      "epoch": 0.9835734142026787,
      "grad_norm": 1.8185721635818481,
      "learning_rate": 1.0207594936708863e-05,
      "loss": 0.09723731875419617,
      "num_input_tokens_seen": 1831816,
      "step": 973,
      "train_runtime": 2619.7458,
      "train_tokens_per_second": 699.234
    },
    {
      "epoch": 0.9845842810209755,
      "grad_norm": 1.6812020540237427,
      "learning_rate": 1.019746835443038e-05,
      "loss": 0.10204876214265823,
      "num_input_tokens_seen": 1833678,
      "step": 974,
      "train_runtime": 2621.9666,
      "train_tokens_per_second": 699.352
    },
    {
      "epoch": 0.9855951478392722,
      "grad_norm": 1.5610170364379883,
      "learning_rate": 1.01873417721519e-05,
      "loss": 0.060523878782987595,
      "num_input_tokens_seen": 1835588,
      "step": 975,
      "train_runtime": 2624.1735,
      "train_tokens_per_second": 699.492
    },
    {
      "epoch": 0.9866060146575689,
      "grad_norm": 1.766232967376709,
      "learning_rate": 1.017721518987342e-05,
      "loss": 0.0874742791056633,
      "num_input_tokens_seen": 1837268,
      "step": 976,
      "train_runtime": 2626.3732,
      "train_tokens_per_second": 699.546
    },
    {
      "epoch": 0.9876168814758656,
      "grad_norm": 1.7839418649673462,
      "learning_rate": 1.0167088607594938e-05,
      "loss": 0.10308610647916794,
      "num_input_tokens_seen": 1839268,
      "step": 977,
      "train_runtime": 2628.5998,
      "train_tokens_per_second": 699.714
    },
    {
      "epoch": 0.9886277482941622,
      "grad_norm": 1.7231510877609253,
      "learning_rate": 1.0156962025316457e-05,
      "loss": 0.12159512937068939,
      "num_input_tokens_seen": 1841210,
      "step": 978,
      "train_runtime": 2630.8425,
      "train_tokens_per_second": 699.856
    },
    {
      "epoch": 0.989638615112459,
      "grad_norm": 2.347747325897217,
      "learning_rate": 1.0146835443037974e-05,
      "loss": 0.09375277906656265,
      "num_input_tokens_seen": 1842680,
      "step": 979,
      "train_runtime": 2632.9526,
      "train_tokens_per_second": 699.853
    },
    {
      "epoch": 0.9906494819307556,
      "grad_norm": 2.045182466506958,
      "learning_rate": 1.0136708860759495e-05,
      "loss": 0.12371169030666351,
      "num_input_tokens_seen": 1844666,
      "step": 980,
      "train_runtime": 2635.1939,
      "train_tokens_per_second": 700.011
    },
    {
      "epoch": 0.9916603487490523,
      "grad_norm": 1.5627561807632446,
      "learning_rate": 1.0126582278481014e-05,
      "loss": 0.0825665146112442,
      "num_input_tokens_seen": 1846488,
      "step": 981,
      "train_runtime": 2637.3771,
      "train_tokens_per_second": 700.123
    },
    {
      "epoch": 0.992671215567349,
      "grad_norm": 1.7864357233047485,
      "learning_rate": 1.0116455696202531e-05,
      "loss": 0.08284065872430801,
      "num_input_tokens_seen": 1847874,
      "step": 982,
      "train_runtime": 2639.5478,
      "train_tokens_per_second": 700.072
    },
    {
      "epoch": 0.9936820823856457,
      "grad_norm": 2.170645236968994,
      "learning_rate": 1.0106329113924052e-05,
      "loss": 0.13837072253227234,
      "num_input_tokens_seen": 1850558,
      "step": 983,
      "train_runtime": 2641.967,
      "train_tokens_per_second": 700.447
    },
    {
      "epoch": 0.9946929492039424,
      "grad_norm": 1.9412685632705688,
      "learning_rate": 1.0096202531645571e-05,
      "loss": 0.11869412660598755,
      "num_input_tokens_seen": 1852386,
      "step": 984,
      "train_runtime": 2644.2438,
      "train_tokens_per_second": 700.535
    },
    {
      "epoch": 0.9957038160222391,
      "grad_norm": 1.5198384523391724,
      "learning_rate": 1.0086075949367089e-05,
      "loss": 0.07557395100593567,
      "num_input_tokens_seen": 1853996,
      "step": 985,
      "train_runtime": 2646.4346,
      "train_tokens_per_second": 700.564
    },
    {
      "epoch": 0.9967146828405358,
      "grad_norm": 1.7301899194717407,
      "learning_rate": 1.007594936708861e-05,
      "loss": 0.09189169108867645,
      "num_input_tokens_seen": 1855850,
      "step": 986,
      "train_runtime": 2648.6551,
      "train_tokens_per_second": 700.676
    },
    {
      "epoch": 0.9977255496588324,
      "grad_norm": 1.65177583694458,
      "learning_rate": 1.0065822784810127e-05,
      "loss": 0.07830586284399033,
      "num_input_tokens_seen": 1857768,
      "step": 987,
      "train_runtime": 2650.8683,
      "train_tokens_per_second": 700.815
    },
    {
      "epoch": 0.9987364164771292,
      "grad_norm": 1.5787363052368164,
      "learning_rate": 1.0055696202531646e-05,
      "loss": 0.08452092856168747,
      "num_input_tokens_seen": 1859650,
      "step": 988,
      "train_runtime": 2653.0899,
      "train_tokens_per_second": 700.937
    },
    {
      "epoch": 0.9997472832954258,
      "grad_norm": 2.041066884994507,
      "learning_rate": 1.0045569620253167e-05,
      "loss": 0.1346156895160675,
      "num_input_tokens_seen": 1861366,
      "step": 989,
      "train_runtime": 2655.2932,
      "train_tokens_per_second": 701.002
    },
    {
      "epoch": 1.0,
      "grad_norm": 23.19478416442871,
      "learning_rate": 1.0035443037974684e-05,
      "loss": 0.34900957345962524,
      "num_input_tokens_seen": 1861552,
      "step": 990,
      "train_runtime": 2665.3394,
      "train_tokens_per_second": 698.43
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.3873527944087982,
      "eval_runtime": 52.9081,
      "eval_samples_per_second": 16.614,
      "eval_steps_per_second": 8.316,
      "num_input_tokens_seen": 1861552,
      "step": 990
    },
    {
      "epoch": 1.0010108668182967,
      "grad_norm": 1.6322622299194336,
      "learning_rate": 1.0025316455696203e-05,
      "loss": 0.07481811940670013,
      "num_input_tokens_seen": 1863014,
      "step": 991,
      "train_runtime": 2727.6173,
      "train_tokens_per_second": 683.019
    },
    {
      "epoch": 1.0020217336365933,
      "grad_norm": 1.3781479597091675,
      "learning_rate": 1.0015189873417724e-05,
      "loss": 0.1042320504784584,
      "num_input_tokens_seen": 1865072,
      "step": 992,
      "train_runtime": 2730.0183,
      "train_tokens_per_second": 683.172
    },
    {
      "epoch": 1.0030326004548902,
      "grad_norm": 1.300392508506775,
      "learning_rate": 1.0005063291139241e-05,
      "loss": 0.08801354467868805,
      "num_input_tokens_seen": 1867012,
      "step": 993,
      "train_runtime": 2732.239,
      "train_tokens_per_second": 683.327
    },
    {
      "epoch": 1.0040434672731868,
      "grad_norm": 1.5778790712356567,
      "learning_rate": 9.99493670886076e-06,
      "loss": 0.09628298133611679,
      "num_input_tokens_seen": 1868564,
      "step": 994,
      "train_runtime": 2734.3721,
      "train_tokens_per_second": 683.361
    },
    {
      "epoch": 1.0050543340914835,
      "grad_norm": 1.264314889907837,
      "learning_rate": 9.98481012658228e-06,
      "loss": 0.06317774206399918,
      "num_input_tokens_seen": 1870560,
      "step": 995,
      "train_runtime": 2736.5927,
      "train_tokens_per_second": 683.536
    },
    {
      "epoch": 1.0060652009097801,
      "grad_norm": 1.314989447593689,
      "learning_rate": 9.974683544303799e-06,
      "loss": 0.06444258987903595,
      "num_input_tokens_seen": 1872450,
      "step": 996,
      "train_runtime": 2738.8083,
      "train_tokens_per_second": 683.673
    },
    {
      "epoch": 1.0070760677280768,
      "grad_norm": 1.285747766494751,
      "learning_rate": 9.964556962025318e-06,
      "loss": 0.0854601040482521,
      "num_input_tokens_seen": 1874370,
      "step": 997,
      "train_runtime": 2741.0467,
      "train_tokens_per_second": 683.815
    },
    {
      "epoch": 1.0080869345463734,
      "grad_norm": 1.5861116647720337,
      "learning_rate": 9.954430379746837e-06,
      "loss": 0.09532982110977173,
      "num_input_tokens_seen": 1876216,
      "step": 998,
      "train_runtime": 2743.2362,
      "train_tokens_per_second": 683.943
    },
    {
      "epoch": 1.0090978013646703,
      "grad_norm": 1.652181625366211,
      "learning_rate": 9.944303797468356e-06,
      "loss": 0.05342147499322891,
      "num_input_tokens_seen": 1878082,
      "step": 999,
      "train_runtime": 2745.453,
      "train_tokens_per_second": 684.07
    },
    {
      "epoch": 1.010108668182967,
      "grad_norm": 1.6445740461349487,
      "learning_rate": 9.934177215189875e-06,
      "loss": 0.0853542909026146,
      "num_input_tokens_seen": 1879650,
      "step": 1000,
      "train_runtime": 2747.5812,
      "train_tokens_per_second": 684.111
    },
    {
      "epoch": 1.0111195350012636,
      "grad_norm": 1.2546181678771973,
      "learning_rate": 9.924050632911392e-06,
      "loss": 0.06549844145774841,
      "num_input_tokens_seen": 1881368,
      "step": 1001,
      "train_runtime": 2749.7809,
      "train_tokens_per_second": 684.188
    },
    {
      "epoch": 1.0121304018195603,
      "grad_norm": 1.8068426847457886,
      "learning_rate": 9.913924050632913e-06,
      "loss": 0.06808516383171082,
      "num_input_tokens_seen": 1882806,
      "step": 1002,
      "train_runtime": 2751.8812,
      "train_tokens_per_second": 684.189
    },
    {
      "epoch": 1.013141268637857,
      "grad_norm": 1.35905921459198,
      "learning_rate": 9.903797468354432e-06,
      "loss": 0.06314563751220703,
      "num_input_tokens_seen": 1884588,
      "step": 1003,
      "train_runtime": 2754.0677,
      "train_tokens_per_second": 684.293
    },
    {
      "epoch": 1.0141521354561536,
      "grad_norm": 1.603310227394104,
      "learning_rate": 9.89367088607595e-06,
      "loss": 0.08736512064933777,
      "num_input_tokens_seen": 1886554,
      "step": 1004,
      "train_runtime": 2756.2612,
      "train_tokens_per_second": 684.461
    },
    {
      "epoch": 1.0151630022744504,
      "grad_norm": 1.5805072784423828,
      "learning_rate": 9.883544303797469e-06,
      "loss": 0.0734487920999527,
      "num_input_tokens_seen": 1888610,
      "step": 1005,
      "train_runtime": 2758.5089,
      "train_tokens_per_second": 684.649
    },
    {
      "epoch": 1.016173869092747,
      "grad_norm": 1.6719231605529785,
      "learning_rate": 9.87341772151899e-06,
      "loss": 0.06566613167524338,
      "num_input_tokens_seen": 1890282,
      "step": 1006,
      "train_runtime": 2760.6988,
      "train_tokens_per_second": 684.711
    },
    {
      "epoch": 1.0171847359110437,
      "grad_norm": 1.2021580934524536,
      "learning_rate": 9.863291139240507e-06,
      "loss": 0.044181108474731445,
      "num_input_tokens_seen": 1891916,
      "step": 1007,
      "train_runtime": 2762.8328,
      "train_tokens_per_second": 684.774
    },
    {
      "epoch": 1.0181956027293404,
      "grad_norm": 1.5920672416687012,
      "learning_rate": 9.853164556962026e-06,
      "loss": 0.07045289129018784,
      "num_input_tokens_seen": 1893622,
      "step": 1008,
      "train_runtime": 2765.0175,
      "train_tokens_per_second": 684.85
    },
    {
      "epoch": 1.019206469547637,
      "grad_norm": 1.6282637119293213,
      "learning_rate": 9.843037974683545e-06,
      "loss": 0.07991328835487366,
      "num_input_tokens_seen": 1895592,
      "step": 1009,
      "train_runtime": 2767.2855,
      "train_tokens_per_second": 685.001
    },
    {
      "epoch": 1.0202173363659337,
      "grad_norm": 1.5714997053146362,
      "learning_rate": 9.832911392405064e-06,
      "loss": 0.06446757912635803,
      "num_input_tokens_seen": 1897352,
      "step": 1010,
      "train_runtime": 2769.4968,
      "train_tokens_per_second": 685.089
    },
    {
      "epoch": 1.0212282031842306,
      "grad_norm": 1.7102954387664795,
      "learning_rate": 9.822784810126583e-06,
      "loss": 0.07787447422742844,
      "num_input_tokens_seen": 1898760,
      "step": 1011,
      "train_runtime": 2771.6213,
      "train_tokens_per_second": 685.072
    },
    {
      "epoch": 1.0222390700025272,
      "grad_norm": 1.3697980642318726,
      "learning_rate": 9.812658227848102e-06,
      "loss": 0.06436234712600708,
      "num_input_tokens_seen": 1900986,
      "step": 1012,
      "train_runtime": 2773.9398,
      "train_tokens_per_second": 685.302
    },
    {
      "epoch": 1.0232499368208239,
      "grad_norm": 1.313659906387329,
      "learning_rate": 9.802531645569621e-06,
      "loss": 0.0565519705414772,
      "num_input_tokens_seen": 1902980,
      "step": 1013,
      "train_runtime": 2776.2302,
      "train_tokens_per_second": 685.455
    },
    {
      "epoch": 1.0242608036391205,
      "grad_norm": 1.8654744625091553,
      "learning_rate": 9.79240506329114e-06,
      "loss": 0.0803845077753067,
      "num_input_tokens_seen": 1904582,
      "step": 1014,
      "train_runtime": 2778.3688,
      "train_tokens_per_second": 685.504
    },
    {
      "epoch": 1.0252716704574172,
      "grad_norm": 1.4783668518066406,
      "learning_rate": 9.78227848101266e-06,
      "loss": 0.07052680850028992,
      "num_input_tokens_seen": 1906454,
      "step": 1015,
      "train_runtime": 2780.5654,
      "train_tokens_per_second": 685.635
    },
    {
      "epoch": 1.0262825372757138,
      "grad_norm": 1.5979278087615967,
      "learning_rate": 9.772151898734179e-06,
      "loss": 0.07049822062253952,
      "num_input_tokens_seen": 1908566,
      "step": 1016,
      "train_runtime": 2782.9208,
      "train_tokens_per_second": 685.814
    },
    {
      "epoch": 1.0272934040940107,
      "grad_norm": 1.6667790412902832,
      "learning_rate": 9.762025316455696e-06,
      "loss": 0.06343676149845123,
      "num_input_tokens_seen": 1910098,
      "step": 1017,
      "train_runtime": 2785.0856,
      "train_tokens_per_second": 685.831
    },
    {
      "epoch": 1.0283042709123074,
      "grad_norm": 2.0874030590057373,
      "learning_rate": 9.751898734177217e-06,
      "loss": 0.044850949198007584,
      "num_input_tokens_seen": 1911660,
      "step": 1018,
      "train_runtime": 2787.2563,
      "train_tokens_per_second": 685.857
    },
    {
      "epoch": 1.029315137730604,
      "grad_norm": 1.7865285873413086,
      "learning_rate": 9.741772151898736e-06,
      "loss": 0.1001095250248909,
      "num_input_tokens_seen": 1914106,
      "step": 1019,
      "train_runtime": 2789.6004,
      "train_tokens_per_second": 686.158
    },
    {
      "epoch": 1.0303260045489007,
      "grad_norm": 1.8012917041778564,
      "learning_rate": 9.731645569620253e-06,
      "loss": 0.06644564867019653,
      "num_input_tokens_seen": 1916250,
      "step": 1020,
      "train_runtime": 2791.8907,
      "train_tokens_per_second": 686.363
    },
    {
      "epoch": 1.0313368713671973,
      "grad_norm": 1.9052693843841553,
      "learning_rate": 9.721518987341772e-06,
      "loss": 0.05830027908086777,
      "num_input_tokens_seen": 1918076,
      "step": 1021,
      "train_runtime": 2801.4615,
      "train_tokens_per_second": 684.67
    },
    {
      "epoch": 1.032347738185494,
      "grad_norm": 1.31282377243042,
      "learning_rate": 9.711392405063293e-06,
      "loss": 0.06762159615755081,
      "num_input_tokens_seen": 1919904,
      "step": 1022,
      "train_runtime": 2803.7334,
      "train_tokens_per_second": 684.767
    },
    {
      "epoch": 1.0333586050037908,
      "grad_norm": 1.3395659923553467,
      "learning_rate": 9.70126582278481e-06,
      "loss": 0.059425726532936096,
      "num_input_tokens_seen": 1921690,
      "step": 1023,
      "train_runtime": 2805.9302,
      "train_tokens_per_second": 684.867
    },
    {
      "epoch": 1.0343694718220875,
      "grad_norm": 1.5222976207733154,
      "learning_rate": 9.69113924050633e-06,
      "loss": 0.07078788429498672,
      "num_input_tokens_seen": 1923734,
      "step": 1024,
      "train_runtime": 2808.1879,
      "train_tokens_per_second": 685.045
    },
    {
      "epoch": 1.0353803386403841,
      "grad_norm": 1.51506507396698,
      "learning_rate": 9.681012658227849e-06,
      "loss": 0.05457054823637009,
      "num_input_tokens_seen": 1925552,
      "step": 1025,
      "train_runtime": 2810.3814,
      "train_tokens_per_second": 685.157
    },
    {
      "epoch": 1.0363912054586808,
      "grad_norm": 1.5985889434814453,
      "learning_rate": 9.670886075949368e-06,
      "loss": 0.09114349633455276,
      "num_input_tokens_seen": 1927862,
      "step": 1026,
      "train_runtime": 2812.6927,
      "train_tokens_per_second": 685.415
    },
    {
      "epoch": 1.0374020722769774,
      "grad_norm": 1.4708911180496216,
      "learning_rate": 9.660759493670887e-06,
      "loss": 0.07649514824151993,
      "num_input_tokens_seen": 1930110,
      "step": 1027,
      "train_runtime": 2814.9914,
      "train_tokens_per_second": 685.654
    },
    {
      "epoch": 1.038412939095274,
      "grad_norm": 1.7000125646591187,
      "learning_rate": 9.650632911392406e-06,
      "loss": 0.08282623440027237,
      "num_input_tokens_seen": 1932032,
      "step": 1028,
      "train_runtime": 2817.2083,
      "train_tokens_per_second": 685.797
    },
    {
      "epoch": 1.039423805913571,
      "grad_norm": 1.353649616241455,
      "learning_rate": 9.640506329113925e-06,
      "loss": 0.0568917952477932,
      "num_input_tokens_seen": 1934146,
      "step": 1029,
      "train_runtime": 2819.4586,
      "train_tokens_per_second": 685.999
    },
    {
      "epoch": 1.0404346727318676,
      "grad_norm": 1.6784396171569824,
      "learning_rate": 9.630379746835444e-06,
      "loss": 0.0890979915857315,
      "num_input_tokens_seen": 1936018,
      "step": 1030,
      "train_runtime": 2821.6579,
      "train_tokens_per_second": 686.128
    },
    {
      "epoch": 1.0414455395501643,
      "grad_norm": 1.769243597984314,
      "learning_rate": 9.620253164556963e-06,
      "loss": 0.08021090924739838,
      "num_input_tokens_seen": 1937840,
      "step": 1031,
      "train_runtime": 2823.894,
      "train_tokens_per_second": 686.23
    },
    {
      "epoch": 1.042456406368461,
      "grad_norm": 1.3696779012680054,
      "learning_rate": 9.610126582278482e-06,
      "loss": 0.07308374345302582,
      "num_input_tokens_seen": 1939736,
      "step": 1032,
      "train_runtime": 2826.2732,
      "train_tokens_per_second": 686.323
    },
    {
      "epoch": 1.0434672731867576,
      "grad_norm": 1.5478060245513916,
      "learning_rate": 9.600000000000001e-06,
      "loss": 0.07152532041072845,
      "num_input_tokens_seen": 1941232,
      "step": 1033,
      "train_runtime": 2828.4997,
      "train_tokens_per_second": 686.312
    },
    {
      "epoch": 1.0444781400050542,
      "grad_norm": 1.4040976762771606,
      "learning_rate": 9.58987341772152e-06,
      "loss": 0.07309073954820633,
      "num_input_tokens_seen": 1943502,
      "step": 1034,
      "train_runtime": 2830.8529,
      "train_tokens_per_second": 686.543
    },
    {
      "epoch": 1.045489006823351,
      "grad_norm": 1.8501466512680054,
      "learning_rate": 9.57974683544304e-06,
      "loss": 0.09751475602388382,
      "num_input_tokens_seen": 1945588,
      "step": 1035,
      "train_runtime": 2833.18,
      "train_tokens_per_second": 686.715
    },
    {
      "epoch": 1.0464998736416478,
      "grad_norm": 1.5903476476669312,
      "learning_rate": 9.569620253164559e-06,
      "loss": 0.07216263562440872,
      "num_input_tokens_seen": 1947762,
      "step": 1036,
      "train_runtime": 2835.4648,
      "train_tokens_per_second": 686.929
    },
    {
      "epoch": 1.0475107404599444,
      "grad_norm": 1.4771184921264648,
      "learning_rate": 9.559493670886076e-06,
      "loss": 0.06384822726249695,
      "num_input_tokens_seen": 1949644,
      "step": 1037,
      "train_runtime": 2837.686,
      "train_tokens_per_second": 687.054
    },
    {
      "epoch": 1.048521607278241,
      "grad_norm": 1.484239935874939,
      "learning_rate": 9.549367088607597e-06,
      "loss": 0.0580734945833683,
      "num_input_tokens_seen": 1951800,
      "step": 1038,
      "train_runtime": 2839.9898,
      "train_tokens_per_second": 687.256
    },
    {
      "epoch": 1.0495324740965377,
      "grad_norm": 1.7439597845077515,
      "learning_rate": 9.539240506329114e-06,
      "loss": 0.08965416252613068,
      "num_input_tokens_seen": 1953690,
      "step": 1039,
      "train_runtime": 2842.2368,
      "train_tokens_per_second": 687.378
    },
    {
      "epoch": 1.0505433409148344,
      "grad_norm": 1.3571206331253052,
      "learning_rate": 9.529113924050633e-06,
      "loss": 0.06396070867776871,
      "num_input_tokens_seen": 1955684,
      "step": 1040,
      "train_runtime": 2844.5003,
      "train_tokens_per_second": 687.532
    },
    {
      "epoch": 1.0515542077331312,
      "grad_norm": 2.0411770343780518,
      "learning_rate": 9.518987341772152e-06,
      "loss": 0.09158805012702942,
      "num_input_tokens_seen": 1957194,
      "step": 1041,
      "train_runtime": 2846.6497,
      "train_tokens_per_second": 687.543
    },
    {
      "epoch": 1.052565074551428,
      "grad_norm": 1.2673745155334473,
      "learning_rate": 9.508860759493671e-06,
      "loss": 0.0589744858443737,
      "num_input_tokens_seen": 1959200,
      "step": 1042,
      "train_runtime": 2848.9349,
      "train_tokens_per_second": 687.696
    },
    {
      "epoch": 1.0535759413697245,
      "grad_norm": 1.4214316606521606,
      "learning_rate": 9.49873417721519e-06,
      "loss": 0.07307812571525574,
      "num_input_tokens_seen": 1960834,
      "step": 1043,
      "train_runtime": 2851.119,
      "train_tokens_per_second": 687.742
    },
    {
      "epoch": 1.0545868081880212,
      "grad_norm": 1.6550146341323853,
      "learning_rate": 9.48860759493671e-06,
      "loss": 0.07903876900672913,
      "num_input_tokens_seen": 1962374,
      "step": 1044,
      "train_runtime": 2853.294,
      "train_tokens_per_second": 687.757
    },
    {
      "epoch": 1.0555976750063178,
      "grad_norm": 2.3695287704467773,
      "learning_rate": 9.478481012658229e-06,
      "loss": 0.07892453670501709,
      "num_input_tokens_seen": 1964462,
      "step": 1045,
      "train_runtime": 2855.5391,
      "train_tokens_per_second": 687.948
    },
    {
      "epoch": 1.0566085418246145,
      "grad_norm": 1.7488689422607422,
      "learning_rate": 9.468354430379748e-06,
      "loss": 0.09015250205993652,
      "num_input_tokens_seen": 1966108,
      "step": 1046,
      "train_runtime": 2857.7028,
      "train_tokens_per_second": 688.003
    },
    {
      "epoch": 1.0576194086429114,
      "grad_norm": 1.7064481973648071,
      "learning_rate": 9.458227848101267e-06,
      "loss": 0.0758054330945015,
      "num_input_tokens_seen": 1968492,
      "step": 1047,
      "train_runtime": 2860.1156,
      "train_tokens_per_second": 688.256
    },
    {
      "epoch": 1.058630275461208,
      "grad_norm": 1.639689326286316,
      "learning_rate": 9.448101265822786e-06,
      "loss": 0.07452461123466492,
      "num_input_tokens_seen": 1970174,
      "step": 1048,
      "train_runtime": 2862.2955,
      "train_tokens_per_second": 688.32
    },
    {
      "epoch": 1.0596411422795047,
      "grad_norm": 1.6180789470672607,
      "learning_rate": 9.437974683544305e-06,
      "loss": 0.06205502152442932,
      "num_input_tokens_seen": 1972186,
      "step": 1049,
      "train_runtime": 2864.5241,
      "train_tokens_per_second": 688.486
    },
    {
      "epoch": 1.0606520090978013,
      "grad_norm": 1.6575864553451538,
      "learning_rate": 9.427848101265824e-06,
      "loss": 0.09482153505086899,
      "num_input_tokens_seen": 1974348,
      "step": 1050,
      "train_runtime": 2866.8507,
      "train_tokens_per_second": 688.682
    },
    {
      "epoch": 1.061662875916098,
      "grad_norm": 1.3664395809173584,
      "learning_rate": 9.417721518987343e-06,
      "loss": 0.038909636437892914,
      "num_input_tokens_seen": 1976044,
      "step": 1051,
      "train_runtime": 2876.283,
      "train_tokens_per_second": 687.013
    },
    {
      "epoch": 1.0626737427343946,
      "grad_norm": 1.5591120719909668,
      "learning_rate": 9.407594936708862e-06,
      "loss": 0.0615665502846241,
      "num_input_tokens_seen": 1977710,
      "step": 1052,
      "train_runtime": 2878.4783,
      "train_tokens_per_second": 687.068
    },
    {
      "epoch": 1.0636846095526915,
      "grad_norm": 1.8644870519638062,
      "learning_rate": 9.39746835443038e-06,
      "loss": 0.08972452580928802,
      "num_input_tokens_seen": 1979460,
      "step": 1053,
      "train_runtime": 2880.7661,
      "train_tokens_per_second": 687.13
    },
    {
      "epoch": 1.0646954763709882,
      "grad_norm": 2.219061851501465,
      "learning_rate": 9.3873417721519e-06,
      "loss": 0.07273946702480316,
      "num_input_tokens_seen": 1981128,
      "step": 1054,
      "train_runtime": 2882.9032,
      "train_tokens_per_second": 687.199
    },
    {
      "epoch": 1.0657063431892848,
      "grad_norm": 1.6209571361541748,
      "learning_rate": 9.37721518987342e-06,
      "loss": 0.09259951114654541,
      "num_input_tokens_seen": 1983712,
      "step": 1055,
      "train_runtime": 2885.3055,
      "train_tokens_per_second": 687.522
    },
    {
      "epoch": 1.0667172100075815,
      "grad_norm": 1.4539138078689575,
      "learning_rate": 9.367088607594937e-06,
      "loss": 0.062339551746845245,
      "num_input_tokens_seen": 1985418,
      "step": 1056,
      "train_runtime": 2887.5571,
      "train_tokens_per_second": 687.577
    },
    {
      "epoch": 1.0677280768258781,
      "grad_norm": 1.4814327955245972,
      "learning_rate": 9.356962025316456e-06,
      "loss": 0.07369458675384521,
      "num_input_tokens_seen": 1987526,
      "step": 1057,
      "train_runtime": 2889.9762,
      "train_tokens_per_second": 687.731
    },
    {
      "epoch": 1.068738943644175,
      "grad_norm": 1.6171040534973145,
      "learning_rate": 9.346835443037977e-06,
      "loss": 0.06056630611419678,
      "num_input_tokens_seen": 1989352,
      "step": 1058,
      "train_runtime": 2892.3176,
      "train_tokens_per_second": 687.806
    },
    {
      "epoch": 1.0697498104624716,
      "grad_norm": 1.803268551826477,
      "learning_rate": 9.336708860759494e-06,
      "loss": 0.0810079425573349,
      "num_input_tokens_seen": 1991022,
      "step": 1059,
      "train_runtime": 2894.5673,
      "train_tokens_per_second": 687.848
    },
    {
      "epoch": 1.0707606772807683,
      "grad_norm": 1.46821129322052,
      "learning_rate": 9.326582278481013e-06,
      "loss": 0.07994016259908676,
      "num_input_tokens_seen": 1992680,
      "step": 1060,
      "train_runtime": 2896.7319,
      "train_tokens_per_second": 687.906
    },
    {
      "epoch": 1.071771544099065,
      "grad_norm": 1.612409234046936,
      "learning_rate": 9.316455696202532e-06,
      "loss": 0.07581929862499237,
      "num_input_tokens_seen": 1995068,
      "step": 1061,
      "train_runtime": 2899.0779,
      "train_tokens_per_second": 688.173
    },
    {
      "epoch": 1.0727824109173616,
      "grad_norm": 1.5073354244232178,
      "learning_rate": 9.306329113924051e-06,
      "loss": 0.0688956081867218,
      "num_input_tokens_seen": 1996926,
      "step": 1062,
      "train_runtime": 2901.3747,
      "train_tokens_per_second": 688.269
    },
    {
      "epoch": 1.0737932777356582,
      "grad_norm": 1.8344887495040894,
      "learning_rate": 9.29620253164557e-06,
      "loss": 0.05343888700008392,
      "num_input_tokens_seen": 1998294,
      "step": 1063,
      "train_runtime": 2903.5353,
      "train_tokens_per_second": 688.228
    },
    {
      "epoch": 1.074804144553955,
      "grad_norm": 2.1024980545043945,
      "learning_rate": 9.28607594936709e-06,
      "loss": 0.0953790545463562,
      "num_input_tokens_seen": 2000332,
      "step": 1064,
      "train_runtime": 2905.8605,
      "train_tokens_per_second": 688.379
    },
    {
      "epoch": 1.0758150113722518,
      "grad_norm": 1.8216012716293335,
      "learning_rate": 9.275949367088609e-06,
      "loss": 0.07662083953619003,
      "num_input_tokens_seen": 2002028,
      "step": 1065,
      "train_runtime": 2907.9888,
      "train_tokens_per_second": 688.458
    },
    {
      "epoch": 1.0768258781905484,
      "grad_norm": 1.4832439422607422,
      "learning_rate": 9.265822784810128e-06,
      "loss": 0.06126931309700012,
      "num_input_tokens_seen": 2004148,
      "step": 1066,
      "train_runtime": 2910.2692,
      "train_tokens_per_second": 688.647
    },
    {
      "epoch": 1.077836745008845,
      "grad_norm": 1.6016371250152588,
      "learning_rate": 9.255696202531647e-06,
      "loss": 0.07705452293157578,
      "num_input_tokens_seen": 2006290,
      "step": 1067,
      "train_runtime": 2912.5295,
      "train_tokens_per_second": 688.848
    },
    {
      "epoch": 1.0788476118271417,
      "grad_norm": 1.9054425954818726,
      "learning_rate": 9.245569620253166e-06,
      "loss": 0.07443057000637054,
      "num_input_tokens_seen": 2008128,
      "step": 1068,
      "train_runtime": 2914.7102,
      "train_tokens_per_second": 688.963
    },
    {
      "epoch": 1.0798584786454384,
      "grad_norm": 1.7080419063568115,
      "learning_rate": 9.235443037974683e-06,
      "loss": 0.06730297952890396,
      "num_input_tokens_seen": 2010360,
      "step": 1069,
      "train_runtime": 2917.0117,
      "train_tokens_per_second": 689.185
    },
    {
      "epoch": 1.0808693454637353,
      "grad_norm": 1.4211933612823486,
      "learning_rate": 9.225316455696204e-06,
      "loss": 0.06265653669834137,
      "num_input_tokens_seen": 2012614,
      "step": 1070,
      "train_runtime": 2919.3097,
      "train_tokens_per_second": 689.414
    },
    {
      "epoch": 1.081880212282032,
      "grad_norm": 1.900034785270691,
      "learning_rate": 9.215189873417723e-06,
      "loss": 0.06608742475509644,
      "num_input_tokens_seen": 2014350,
      "step": 1071,
      "train_runtime": 2921.5228,
      "train_tokens_per_second": 689.486
    },
    {
      "epoch": 1.0828910791003286,
      "grad_norm": 1.803638219833374,
      "learning_rate": 9.20506329113924e-06,
      "loss": 0.0830145925283432,
      "num_input_tokens_seen": 2016280,
      "step": 1072,
      "train_runtime": 2923.7565,
      "train_tokens_per_second": 689.62
    },
    {
      "epoch": 1.0839019459186252,
      "grad_norm": 1.8618730306625366,
      "learning_rate": 9.19493670886076e-06,
      "loss": 0.08237496018409729,
      "num_input_tokens_seen": 2018160,
      "step": 1073,
      "train_runtime": 2925.9549,
      "train_tokens_per_second": 689.744
    },
    {
      "epoch": 1.0849128127369219,
      "grad_norm": 1.6639076471328735,
      "learning_rate": 9.18481012658228e-06,
      "loss": 0.07322726398706436,
      "num_input_tokens_seen": 2020416,
      "step": 1074,
      "train_runtime": 2928.3337,
      "train_tokens_per_second": 689.954
    },
    {
      "epoch": 1.0859236795552185,
      "grad_norm": 1.3883980512619019,
      "learning_rate": 9.174683544303798e-06,
      "loss": 0.07084261626005173,
      "num_input_tokens_seen": 2022348,
      "step": 1075,
      "train_runtime": 2930.5483,
      "train_tokens_per_second": 690.092
    },
    {
      "epoch": 1.0869345463735154,
      "grad_norm": 1.5835883617401123,
      "learning_rate": 9.164556962025317e-06,
      "loss": 0.06844441592693329,
      "num_input_tokens_seen": 2024030,
      "step": 1076,
      "train_runtime": 2932.7211,
      "train_tokens_per_second": 690.154
    },
    {
      "epoch": 1.087945413191812,
      "grad_norm": 1.8336513042449951,
      "learning_rate": 9.154430379746836e-06,
      "loss": 0.08951697498559952,
      "num_input_tokens_seen": 2025526,
      "step": 1077,
      "train_runtime": 2934.8198,
      "train_tokens_per_second": 690.17
    },
    {
      "epoch": 1.0889562800101087,
      "grad_norm": 1.4816232919692993,
      "learning_rate": 9.144303797468355e-06,
      "loss": 0.05875733122229576,
      "num_input_tokens_seen": 2027310,
      "step": 1078,
      "train_runtime": 2937.003,
      "train_tokens_per_second": 690.265
    },
    {
      "epoch": 1.0899671468284053,
      "grad_norm": 1.3453950881958008,
      "learning_rate": 9.134177215189874e-06,
      "loss": 0.057678379118442535,
      "num_input_tokens_seen": 2029130,
      "step": 1079,
      "train_runtime": 2939.2305,
      "train_tokens_per_second": 690.361
    },
    {
      "epoch": 1.090978013646702,
      "grad_norm": 1.2752723693847656,
      "learning_rate": 9.124050632911393e-06,
      "loss": 0.06412097811698914,
      "num_input_tokens_seen": 2031392,
      "step": 1080,
      "train_runtime": 2941.5368,
      "train_tokens_per_second": 690.589
    },
    {
      "epoch": 1.0919888804649986,
      "grad_norm": 1.6641526222229004,
      "learning_rate": 9.113924050632912e-06,
      "loss": 0.08673028647899628,
      "num_input_tokens_seen": 2033048,
      "step": 1081,
      "train_runtime": 2950.9052,
      "train_tokens_per_second": 688.957
    },
    {
      "epoch": 1.0929997472832955,
      "grad_norm": 1.4624192714691162,
      "learning_rate": 9.103797468354431e-06,
      "loss": 0.052173640578985214,
      "num_input_tokens_seen": 2034860,
      "step": 1082,
      "train_runtime": 2953.1111,
      "train_tokens_per_second": 689.056
    },
    {
      "epoch": 1.0940106141015922,
      "grad_norm": 1.6662530899047852,
      "learning_rate": 9.09367088607595e-06,
      "loss": 0.0618804469704628,
      "num_input_tokens_seen": 2036578,
      "step": 1083,
      "train_runtime": 2955.334,
      "train_tokens_per_second": 689.119
    },
    {
      "epoch": 1.0950214809198888,
      "grad_norm": 1.7030669450759888,
      "learning_rate": 9.08354430379747e-06,
      "loss": 0.06165473163127899,
      "num_input_tokens_seen": 2038364,
      "step": 1084,
      "train_runtime": 2957.5348,
      "train_tokens_per_second": 689.21
    },
    {
      "epoch": 1.0960323477381855,
      "grad_norm": 1.3679355382919312,
      "learning_rate": 9.073417721518989e-06,
      "loss": 0.03944582864642143,
      "num_input_tokens_seen": 2040330,
      "step": 1085,
      "train_runtime": 2959.7266,
      "train_tokens_per_second": 689.364
    },
    {
      "epoch": 1.0970432145564821,
      "grad_norm": 1.2052851915359497,
      "learning_rate": 9.063291139240506e-06,
      "loss": 0.04675544425845146,
      "num_input_tokens_seen": 2042242,
      "step": 1086,
      "train_runtime": 2961.9984,
      "train_tokens_per_second": 689.481
    },
    {
      "epoch": 1.0980540813747788,
      "grad_norm": 1.8082427978515625,
      "learning_rate": 9.053164556962027e-06,
      "loss": 0.08575178682804108,
      "num_input_tokens_seen": 2044062,
      "step": 1087,
      "train_runtime": 2964.2939,
      "train_tokens_per_second": 689.561
    },
    {
      "epoch": 1.0990649481930757,
      "grad_norm": 1.5417779684066772,
      "learning_rate": 9.043037974683546e-06,
      "loss": 0.06239135563373566,
      "num_input_tokens_seen": 2045578,
      "step": 1088,
      "train_runtime": 2966.457,
      "train_tokens_per_second": 689.569
    },
    {
      "epoch": 1.1000758150113723,
      "grad_norm": 1.6572215557098389,
      "learning_rate": 9.032911392405063e-06,
      "loss": 0.07030152529478073,
      "num_input_tokens_seen": 2047120,
      "step": 1089,
      "train_runtime": 2968.6522,
      "train_tokens_per_second": 689.579
    },
    {
      "epoch": 1.101086681829669,
      "grad_norm": 2.1292786598205566,
      "learning_rate": 9.022784810126582e-06,
      "loss": 0.08384208381175995,
      "num_input_tokens_seen": 2048692,
      "step": 1090,
      "train_runtime": 2970.8242,
      "train_tokens_per_second": 689.604
    },
    {
      "epoch": 1.1020975486479656,
      "grad_norm": 1.7668535709381104,
      "learning_rate": 9.012658227848103e-06,
      "loss": 0.0798768550157547,
      "num_input_tokens_seen": 2050836,
      "step": 1091,
      "train_runtime": 2973.1606,
      "train_tokens_per_second": 689.783
    },
    {
      "epoch": 1.1031084154662623,
      "grad_norm": 1.7629799842834473,
      "learning_rate": 9.00253164556962e-06,
      "loss": 0.09637036919593811,
      "num_input_tokens_seen": 2052972,
      "step": 1092,
      "train_runtime": 2975.4611,
      "train_tokens_per_second": 689.968
    },
    {
      "epoch": 1.104119282284559,
      "grad_norm": 1.5138276815414429,
      "learning_rate": 8.99240506329114e-06,
      "loss": 0.06796854734420776,
      "num_input_tokens_seen": 2055178,
      "step": 1093,
      "train_runtime": 2977.8406,
      "train_tokens_per_second": 690.157
    },
    {
      "epoch": 1.1051301491028558,
      "grad_norm": 1.6930639743804932,
      "learning_rate": 8.982278481012659e-06,
      "loss": 0.07748739421367645,
      "num_input_tokens_seen": 2056806,
      "step": 1094,
      "train_runtime": 2980.0242,
      "train_tokens_per_second": 690.198
    },
    {
      "epoch": 1.1061410159211524,
      "grad_norm": 1.4320027828216553,
      "learning_rate": 8.972151898734178e-06,
      "loss": 0.06378726661205292,
      "num_input_tokens_seen": 2059252,
      "step": 1095,
      "train_runtime": 2982.3803,
      "train_tokens_per_second": 690.473
    },
    {
      "epoch": 1.107151882739449,
      "grad_norm": 1.3486897945404053,
      "learning_rate": 8.962025316455697e-06,
      "loss": 0.04433378204703331,
      "num_input_tokens_seen": 2061036,
      "step": 1096,
      "train_runtime": 2984.5739,
      "train_tokens_per_second": 690.563
    },
    {
      "epoch": 1.1081627495577457,
      "grad_norm": 1.6590334177017212,
      "learning_rate": 8.951898734177216e-06,
      "loss": 0.06255605816841125,
      "num_input_tokens_seen": 2063098,
      "step": 1097,
      "train_runtime": 2986.846,
      "train_tokens_per_second": 690.728
    },
    {
      "epoch": 1.1091736163760424,
      "grad_norm": 1.7285943031311035,
      "learning_rate": 8.941772151898735e-06,
      "loss": 0.08358632773160934,
      "num_input_tokens_seen": 2065018,
      "step": 1098,
      "train_runtime": 2989.0721,
      "train_tokens_per_second": 690.856
    },
    {
      "epoch": 1.110184483194339,
      "grad_norm": 1.3904027938842773,
      "learning_rate": 8.931645569620254e-06,
      "loss": 0.0616244338452816,
      "num_input_tokens_seen": 2067042,
      "step": 1099,
      "train_runtime": 2991.2846,
      "train_tokens_per_second": 691.022
    },
    {
      "epoch": 1.111195350012636,
      "grad_norm": 1.6543943881988525,
      "learning_rate": 8.921518987341773e-06,
      "loss": 0.06925375759601593,
      "num_input_tokens_seen": 2068762,
      "step": 1100,
      "train_runtime": 2993.4889,
      "train_tokens_per_second": 691.087
    },
    {
      "epoch": 1.1122062168309326,
      "grad_norm": 1.649613618850708,
      "learning_rate": 8.911392405063292e-06,
      "loss": 0.07829856872558594,
      "num_input_tokens_seen": 2070476,
      "step": 1101,
      "train_runtime": 2995.6911,
      "train_tokens_per_second": 691.151
    },
    {
      "epoch": 1.1132170836492292,
      "grad_norm": 1.4735194444656372,
      "learning_rate": 8.90126582278481e-06,
      "loss": 0.06881175190210342,
      "num_input_tokens_seen": 2072284,
      "step": 1102,
      "train_runtime": 2997.8727,
      "train_tokens_per_second": 691.252
    },
    {
      "epoch": 1.1142279504675259,
      "grad_norm": 1.8341642618179321,
      "learning_rate": 8.89113924050633e-06,
      "loss": 0.0606992207467556,
      "num_input_tokens_seen": 2074296,
      "step": 1103,
      "train_runtime": 3000.1202,
      "train_tokens_per_second": 691.404
    },
    {
      "epoch": 1.1152388172858225,
      "grad_norm": 1.4089767932891846,
      "learning_rate": 8.88101265822785e-06,
      "loss": 0.053868427872657776,
      "num_input_tokens_seen": 2076596,
      "step": 1104,
      "train_runtime": 3002.4244,
      "train_tokens_per_second": 691.64
    },
    {
      "epoch": 1.1162496841041194,
      "grad_norm": 1.8250232934951782,
      "learning_rate": 8.870886075949367e-06,
      "loss": 0.06492453813552856,
      "num_input_tokens_seen": 2078276,
      "step": 1105,
      "train_runtime": 3004.5811,
      "train_tokens_per_second": 691.702
    },
    {
      "epoch": 1.117260550922416,
      "grad_norm": 1.3938127756118774,
      "learning_rate": 8.860759493670886e-06,
      "loss": 0.07209944725036621,
      "num_input_tokens_seen": 2080060,
      "step": 1106,
      "train_runtime": 3006.7729,
      "train_tokens_per_second": 691.792
    },
    {
      "epoch": 1.1182714177407127,
      "grad_norm": 1.645253300666809,
      "learning_rate": 8.850632911392407e-06,
      "loss": 0.08678440749645233,
      "num_input_tokens_seen": 2082446,
      "step": 1107,
      "train_runtime": 3009.074,
      "train_tokens_per_second": 692.055
    },
    {
      "epoch": 1.1192822845590094,
      "grad_norm": 1.487198829650879,
      "learning_rate": 8.840506329113924e-06,
      "loss": 0.06331930309534073,
      "num_input_tokens_seen": 2083832,
      "step": 1108,
      "train_runtime": 3011.2099,
      "train_tokens_per_second": 692.025
    },
    {
      "epoch": 1.120293151377306,
      "grad_norm": 1.535788655281067,
      "learning_rate": 8.830379746835443e-06,
      "loss": 0.08061740547418594,
      "num_input_tokens_seen": 2086060,
      "step": 1109,
      "train_runtime": 3013.493,
      "train_tokens_per_second": 692.24
    },
    {
      "epoch": 1.1213040181956027,
      "grad_norm": 1.3998849391937256,
      "learning_rate": 8.820253164556962e-06,
      "loss": 0.04212111234664917,
      "num_input_tokens_seen": 2087998,
      "step": 1110,
      "train_runtime": 3015.7254,
      "train_tokens_per_second": 692.37
    },
    {
      "epoch": 1.1223148850138993,
      "grad_norm": 1.6482014656066895,
      "learning_rate": 8.810126582278481e-06,
      "loss": 0.06838325411081314,
      "num_input_tokens_seen": 2090364,
      "step": 1111,
      "train_runtime": 3024.8933,
      "train_tokens_per_second": 691.054
    },
    {
      "epoch": 1.1233257518321962,
      "grad_norm": 1.3028661012649536,
      "learning_rate": 8.8e-06,
      "loss": 0.045957691967487335,
      "num_input_tokens_seen": 2092166,
      "step": 1112,
      "train_runtime": 3027.1036,
      "train_tokens_per_second": 691.144
    },
    {
      "epoch": 1.1243366186504928,
      "grad_norm": 1.5552278757095337,
      "learning_rate": 8.78987341772152e-06,
      "loss": 0.05935283750295639,
      "num_input_tokens_seen": 2094464,
      "step": 1113,
      "train_runtime": 3029.3872,
      "train_tokens_per_second": 691.382
    },
    {
      "epoch": 1.1253474854687895,
      "grad_norm": 2.1640241146087646,
      "learning_rate": 8.779746835443039e-06,
      "loss": 0.13538388907909393,
      "num_input_tokens_seen": 2096704,
      "step": 1114,
      "train_runtime": 3031.6688,
      "train_tokens_per_second": 691.601
    },
    {
      "epoch": 1.1263583522870861,
      "grad_norm": 1.8794739246368408,
      "learning_rate": 8.769620253164558e-06,
      "loss": 0.08418547362089157,
      "num_input_tokens_seen": 2098626,
      "step": 1115,
      "train_runtime": 3033.9238,
      "train_tokens_per_second": 691.72
    },
    {
      "epoch": 1.1273692191053828,
      "grad_norm": 1.6963447332382202,
      "learning_rate": 8.759493670886077e-06,
      "loss": 0.07459042966365814,
      "num_input_tokens_seen": 2100258,
      "step": 1116,
      "train_runtime": 3036.0944,
      "train_tokens_per_second": 691.763
    },
    {
      "epoch": 1.1283800859236797,
      "grad_norm": 1.5021796226501465,
      "learning_rate": 8.749367088607596e-06,
      "loss": 0.05739009380340576,
      "num_input_tokens_seen": 2102152,
      "step": 1117,
      "train_runtime": 3038.3378,
      "train_tokens_per_second": 691.876
    },
    {
      "epoch": 1.1293909527419763,
      "grad_norm": 1.7058435678482056,
      "learning_rate": 8.739240506329115e-06,
      "loss": 0.10165772587060928,
      "num_input_tokens_seen": 2104124,
      "step": 1118,
      "train_runtime": 3040.5956,
      "train_tokens_per_second": 692.01
    },
    {
      "epoch": 1.130401819560273,
      "grad_norm": 1.438440203666687,
      "learning_rate": 8.729113924050634e-06,
      "loss": 0.08349044620990753,
      "num_input_tokens_seen": 2106328,
      "step": 1119,
      "train_runtime": 3042.8964,
      "train_tokens_per_second": 692.212
    },
    {
      "epoch": 1.1314126863785696,
      "grad_norm": 1.3999005556106567,
      "learning_rate": 8.718987341772153e-06,
      "loss": 0.06760846078395844,
      "num_input_tokens_seen": 2108402,
      "step": 1120,
      "train_runtime": 3045.1684,
      "train_tokens_per_second": 692.376
    },
    {
      "epoch": 1.1324235531968663,
      "grad_norm": 1.9466845989227295,
      "learning_rate": 8.708860759493672e-06,
      "loss": 0.06737107783555984,
      "num_input_tokens_seen": 2110132,
      "step": 1121,
      "train_runtime": 3047.4391,
      "train_tokens_per_second": 692.428
    },
    {
      "epoch": 1.133434420015163,
      "grad_norm": 1.329634189605713,
      "learning_rate": 8.69873417721519e-06,
      "loss": 0.06647045165300369,
      "num_input_tokens_seen": 2112090,
      "step": 1122,
      "train_runtime": 3049.7361,
      "train_tokens_per_second": 692.548
    },
    {
      "epoch": 1.1344452868334596,
      "grad_norm": 1.5675692558288574,
      "learning_rate": 8.68860759493671e-06,
      "loss": 0.06497927010059357,
      "num_input_tokens_seen": 2113744,
      "step": 1123,
      "train_runtime": 3051.8992,
      "train_tokens_per_second": 692.6
    },
    {
      "epoch": 1.1354561536517565,
      "grad_norm": 1.6912192106246948,
      "learning_rate": 8.67848101265823e-06,
      "loss": 0.08475741744041443,
      "num_input_tokens_seen": 2115786,
      "step": 1124,
      "train_runtime": 3054.1829,
      "train_tokens_per_second": 692.75
    },
    {
      "epoch": 1.136467020470053,
      "grad_norm": 1.5909514427185059,
      "learning_rate": 8.668354430379747e-06,
      "loss": 0.06388218700885773,
      "num_input_tokens_seen": 2117714,
      "step": 1125,
      "train_runtime": 3056.4061,
      "train_tokens_per_second": 692.877
    },
    {
      "epoch": 1.1374778872883498,
      "grad_norm": 2.9671549797058105,
      "learning_rate": 8.658227848101266e-06,
      "loss": 0.0821935385465622,
      "num_input_tokens_seen": 2119428,
      "step": 1126,
      "train_runtime": 3058.562,
      "train_tokens_per_second": 692.949
    },
    {
      "epoch": 1.1384887541066464,
      "grad_norm": 1.879675030708313,
      "learning_rate": 8.648101265822785e-06,
      "loss": 0.08488821983337402,
      "num_input_tokens_seen": 2121072,
      "step": 1127,
      "train_runtime": 3060.7569,
      "train_tokens_per_second": 692.989
    },
    {
      "epoch": 1.139499620924943,
      "grad_norm": 1.5213686227798462,
      "learning_rate": 8.637974683544304e-06,
      "loss": 0.08338913321495056,
      "num_input_tokens_seen": 2123588,
      "step": 1128,
      "train_runtime": 3063.1233,
      "train_tokens_per_second": 693.275
    },
    {
      "epoch": 1.14051048774324,
      "grad_norm": 1.3064101934432983,
      "learning_rate": 8.627848101265823e-06,
      "loss": 0.06499403715133667,
      "num_input_tokens_seen": 2125496,
      "step": 1129,
      "train_runtime": 3065.3746,
      "train_tokens_per_second": 693.389
    },
    {
      "epoch": 1.1415213545615366,
      "grad_norm": 1.5557777881622314,
      "learning_rate": 8.617721518987342e-06,
      "loss": 0.0621199756860733,
      "num_input_tokens_seen": 2127294,
      "step": 1130,
      "train_runtime": 3067.5581,
      "train_tokens_per_second": 693.481
    },
    {
      "epoch": 1.1425322213798332,
      "grad_norm": 1.6157169342041016,
      "learning_rate": 8.607594936708861e-06,
      "loss": 0.09719564765691757,
      "num_input_tokens_seen": 2130054,
      "step": 1131,
      "train_runtime": 3070.1017,
      "train_tokens_per_second": 693.806
    },
    {
      "epoch": 1.1435430881981299,
      "grad_norm": 1.687367558479309,
      "learning_rate": 8.59746835443038e-06,
      "loss": 0.08943184465169907,
      "num_input_tokens_seen": 2131806,
      "step": 1132,
      "train_runtime": 3072.2596,
      "train_tokens_per_second": 693.889
    },
    {
      "epoch": 1.1445539550164265,
      "grad_norm": 1.8464778661727905,
      "learning_rate": 8.5873417721519e-06,
      "loss": 0.06853809952735901,
      "num_input_tokens_seen": 2133332,
      "step": 1133,
      "train_runtime": 3074.4136,
      "train_tokens_per_second": 693.899
    },
    {
      "epoch": 1.1455648218347232,
      "grad_norm": 1.7622979879379272,
      "learning_rate": 8.577215189873419e-06,
      "loss": 0.07444445788860321,
      "num_input_tokens_seen": 2134852,
      "step": 1134,
      "train_runtime": 3076.6215,
      "train_tokens_per_second": 693.895
    },
    {
      "epoch": 1.1465756886530198,
      "grad_norm": 1.74332594871521,
      "learning_rate": 8.567088607594938e-06,
      "loss": 0.07678605616092682,
      "num_input_tokens_seen": 2136526,
      "step": 1135,
      "train_runtime": 3078.7916,
      "train_tokens_per_second": 693.95
    },
    {
      "epoch": 1.1475865554713167,
      "grad_norm": 1.6526131629943848,
      "learning_rate": 8.556962025316457e-06,
      "loss": 0.07518184930086136,
      "num_input_tokens_seen": 2138368,
      "step": 1136,
      "train_runtime": 3080.9896,
      "train_tokens_per_second": 694.052
    },
    {
      "epoch": 1.1485974222896134,
      "grad_norm": 1.5729053020477295,
      "learning_rate": 8.546835443037976e-06,
      "loss": 0.08650077879428864,
      "num_input_tokens_seen": 2140978,
      "step": 1137,
      "train_runtime": 3083.3943,
      "train_tokens_per_second": 694.358
    },
    {
      "epoch": 1.14960828910791,
      "grad_norm": 1.495152473449707,
      "learning_rate": 8.536708860759493e-06,
      "loss": 0.06230098009109497,
      "num_input_tokens_seen": 2143190,
      "step": 1138,
      "train_runtime": 3085.7032,
      "train_tokens_per_second": 694.555
    },
    {
      "epoch": 1.1506191559262067,
      "grad_norm": 1.5127619504928589,
      "learning_rate": 8.526582278481014e-06,
      "loss": 0.055189043283462524,
      "num_input_tokens_seen": 2144560,
      "step": 1139,
      "train_runtime": 3087.7746,
      "train_tokens_per_second": 694.533
    },
    {
      "epoch": 1.1516300227445033,
      "grad_norm": 1.6251533031463623,
      "learning_rate": 8.516455696202533e-06,
      "loss": 0.06701746582984924,
      "num_input_tokens_seen": 2146132,
      "step": 1140,
      "train_runtime": 3089.9262,
      "train_tokens_per_second": 694.558
    },
    {
      "epoch": 1.1526408895628002,
      "grad_norm": 1.6752350330352783,
      "learning_rate": 8.50632911392405e-06,
      "loss": 0.07148720324039459,
      "num_input_tokens_seen": 2148060,
      "step": 1141,
      "train_runtime": 3098.6284,
      "train_tokens_per_second": 693.229
    },
    {
      "epoch": 1.1536517563810968,
      "grad_norm": 2.347093105316162,
      "learning_rate": 8.49620253164557e-06,
      "loss": 0.1034323126077652,
      "num_input_tokens_seen": 2149560,
      "step": 1142,
      "train_runtime": 3100.7289,
      "train_tokens_per_second": 693.243
    },
    {
      "epoch": 1.1546626231993935,
      "grad_norm": 1.5739468336105347,
      "learning_rate": 8.48607594936709e-06,
      "loss": 0.0766829252243042,
      "num_input_tokens_seen": 2151656,
      "step": 1143,
      "train_runtime": 3102.9973,
      "train_tokens_per_second": 693.412
    },
    {
      "epoch": 1.1556734900176902,
      "grad_norm": 1.3017951250076294,
      "learning_rate": 8.475949367088608e-06,
      "loss": 0.06286372244358063,
      "num_input_tokens_seen": 2153778,
      "step": 1144,
      "train_runtime": 3105.2836,
      "train_tokens_per_second": 693.585
    },
    {
      "epoch": 1.1566843568359868,
      "grad_norm": 1.3676447868347168,
      "learning_rate": 8.465822784810127e-06,
      "loss": 0.059599995613098145,
      "num_input_tokens_seen": 2156104,
      "step": 1145,
      "train_runtime": 3107.5775,
      "train_tokens_per_second": 693.821
    },
    {
      "epoch": 1.1576952236542835,
      "grad_norm": 1.4337157011032104,
      "learning_rate": 8.455696202531646e-06,
      "loss": 0.06387347728013992,
      "num_input_tokens_seen": 2158052,
      "step": 1146,
      "train_runtime": 3109.803,
      "train_tokens_per_second": 693.951
    },
    {
      "epoch": 1.1587060904725803,
      "grad_norm": 1.5866986513137817,
      "learning_rate": 8.445569620253165e-06,
      "loss": 0.07430361956357956,
      "num_input_tokens_seen": 2159532,
      "step": 1147,
      "train_runtime": 3111.914,
      "train_tokens_per_second": 693.956
    },
    {
      "epoch": 1.159716957290877,
      "grad_norm": 1.9716557264328003,
      "learning_rate": 8.435443037974684e-06,
      "loss": 0.10540448874235153,
      "num_input_tokens_seen": 2161148,
      "step": 1148,
      "train_runtime": 3114.0679,
      "train_tokens_per_second": 693.995
    },
    {
      "epoch": 1.1607278241091736,
      "grad_norm": 1.620793104171753,
      "learning_rate": 8.425316455696203e-06,
      "loss": 0.07015478610992432,
      "num_input_tokens_seen": 2163152,
      "step": 1149,
      "train_runtime": 3116.3359,
      "train_tokens_per_second": 694.133
    },
    {
      "epoch": 1.1617386909274703,
      "grad_norm": 1.8051007986068726,
      "learning_rate": 8.415189873417722e-06,
      "loss": 0.06369549036026001,
      "num_input_tokens_seen": 2165346,
      "step": 1150,
      "train_runtime": 3118.6247,
      "train_tokens_per_second": 694.327
    },
    {
      "epoch": 1.162749557745767,
      "grad_norm": 1.872153878211975,
      "learning_rate": 8.405063291139241e-06,
      "loss": 0.08213125169277191,
      "num_input_tokens_seen": 2167748,
      "step": 1151,
      "train_runtime": 3120.9913,
      "train_tokens_per_second": 694.57
    },
    {
      "epoch": 1.1637604245640636,
      "grad_norm": 1.3431483507156372,
      "learning_rate": 8.39493670886076e-06,
      "loss": 0.0798673927783966,
      "num_input_tokens_seen": 2169900,
      "step": 1152,
      "train_runtime": 3123.2422,
      "train_tokens_per_second": 694.759
    },
    {
      "epoch": 1.1647712913823605,
      "grad_norm": 1.8245978355407715,
      "learning_rate": 8.38481012658228e-06,
      "loss": 0.08059132844209671,
      "num_input_tokens_seen": 2171650,
      "step": 1153,
      "train_runtime": 3125.4608,
      "train_tokens_per_second": 694.826
    },
    {
      "epoch": 1.1657821582006571,
      "grad_norm": 1.4516791105270386,
      "learning_rate": 8.374683544303797e-06,
      "loss": 0.0558345727622509,
      "num_input_tokens_seen": 2173872,
      "step": 1154,
      "train_runtime": 3127.7393,
      "train_tokens_per_second": 695.03
    },
    {
      "epoch": 1.1667930250189538,
      "grad_norm": 1.4005495309829712,
      "learning_rate": 8.364556962025318e-06,
      "loss": 0.0451105460524559,
      "num_input_tokens_seen": 2175528,
      "step": 1155,
      "train_runtime": 3129.9274,
      "train_tokens_per_second": 695.073
    },
    {
      "epoch": 1.1678038918372504,
      "grad_norm": 1.5774210691452026,
      "learning_rate": 8.354430379746837e-06,
      "loss": 0.07367590069770813,
      "num_input_tokens_seen": 2177430,
      "step": 1156,
      "train_runtime": 3132.1179,
      "train_tokens_per_second": 695.194
    },
    {
      "epoch": 1.168814758655547,
      "grad_norm": 1.8974392414093018,
      "learning_rate": 8.344303797468354e-06,
      "loss": 0.06681827455759048,
      "num_input_tokens_seen": 2178862,
      "step": 1157,
      "train_runtime": 3134.2574,
      "train_tokens_per_second": 695.176
    },
    {
      "epoch": 1.1698256254738437,
      "grad_norm": 1.4887057542800903,
      "learning_rate": 8.334177215189873e-06,
      "loss": 0.06238287687301636,
      "num_input_tokens_seen": 2180700,
      "step": 1158,
      "train_runtime": 3136.4886,
      "train_tokens_per_second": 695.268
    },
    {
      "epoch": 1.1708364922921406,
      "grad_norm": 1.715378761291504,
      "learning_rate": 8.324050632911394e-06,
      "loss": 0.08617313206195831,
      "num_input_tokens_seen": 2182790,
      "step": 1159,
      "train_runtime": 3138.7448,
      "train_tokens_per_second": 695.434
    },
    {
      "epoch": 1.1718473591104372,
      "grad_norm": 2.034440755844116,
      "learning_rate": 8.313924050632911e-06,
      "loss": 0.1255088746547699,
      "num_input_tokens_seen": 2184616,
      "step": 1160,
      "train_runtime": 3140.9391,
      "train_tokens_per_second": 695.53
    },
    {
      "epoch": 1.172858225928734,
      "grad_norm": 1.7983278036117554,
      "learning_rate": 8.30379746835443e-06,
      "loss": 0.07602868229150772,
      "num_input_tokens_seen": 2186642,
      "step": 1161,
      "train_runtime": 3143.1914,
      "train_tokens_per_second": 695.676
    },
    {
      "epoch": 1.1738690927470306,
      "grad_norm": 2.0422070026397705,
      "learning_rate": 8.29367088607595e-06,
      "loss": 0.06293034553527832,
      "num_input_tokens_seen": 2188852,
      "step": 1162,
      "train_runtime": 3145.5159,
      "train_tokens_per_second": 695.864
    },
    {
      "epoch": 1.1748799595653272,
      "grad_norm": 1.6136746406555176,
      "learning_rate": 8.283544303797469e-06,
      "loss": 0.07936811447143555,
      "num_input_tokens_seen": 2190730,
      "step": 1163,
      "train_runtime": 3147.7689,
      "train_tokens_per_second": 695.963
    },
    {
      "epoch": 1.175890826383624,
      "grad_norm": 1.4614843130111694,
      "learning_rate": 8.273417721518988e-06,
      "loss": 0.04913468658924103,
      "num_input_tokens_seen": 2192420,
      "step": 1164,
      "train_runtime": 3150.0158,
      "train_tokens_per_second": 696.003
    },
    {
      "epoch": 1.1769016932019207,
      "grad_norm": 1.6033903360366821,
      "learning_rate": 8.263291139240507e-06,
      "loss": 0.048490870743989944,
      "num_input_tokens_seen": 2194292,
      "step": 1165,
      "train_runtime": 3152.2455,
      "train_tokens_per_second": 696.104
    },
    {
      "epoch": 1.1779125600202174,
      "grad_norm": 1.5246518850326538,
      "learning_rate": 8.253164556962026e-06,
      "loss": 0.04833872616291046,
      "num_input_tokens_seen": 2196356,
      "step": 1166,
      "train_runtime": 3154.5205,
      "train_tokens_per_second": 696.257
    },
    {
      "epoch": 1.178923426838514,
      "grad_norm": 1.6408116817474365,
      "learning_rate": 8.243037974683545e-06,
      "loss": 0.09846121072769165,
      "num_input_tokens_seen": 2198536,
      "step": 1167,
      "train_runtime": 3156.8459,
      "train_tokens_per_second": 696.434
    },
    {
      "epoch": 1.1799342936568107,
      "grad_norm": 1.8436825275421143,
      "learning_rate": 8.232911392405064e-06,
      "loss": 0.06969211995601654,
      "num_input_tokens_seen": 2200490,
      "step": 1168,
      "train_runtime": 3159.1275,
      "train_tokens_per_second": 696.55
    },
    {
      "epoch": 1.1809451604751073,
      "grad_norm": 1.651782512664795,
      "learning_rate": 8.222784810126583e-06,
      "loss": 0.06236081197857857,
      "num_input_tokens_seen": 2202402,
      "step": 1169,
      "train_runtime": 3161.3711,
      "train_tokens_per_second": 696.66
    },
    {
      "epoch": 1.181956027293404,
      "grad_norm": 1.5295475721359253,
      "learning_rate": 8.212658227848102e-06,
      "loss": 0.05375359207391739,
      "num_input_tokens_seen": 2204174,
      "step": 1170,
      "train_runtime": 3163.5712,
      "train_tokens_per_second": 696.736
    },
    {
      "epoch": 1.1829668941117009,
      "grad_norm": 1.7514253854751587,
      "learning_rate": 8.202531645569621e-06,
      "loss": 0.07153820246458054,
      "num_input_tokens_seen": 2206006,
      "step": 1171,
      "train_runtime": 3172.0097,
      "train_tokens_per_second": 695.46
    },
    {
      "epoch": 1.1839777609299975,
      "grad_norm": 1.4838099479675293,
      "learning_rate": 8.19240506329114e-06,
      "loss": 0.06446850299835205,
      "num_input_tokens_seen": 2207678,
      "step": 1172,
      "train_runtime": 3174.251,
      "train_tokens_per_second": 695.496
    },
    {
      "epoch": 1.1849886277482942,
      "grad_norm": 1.6390488147735596,
      "learning_rate": 8.18227848101266e-06,
      "loss": 0.06537766754627228,
      "num_input_tokens_seen": 2209704,
      "step": 1173,
      "train_runtime": 3176.5177,
      "train_tokens_per_second": 695.637
    },
    {
      "epoch": 1.1859994945665908,
      "grad_norm": 1.5963364839553833,
      "learning_rate": 8.172151898734177e-06,
      "loss": 0.07196581363677979,
      "num_input_tokens_seen": 2211704,
      "step": 1174,
      "train_runtime": 3178.7396,
      "train_tokens_per_second": 695.78
    },
    {
      "epoch": 1.1870103613848875,
      "grad_norm": 1.6719778776168823,
      "learning_rate": 8.162025316455698e-06,
      "loss": 0.07993964850902557,
      "num_input_tokens_seen": 2213624,
      "step": 1175,
      "train_runtime": 3180.9662,
      "train_tokens_per_second": 695.897
    },
    {
      "epoch": 1.1880212282031843,
      "grad_norm": 1.875149130821228,
      "learning_rate": 8.151898734177217e-06,
      "loss": 0.0808754414319992,
      "num_input_tokens_seen": 2215790,
      "step": 1176,
      "train_runtime": 3183.2524,
      "train_tokens_per_second": 696.077
    },
    {
      "epoch": 1.189032095021481,
      "grad_norm": 1.675183892250061,
      "learning_rate": 8.141772151898734e-06,
      "loss": 0.07747514545917511,
      "num_input_tokens_seen": 2217462,
      "step": 1177,
      "train_runtime": 3185.388,
      "train_tokens_per_second": 696.136
    },
    {
      "epoch": 1.1900429618397776,
      "grad_norm": 1.878333330154419,
      "learning_rate": 8.131645569620253e-06,
      "loss": 0.08744428306818008,
      "num_input_tokens_seen": 2219292,
      "step": 1178,
      "train_runtime": 3187.5866,
      "train_tokens_per_second": 696.23
    },
    {
      "epoch": 1.1910538286580743,
      "grad_norm": 2.005580186843872,
      "learning_rate": 8.121518987341774e-06,
      "loss": 0.07720157504081726,
      "num_input_tokens_seen": 2221190,
      "step": 1179,
      "train_runtime": 3189.8116,
      "train_tokens_per_second": 696.339
    },
    {
      "epoch": 1.192064695476371,
      "grad_norm": 2.0327203273773193,
      "learning_rate": 8.111392405063291e-06,
      "loss": 0.09617538005113602,
      "num_input_tokens_seen": 2223196,
      "step": 1180,
      "train_runtime": 3192.0674,
      "train_tokens_per_second": 696.475
    },
    {
      "epoch": 1.1930755622946676,
      "grad_norm": 1.6846823692321777,
      "learning_rate": 8.10126582278481e-06,
      "loss": 0.06675636768341064,
      "num_input_tokens_seen": 2224970,
      "step": 1181,
      "train_runtime": 3194.2609,
      "train_tokens_per_second": 696.552
    },
    {
      "epoch": 1.1940864291129643,
      "grad_norm": 2.134172201156616,
      "learning_rate": 8.09113924050633e-06,
      "loss": 0.1031813770532608,
      "num_input_tokens_seen": 2227122,
      "step": 1182,
      "train_runtime": 3196.5133,
      "train_tokens_per_second": 696.735
    },
    {
      "epoch": 1.1950972959312611,
      "grad_norm": 1.6644850969314575,
      "learning_rate": 8.081012658227849e-06,
      "loss": 0.053227800875902176,
      "num_input_tokens_seen": 2229118,
      "step": 1183,
      "train_runtime": 3198.7373,
      "train_tokens_per_second": 696.874
    },
    {
      "epoch": 1.1961081627495578,
      "grad_norm": 1.7365634441375732,
      "learning_rate": 8.070886075949368e-06,
      "loss": 0.05617215484380722,
      "num_input_tokens_seen": 2230492,
      "step": 1184,
      "train_runtime": 3200.8117,
      "train_tokens_per_second": 696.852
    },
    {
      "epoch": 1.1971190295678544,
      "grad_norm": 1.5770543813705444,
      "learning_rate": 8.060759493670887e-06,
      "loss": 0.07203859090805054,
      "num_input_tokens_seen": 2232598,
      "step": 1185,
      "train_runtime": 3203.0823,
      "train_tokens_per_second": 697.016
    },
    {
      "epoch": 1.198129896386151,
      "grad_norm": 1.6365536451339722,
      "learning_rate": 8.050632911392406e-06,
      "loss": 0.05470001697540283,
      "num_input_tokens_seen": 2234364,
      "step": 1186,
      "train_runtime": 3205.2707,
      "train_tokens_per_second": 697.091
    },
    {
      "epoch": 1.1991407632044477,
      "grad_norm": 1.6714481115341187,
      "learning_rate": 8.040506329113925e-06,
      "loss": 0.08494721353054047,
      "num_input_tokens_seen": 2236454,
      "step": 1187,
      "train_runtime": 3207.5649,
      "train_tokens_per_second": 697.244
    },
    {
      "epoch": 1.2001516300227446,
      "grad_norm": 2.886770248413086,
      "learning_rate": 8.030379746835444e-06,
      "loss": 0.10024242103099823,
      "num_input_tokens_seen": 2237934,
      "step": 1188,
      "train_runtime": 3209.6917,
      "train_tokens_per_second": 697.243
    },
    {
      "epoch": 1.2001516300227446,
      "eval_loss": 0.39543792605400085,
      "eval_runtime": 52.3082,
      "eval_samples_per_second": 16.804,
      "eval_steps_per_second": 8.412,
      "num_input_tokens_seen": 2237934,
      "step": 1188
    },
    {
      "epoch": 1.2011624968410413,
      "grad_norm": 1.674844741821289,
      "learning_rate": 8.020253164556963e-06,
      "loss": 0.0708998516201973,
      "num_input_tokens_seen": 2239668,
      "step": 1189,
      "train_runtime": 3264.2019,
      "train_tokens_per_second": 686.13
    },
    {
      "epoch": 1.202173363659338,
      "grad_norm": 1.280044674873352,
      "learning_rate": 8.01012658227848e-06,
      "loss": 0.06288723647594452,
      "num_input_tokens_seen": 2241998,
      "step": 1190,
      "train_runtime": 3266.5422,
      "train_tokens_per_second": 686.352
    },
    {
      "epoch": 1.2031842304776346,
      "grad_norm": 2.1572518348693848,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.07750942558050156,
      "num_input_tokens_seen": 2243886,
      "step": 1191,
      "train_runtime": 3268.7357,
      "train_tokens_per_second": 686.469
    },
    {
      "epoch": 1.2041950972959312,
      "grad_norm": 1.4469211101531982,
      "learning_rate": 7.98987341772152e-06,
      "loss": 0.05743473768234253,
      "num_input_tokens_seen": 2245630,
      "step": 1192,
      "train_runtime": 3270.9389,
      "train_tokens_per_second": 686.54
    },
    {
      "epoch": 1.2052059641142279,
      "grad_norm": 1.8174983263015747,
      "learning_rate": 7.979746835443038e-06,
      "loss": 0.08593261986970901,
      "num_input_tokens_seen": 2247608,
      "step": 1193,
      "train_runtime": 3273.2755,
      "train_tokens_per_second": 686.654
    },
    {
      "epoch": 1.2062168309325245,
      "grad_norm": 1.115770936012268,
      "learning_rate": 7.969620253164557e-06,
      "loss": 0.04833998158574104,
      "num_input_tokens_seen": 2249956,
      "step": 1194,
      "train_runtime": 3275.6114,
      "train_tokens_per_second": 686.881
    },
    {
      "epoch": 1.2072276977508214,
      "grad_norm": 1.8248049020767212,
      "learning_rate": 7.959493670886078e-06,
      "loss": 0.08127608895301819,
      "num_input_tokens_seen": 2251722,
      "step": 1195,
      "train_runtime": 3277.8099,
      "train_tokens_per_second": 686.959
    },
    {
      "epoch": 1.208238564569118,
      "grad_norm": 1.9218639135360718,
      "learning_rate": 7.949367088607595e-06,
      "loss": 0.0896994099020958,
      "num_input_tokens_seen": 2253246,
      "step": 1196,
      "train_runtime": 3279.9467,
      "train_tokens_per_second": 686.976
    },
    {
      "epoch": 1.2092494313874147,
      "grad_norm": 1.6879656314849854,
      "learning_rate": 7.939240506329114e-06,
      "loss": 0.06548882275819778,
      "num_input_tokens_seen": 2254870,
      "step": 1197,
      "train_runtime": 3282.1331,
      "train_tokens_per_second": 687.014
    },
    {
      "epoch": 1.2102602982057113,
      "grad_norm": 1.4788695573806763,
      "learning_rate": 7.929113924050633e-06,
      "loss": 0.05680616945028305,
      "num_input_tokens_seen": 2256480,
      "step": 1198,
      "train_runtime": 3284.3035,
      "train_tokens_per_second": 687.05
    },
    {
      "epoch": 1.211271165024008,
      "grad_norm": 1.8426119089126587,
      "learning_rate": 7.918987341772152e-06,
      "loss": 0.061318475753068924,
      "num_input_tokens_seen": 2257978,
      "step": 1199,
      "train_runtime": 3286.4056,
      "train_tokens_per_second": 687.066
    },
    {
      "epoch": 1.2122820318423049,
      "grad_norm": 2.1089558601379395,
      "learning_rate": 7.908860759493671e-06,
      "loss": 0.09348195791244507,
      "num_input_tokens_seen": 2259402,
      "step": 1200,
      "train_runtime": 3288.5173,
      "train_tokens_per_second": 687.058
    },
    {
      "epoch": 1.2132928986606015,
      "grad_norm": 1.671409010887146,
      "learning_rate": 7.89873417721519e-06,
      "loss": 0.060666583478450775,
      "num_input_tokens_seen": 2260964,
      "step": 1201,
      "train_runtime": 3296.9542,
      "train_tokens_per_second": 685.774
    },
    {
      "epoch": 1.2143037654788982,
      "grad_norm": 1.1846647262573242,
      "learning_rate": 7.88860759493671e-06,
      "loss": 0.04691457003355026,
      "num_input_tokens_seen": 2262478,
      "step": 1202,
      "train_runtime": 3299.137,
      "train_tokens_per_second": 685.779
    },
    {
      "epoch": 1.2153146322971948,
      "grad_norm": 1.8600661754608154,
      "learning_rate": 7.878481012658229e-06,
      "loss": 0.06159030646085739,
      "num_input_tokens_seen": 2263914,
      "step": 1203,
      "train_runtime": 3301.234,
      "train_tokens_per_second": 685.778
    },
    {
      "epoch": 1.2163254991154915,
      "grad_norm": 1.452722191810608,
      "learning_rate": 7.868354430379748e-06,
      "loss": 0.06574209034442902,
      "num_input_tokens_seen": 2266010,
      "step": 1204,
      "train_runtime": 3303.5207,
      "train_tokens_per_second": 685.938
    },
    {
      "epoch": 1.2173363659337881,
      "grad_norm": 1.654717206954956,
      "learning_rate": 7.858227848101267e-06,
      "loss": 0.08449318259954453,
      "num_input_tokens_seen": 2267862,
      "step": 1205,
      "train_runtime": 3305.6955,
      "train_tokens_per_second": 686.047
    },
    {
      "epoch": 1.218347232752085,
      "grad_norm": 1.8064467906951904,
      "learning_rate": 7.848101265822786e-06,
      "loss": 0.06694860756397247,
      "num_input_tokens_seen": 2269564,
      "step": 1206,
      "train_runtime": 3307.9064,
      "train_tokens_per_second": 686.103
    },
    {
      "epoch": 1.2193580995703817,
      "grad_norm": 1.4400807619094849,
      "learning_rate": 7.837974683544305e-06,
      "loss": 0.07201480865478516,
      "num_input_tokens_seen": 2272080,
      "step": 1207,
      "train_runtime": 3310.337,
      "train_tokens_per_second": 686.359
    },
    {
      "epoch": 1.2203689663886783,
      "grad_norm": 1.7585588693618774,
      "learning_rate": 7.827848101265824e-06,
      "loss": 0.06069216877222061,
      "num_input_tokens_seen": 2273334,
      "step": 1208,
      "train_runtime": 3312.3935,
      "train_tokens_per_second": 686.312
    },
    {
      "epoch": 1.221379833206975,
      "grad_norm": 1.5639764070510864,
      "learning_rate": 7.817721518987343e-06,
      "loss": 0.0779055655002594,
      "num_input_tokens_seen": 2275454,
      "step": 1209,
      "train_runtime": 3314.6832,
      "train_tokens_per_second": 686.477
    },
    {
      "epoch": 1.2223907000252716,
      "grad_norm": 2.509942054748535,
      "learning_rate": 7.80759493670886e-06,
      "loss": 0.05721501260995865,
      "num_input_tokens_seen": 2277052,
      "step": 1210,
      "train_runtime": 3316.906,
      "train_tokens_per_second": 686.499
    },
    {
      "epoch": 1.2234015668435683,
      "grad_norm": 1.4947055578231812,
      "learning_rate": 7.797468354430381e-06,
      "loss": 0.048078425228595734,
      "num_input_tokens_seen": 2278706,
      "step": 1211,
      "train_runtime": 3319.0757,
      "train_tokens_per_second": 686.548
    },
    {
      "epoch": 1.2244124336618651,
      "grad_norm": 2.03582501411438,
      "learning_rate": 7.7873417721519e-06,
      "loss": 0.07997502386569977,
      "num_input_tokens_seen": 2280384,
      "step": 1212,
      "train_runtime": 3321.2464,
      "train_tokens_per_second": 686.605
    },
    {
      "epoch": 1.2254233004801618,
      "grad_norm": 1.7681838274002075,
      "learning_rate": 7.777215189873418e-06,
      "loss": 0.07857958227396011,
      "num_input_tokens_seen": 2282836,
      "step": 1213,
      "train_runtime": 3323.6181,
      "train_tokens_per_second": 686.853
    },
    {
      "epoch": 1.2264341672984584,
      "grad_norm": 1.890169620513916,
      "learning_rate": 7.767088607594937e-06,
      "loss": 0.07384022325277328,
      "num_input_tokens_seen": 2284726,
      "step": 1214,
      "train_runtime": 3325.8392,
      "train_tokens_per_second": 686.962
    },
    {
      "epoch": 1.227445034116755,
      "grad_norm": 1.5807586908340454,
      "learning_rate": 7.756962025316456e-06,
      "loss": 0.07555880397558212,
      "num_input_tokens_seen": 2287752,
      "step": 1215,
      "train_runtime": 3328.3809,
      "train_tokens_per_second": 687.347
    },
    {
      "epoch": 1.2284559009350517,
      "grad_norm": 1.5621229410171509,
      "learning_rate": 7.746835443037975e-06,
      "loss": 0.05525794252753258,
      "num_input_tokens_seen": 2289266,
      "step": 1216,
      "train_runtime": 3330.5221,
      "train_tokens_per_second": 687.359
    },
    {
      "epoch": 1.2294667677533484,
      "grad_norm": 3.0601284503936768,
      "learning_rate": 7.736708860759494e-06,
      "loss": 0.06457633525133133,
      "num_input_tokens_seen": 2291274,
      "step": 1217,
      "train_runtime": 3332.7689,
      "train_tokens_per_second": 687.499
    },
    {
      "epoch": 1.2304776345716453,
      "grad_norm": 1.5983930826187134,
      "learning_rate": 7.726582278481013e-06,
      "loss": 0.06922073662281036,
      "num_input_tokens_seen": 2293442,
      "step": 1218,
      "train_runtime": 3335.0849,
      "train_tokens_per_second": 687.671
    },
    {
      "epoch": 1.231488501389942,
      "grad_norm": 1.310104489326477,
      "learning_rate": 7.716455696202532e-06,
      "loss": 0.06207498908042908,
      "num_input_tokens_seen": 2295438,
      "step": 1219,
      "train_runtime": 3337.3263,
      "train_tokens_per_second": 687.807
    },
    {
      "epoch": 1.2324993682082386,
      "grad_norm": 1.3984287977218628,
      "learning_rate": 7.706329113924051e-06,
      "loss": 0.053548961877822876,
      "num_input_tokens_seen": 2297814,
      "step": 1220,
      "train_runtime": 3339.6966,
      "train_tokens_per_second": 688.031
    },
    {
      "epoch": 1.2335102350265352,
      "grad_norm": 1.6000895500183105,
      "learning_rate": 7.69620253164557e-06,
      "loss": 0.06268253922462463,
      "num_input_tokens_seen": 2299686,
      "step": 1221,
      "train_runtime": 3341.9541,
      "train_tokens_per_second": 688.126
    },
    {
      "epoch": 1.2345211018448319,
      "grad_norm": 1.686050534248352,
      "learning_rate": 7.68607594936709e-06,
      "loss": 0.05675904080271721,
      "num_input_tokens_seen": 2301578,
      "step": 1222,
      "train_runtime": 3344.1509,
      "train_tokens_per_second": 688.24
    },
    {
      "epoch": 1.2355319686631288,
      "grad_norm": 1.8062411546707153,
      "learning_rate": 7.675949367088609e-06,
      "loss": 0.074558787047863,
      "num_input_tokens_seen": 2303610,
      "step": 1223,
      "train_runtime": 3346.3795,
      "train_tokens_per_second": 688.389
    },
    {
      "epoch": 1.2365428354814254,
      "grad_norm": 1.4807677268981934,
      "learning_rate": 7.665822784810128e-06,
      "loss": 0.05706555023789406,
      "num_input_tokens_seen": 2305512,
      "step": 1224,
      "train_runtime": 3348.5706,
      "train_tokens_per_second": 688.506
    },
    {
      "epoch": 1.237553702299722,
      "grad_norm": 1.8867607116699219,
      "learning_rate": 7.655696202531647e-06,
      "loss": 0.06920222193002701,
      "num_input_tokens_seen": 2307674,
      "step": 1225,
      "train_runtime": 3350.8528,
      "train_tokens_per_second": 688.683
    },
    {
      "epoch": 1.2385645691180187,
      "grad_norm": 2.331836462020874,
      "learning_rate": 7.645569620253164e-06,
      "loss": 0.07282140105962753,
      "num_input_tokens_seen": 2309670,
      "step": 1226,
      "train_runtime": 3353.0765,
      "train_tokens_per_second": 688.821
    },
    {
      "epoch": 1.2395754359363154,
      "grad_norm": 1.5728611946105957,
      "learning_rate": 7.635443037974685e-06,
      "loss": 0.05360238626599312,
      "num_input_tokens_seen": 2311322,
      "step": 1227,
      "train_runtime": 3355.2675,
      "train_tokens_per_second": 688.864
    },
    {
      "epoch": 1.240586302754612,
      "grad_norm": 2.0821712017059326,
      "learning_rate": 7.625316455696203e-06,
      "loss": 0.06303103268146515,
      "num_input_tokens_seen": 2312726,
      "step": 1228,
      "train_runtime": 3357.3611,
      "train_tokens_per_second": 688.852
    },
    {
      "epoch": 1.2415971695729087,
      "grad_norm": 1.519388198852539,
      "learning_rate": 7.615189873417722e-06,
      "loss": 0.06824515759944916,
      "num_input_tokens_seen": 2314644,
      "step": 1229,
      "train_runtime": 3359.5486,
      "train_tokens_per_second": 688.975
    },
    {
      "epoch": 1.2426080363912055,
      "grad_norm": 1.7579048871994019,
      "learning_rate": 7.6050632911392405e-06,
      "loss": 0.06210828199982643,
      "num_input_tokens_seen": 2316536,
      "step": 1230,
      "train_runtime": 3361.7631,
      "train_tokens_per_second": 689.084
    },
    {
      "epoch": 1.2436189032095022,
      "grad_norm": 2.1102397441864014,
      "learning_rate": 7.5949367088607605e-06,
      "loss": 0.06933826208114624,
      "num_input_tokens_seen": 2318024,
      "step": 1231,
      "train_runtime": 3370.5171,
      "train_tokens_per_second": 687.735
    },
    {
      "epoch": 1.2446297700277988,
      "grad_norm": 2.117542028427124,
      "learning_rate": 7.5848101265822796e-06,
      "loss": 0.09398352354764938,
      "num_input_tokens_seen": 2320428,
      "step": 1232,
      "train_runtime": 3372.8486,
      "train_tokens_per_second": 687.973
    },
    {
      "epoch": 1.2456406368460955,
      "grad_norm": 1.7313145399093628,
      "learning_rate": 7.574683544303798e-06,
      "loss": 0.080222949385643,
      "num_input_tokens_seen": 2322182,
      "step": 1233,
      "train_runtime": 3375.063,
      "train_tokens_per_second": 688.041
    },
    {
      "epoch": 1.2466515036643921,
      "grad_norm": 1.8148479461669922,
      "learning_rate": 7.564556962025317e-06,
      "loss": 0.07266001403331757,
      "num_input_tokens_seen": 2323616,
      "step": 1234,
      "train_runtime": 3377.1932,
      "train_tokens_per_second": 688.032
    },
    {
      "epoch": 1.247662370482689,
      "grad_norm": 1.7799381017684937,
      "learning_rate": 7.554430379746837e-06,
      "loss": 0.09345235675573349,
      "num_input_tokens_seen": 2325408,
      "step": 1235,
      "train_runtime": 3379.3963,
      "train_tokens_per_second": 688.113
    },
    {
      "epoch": 1.2486732373009857,
      "grad_norm": 1.1585361957550049,
      "learning_rate": 7.544303797468355e-06,
      "loss": 0.05592569708824158,
      "num_input_tokens_seen": 2327858,
      "step": 1236,
      "train_runtime": 3381.7409,
      "train_tokens_per_second": 688.361
    },
    {
      "epoch": 1.2496841041192823,
      "grad_norm": 1.718593716621399,
      "learning_rate": 7.534177215189874e-06,
      "loss": 0.05969762057065964,
      "num_input_tokens_seen": 2329798,
      "step": 1237,
      "train_runtime": 3383.9627,
      "train_tokens_per_second": 688.482
    },
    {
      "epoch": 1.250694970937579,
      "grad_norm": 1.499233365058899,
      "learning_rate": 7.524050632911392e-06,
      "loss": 0.06824598461389542,
      "num_input_tokens_seen": 2332188,
      "step": 1238,
      "train_runtime": 3386.3049,
      "train_tokens_per_second": 688.712
    },
    {
      "epoch": 1.2517058377558756,
      "grad_norm": 1.586361050605774,
      "learning_rate": 7.513924050632912e-06,
      "loss": 0.07227626442909241,
      "num_input_tokens_seen": 2334310,
      "step": 1239,
      "train_runtime": 3388.5468,
      "train_tokens_per_second": 688.882
    },
    {
      "epoch": 1.2527167045741723,
      "grad_norm": 1.6351972818374634,
      "learning_rate": 7.503797468354431e-06,
      "loss": 0.06726763397455215,
      "num_input_tokens_seen": 2336354,
      "step": 1240,
      "train_runtime": 3390.7876,
      "train_tokens_per_second": 689.03
    },
    {
      "epoch": 1.253727571392469,
      "grad_norm": 1.3885319232940674,
      "learning_rate": 7.49367088607595e-06,
      "loss": 0.05118235945701599,
      "num_input_tokens_seen": 2338472,
      "step": 1241,
      "train_runtime": 3393.1025,
      "train_tokens_per_second": 689.184
    },
    {
      "epoch": 1.2547384382107658,
      "grad_norm": 1.7103335857391357,
      "learning_rate": 7.483544303797469e-06,
      "loss": 0.07767714560031891,
      "num_input_tokens_seen": 2340730,
      "step": 1242,
      "train_runtime": 3395.3934,
      "train_tokens_per_second": 689.384
    },
    {
      "epoch": 1.2557493050290625,
      "grad_norm": 1.4612255096435547,
      "learning_rate": 7.473417721518989e-06,
      "loss": 0.07450965791940689,
      "num_input_tokens_seen": 2342948,
      "step": 1243,
      "train_runtime": 3397.6933,
      "train_tokens_per_second": 689.57
    },
    {
      "epoch": 1.256760171847359,
      "grad_norm": 1.659128189086914,
      "learning_rate": 7.463291139240507e-06,
      "loss": 0.06260944157838821,
      "num_input_tokens_seen": 2345132,
      "step": 1244,
      "train_runtime": 3399.999,
      "train_tokens_per_second": 689.745
    },
    {
      "epoch": 1.2577710386656558,
      "grad_norm": 1.879957675933838,
      "learning_rate": 7.453164556962026e-06,
      "loss": 0.08181922882795334,
      "num_input_tokens_seen": 2346772,
      "step": 1245,
      "train_runtime": 3402.1789,
      "train_tokens_per_second": 689.785
    },
    {
      "epoch": 1.2587819054839524,
      "grad_norm": 1.8026577234268188,
      "learning_rate": 7.443037974683544e-06,
      "loss": 0.07521645724773407,
      "num_input_tokens_seen": 2348770,
      "step": 1246,
      "train_runtime": 3404.4302,
      "train_tokens_per_second": 689.916
    },
    {
      "epoch": 1.2597927723022493,
      "grad_norm": 1.6696979999542236,
      "learning_rate": 7.432911392405064e-06,
      "loss": 0.08103272318840027,
      "num_input_tokens_seen": 2350794,
      "step": 1247,
      "train_runtime": 3406.6848,
      "train_tokens_per_second": 690.053
    },
    {
      "epoch": 1.260803639120546,
      "grad_norm": 1.6551284790039062,
      "learning_rate": 7.422784810126583e-06,
      "loss": 0.06993693858385086,
      "num_input_tokens_seen": 2352492,
      "step": 1248,
      "train_runtime": 3408.8427,
      "train_tokens_per_second": 690.115
    },
    {
      "epoch": 1.2618145059388426,
      "grad_norm": 2.096829891204834,
      "learning_rate": 7.4126582278481014e-06,
      "loss": 0.09812723100185394,
      "num_input_tokens_seen": 2354176,
      "step": 1249,
      "train_runtime": 3411.0168,
      "train_tokens_per_second": 690.168
    },
    {
      "epoch": 1.2628253727571392,
      "grad_norm": 1.4124011993408203,
      "learning_rate": 7.4025316455696205e-06,
      "loss": 0.06310103833675385,
      "num_input_tokens_seen": 2356116,
      "step": 1250,
      "train_runtime": 3413.2703,
      "train_tokens_per_second": 690.281
    },
    {
      "epoch": 1.263836239575436,
      "grad_norm": 2.1522560119628906,
      "learning_rate": 7.3924050632911405e-06,
      "loss": 0.1018192321062088,
      "num_input_tokens_seen": 2357662,
      "step": 1251,
      "train_runtime": 3415.4392,
      "train_tokens_per_second": 690.295
    },
    {
      "epoch": 1.2648471063937325,
      "grad_norm": 1.7819515466690063,
      "learning_rate": 7.382278481012659e-06,
      "loss": 0.06440679728984833,
      "num_input_tokens_seen": 2359266,
      "step": 1252,
      "train_runtime": 3417.5953,
      "train_tokens_per_second": 690.329
    },
    {
      "epoch": 1.2658579732120292,
      "grad_norm": 1.6107395887374878,
      "learning_rate": 7.372151898734178e-06,
      "loss": 0.05555751174688339,
      "num_input_tokens_seen": 2361054,
      "step": 1253,
      "train_runtime": 3419.8092,
      "train_tokens_per_second": 690.405
    },
    {
      "epoch": 1.266868840030326,
      "grad_norm": 1.7699815034866333,
      "learning_rate": 7.362025316455697e-06,
      "loss": 0.07139113545417786,
      "num_input_tokens_seen": 2362734,
      "step": 1254,
      "train_runtime": 3421.9669,
      "train_tokens_per_second": 690.461
    },
    {
      "epoch": 1.2678797068486227,
      "grad_norm": 1.6306917667388916,
      "learning_rate": 7.351898734177216e-06,
      "loss": 0.08092683553695679,
      "num_input_tokens_seen": 2364500,
      "step": 1255,
      "train_runtime": 3424.1596,
      "train_tokens_per_second": 690.534
    },
    {
      "epoch": 1.2688905736669194,
      "grad_norm": 1.4850523471832275,
      "learning_rate": 7.341772151898735e-06,
      "loss": 0.05796663090586662,
      "num_input_tokens_seen": 2366698,
      "step": 1256,
      "train_runtime": 3426.4319,
      "train_tokens_per_second": 690.718
    },
    {
      "epoch": 1.269901440485216,
      "grad_norm": 2.2261545658111572,
      "learning_rate": 7.331645569620254e-06,
      "loss": 0.06608270108699799,
      "num_input_tokens_seen": 2368722,
      "step": 1257,
      "train_runtime": 3428.6924,
      "train_tokens_per_second": 690.853
    },
    {
      "epoch": 1.270912307303513,
      "grad_norm": 1.7036511898040771,
      "learning_rate": 7.321518987341772e-06,
      "loss": 0.07144436240196228,
      "num_input_tokens_seen": 2370806,
      "step": 1258,
      "train_runtime": 3430.9664,
      "train_tokens_per_second": 691.002
    },
    {
      "epoch": 1.2719231741218096,
      "grad_norm": 1.2830642461776733,
      "learning_rate": 7.311392405063292e-06,
      "loss": 0.05315488949418068,
      "num_input_tokens_seen": 2372522,
      "step": 1259,
      "train_runtime": 3433.1884,
      "train_tokens_per_second": 691.055
    },
    {
      "epoch": 1.2729340409401062,
      "grad_norm": 1.4549801349639893,
      "learning_rate": 7.301265822784811e-06,
      "loss": 0.056591495871543884,
      "num_input_tokens_seen": 2374258,
      "step": 1260,
      "train_runtime": 3435.3868,
      "train_tokens_per_second": 691.118
    },
    {
      "epoch": 1.2739449077584029,
      "grad_norm": 1.6096266508102417,
      "learning_rate": 7.29113924050633e-06,
      "loss": 0.05757144093513489,
      "num_input_tokens_seen": 2375890,
      "step": 1261,
      "train_runtime": 3444.0821,
      "train_tokens_per_second": 689.847
    },
    {
      "epoch": 1.2749557745766995,
      "grad_norm": 1.963739037513733,
      "learning_rate": 7.281012658227849e-06,
      "loss": 0.04849129542708397,
      "num_input_tokens_seen": 2377974,
      "step": 1262,
      "train_runtime": 3446.4063,
      "train_tokens_per_second": 689.987
    },
    {
      "epoch": 1.2759666413949962,
      "grad_norm": 1.474272608757019,
      "learning_rate": 7.270886075949367e-06,
      "loss": 0.06390251219272614,
      "num_input_tokens_seen": 2379908,
      "step": 1263,
      "train_runtime": 3448.6525,
      "train_tokens_per_second": 690.098
    },
    {
      "epoch": 1.2769775082132928,
      "grad_norm": 1.9414095878601074,
      "learning_rate": 7.260759493670887e-06,
      "loss": 0.08182346820831299,
      "num_input_tokens_seen": 2382120,
      "step": 1264,
      "train_runtime": 3450.9357,
      "train_tokens_per_second": 690.282
    },
    {
      "epoch": 1.2779883750315895,
      "grad_norm": 1.2520118951797485,
      "learning_rate": 7.250632911392406e-06,
      "loss": 0.049603234976530075,
      "num_input_tokens_seen": 2383738,
      "step": 1265,
      "train_runtime": 3453.0761,
      "train_tokens_per_second": 690.323
    },
    {
      "epoch": 1.2789992418498863,
      "grad_norm": 1.4932117462158203,
      "learning_rate": 7.240506329113924e-06,
      "loss": 0.06469394266605377,
      "num_input_tokens_seen": 2385874,
      "step": 1266,
      "train_runtime": 3455.3244,
      "train_tokens_per_second": 690.492
    },
    {
      "epoch": 1.280010108668183,
      "grad_norm": 1.7077187299728394,
      "learning_rate": 7.230379746835443e-06,
      "loss": 0.0767069011926651,
      "num_input_tokens_seen": 2388636,
      "step": 1267,
      "train_runtime": 3457.7679,
      "train_tokens_per_second": 690.803
    },
    {
      "epoch": 1.2810209754864796,
      "grad_norm": 1.7286819219589233,
      "learning_rate": 7.220253164556963e-06,
      "loss": 0.0681333988904953,
      "num_input_tokens_seen": 2390426,
      "step": 1268,
      "train_runtime": 3459.9379,
      "train_tokens_per_second": 690.887
    },
    {
      "epoch": 1.2820318423047763,
      "grad_norm": 1.715703010559082,
      "learning_rate": 7.2101265822784814e-06,
      "loss": 0.08162987977266312,
      "num_input_tokens_seen": 2392536,
      "step": 1269,
      "train_runtime": 3462.1887,
      "train_tokens_per_second": 691.047
    },
    {
      "epoch": 1.2830427091230732,
      "grad_norm": 2.2412264347076416,
      "learning_rate": 7.2000000000000005e-06,
      "loss": 0.06202499195933342,
      "num_input_tokens_seen": 2394276,
      "step": 1270,
      "train_runtime": 3464.3776,
      "train_tokens_per_second": 691.113
    },
    {
      "epoch": 1.2840535759413698,
      "grad_norm": 1.5789408683776855,
      "learning_rate": 7.189873417721519e-06,
      "loss": 0.07125669717788696,
      "num_input_tokens_seen": 2396094,
      "step": 1271,
      "train_runtime": 3466.5662,
      "train_tokens_per_second": 691.201
    },
    {
      "epoch": 1.2850644427596665,
      "grad_norm": 1.8839894533157349,
      "learning_rate": 7.179746835443039e-06,
      "loss": 0.10887087136507034,
      "num_input_tokens_seen": 2398042,
      "step": 1272,
      "train_runtime": 3468.7489,
      "train_tokens_per_second": 691.328
    },
    {
      "epoch": 1.2860753095779631,
      "grad_norm": 1.8170326948165894,
      "learning_rate": 7.169620253164558e-06,
      "loss": 0.0829956978559494,
      "num_input_tokens_seen": 2400290,
      "step": 1273,
      "train_runtime": 3471.0668,
      "train_tokens_per_second": 691.514
    },
    {
      "epoch": 1.2870861763962598,
      "grad_norm": 1.7981148958206177,
      "learning_rate": 7.159493670886076e-06,
      "loss": 0.06110212579369545,
      "num_input_tokens_seen": 2402024,
      "step": 1274,
      "train_runtime": 3473.2338,
      "train_tokens_per_second": 691.581
    },
    {
      "epoch": 1.2880970432145564,
      "grad_norm": 1.662429928779602,
      "learning_rate": 7.149367088607595e-06,
      "loss": 0.07427185028791428,
      "num_input_tokens_seen": 2403846,
      "step": 1275,
      "train_runtime": 3475.4589,
      "train_tokens_per_second": 691.663
    },
    {
      "epoch": 1.289107910032853,
      "grad_norm": 1.905470371246338,
      "learning_rate": 7.139240506329115e-06,
      "loss": 0.07187589257955551,
      "num_input_tokens_seen": 2405930,
      "step": 1276,
      "train_runtime": 3477.7463,
      "train_tokens_per_second": 691.807
    },
    {
      "epoch": 1.2901187768511497,
      "grad_norm": 1.6884658336639404,
      "learning_rate": 7.129113924050633e-06,
      "loss": 0.07680030167102814,
      "num_input_tokens_seen": 2407904,
      "step": 1277,
      "train_runtime": 3479.9972,
      "train_tokens_per_second": 691.927
    },
    {
      "epoch": 1.2911296436694466,
      "grad_norm": 1.6698951721191406,
      "learning_rate": 7.118987341772152e-06,
      "loss": 0.0777999758720398,
      "num_input_tokens_seen": 2410116,
      "step": 1278,
      "train_runtime": 3482.2895,
      "train_tokens_per_second": 692.107
    },
    {
      "epoch": 1.2921405104877433,
      "grad_norm": 1.534956932067871,
      "learning_rate": 7.108860759493671e-06,
      "loss": 0.07324867695569992,
      "num_input_tokens_seen": 2412480,
      "step": 1279,
      "train_runtime": 3484.6353,
      "train_tokens_per_second": 692.319
    },
    {
      "epoch": 1.29315137730604,
      "grad_norm": 2.1732232570648193,
      "learning_rate": 7.0987341772151905e-06,
      "loss": 0.07079564034938812,
      "num_input_tokens_seen": 2413934,
      "step": 1280,
      "train_runtime": 3486.712,
      "train_tokens_per_second": 692.324
    },
    {
      "epoch": 1.2941622441243366,
      "grad_norm": 1.7259658575057983,
      "learning_rate": 7.08860759493671e-06,
      "loss": 0.0802023857831955,
      "num_input_tokens_seen": 2415538,
      "step": 1281,
      "train_runtime": 3488.8518,
      "train_tokens_per_second": 692.359
    },
    {
      "epoch": 1.2951731109426334,
      "grad_norm": 1.2600252628326416,
      "learning_rate": 7.078481012658228e-06,
      "loss": 0.05568865314126015,
      "num_input_tokens_seen": 2418354,
      "step": 1282,
      "train_runtime": 3491.3211,
      "train_tokens_per_second": 692.676
    },
    {
      "epoch": 1.29618397776093,
      "grad_norm": 1.670655369758606,
      "learning_rate": 7.068354430379747e-06,
      "loss": 0.0647815465927124,
      "num_input_tokens_seen": 2420410,
      "step": 1283,
      "train_runtime": 3493.5914,
      "train_tokens_per_second": 692.814
    },
    {
      "epoch": 1.2971948445792267,
      "grad_norm": 1.5092346668243408,
      "learning_rate": 7.058227848101267e-06,
      "loss": 0.053298354148864746,
      "num_input_tokens_seen": 2422074,
      "step": 1284,
      "train_runtime": 3495.7382,
      "train_tokens_per_second": 692.865
    },
    {
      "epoch": 1.2982057113975234,
      "grad_norm": 1.9336142539978027,
      "learning_rate": 7.048101265822785e-06,
      "loss": 0.061506543308496475,
      "num_input_tokens_seen": 2423840,
      "step": 1285,
      "train_runtime": 3497.9079,
      "train_tokens_per_second": 692.94
    },
    {
      "epoch": 1.29921657821582,
      "grad_norm": 1.463113784790039,
      "learning_rate": 7.037974683544304e-06,
      "loss": 0.06923742592334747,
      "num_input_tokens_seen": 2425850,
      "step": 1286,
      "train_runtime": 3500.1456,
      "train_tokens_per_second": 693.071
    },
    {
      "epoch": 1.3002274450341167,
      "grad_norm": 1.5164885520935059,
      "learning_rate": 7.027848101265823e-06,
      "loss": 0.06922876834869385,
      "num_input_tokens_seen": 2428110,
      "step": 1287,
      "train_runtime": 3502.4778,
      "train_tokens_per_second": 693.255
    },
    {
      "epoch": 1.3012383118524133,
      "grad_norm": 1.9097461700439453,
      "learning_rate": 7.017721518987342e-06,
      "loss": 0.06577549129724503,
      "num_input_tokens_seen": 2430236,
      "step": 1288,
      "train_runtime": 3504.7619,
      "train_tokens_per_second": 693.41
    },
    {
      "epoch": 1.3022491786707102,
      "grad_norm": 1.7521017789840698,
      "learning_rate": 7.0075949367088614e-06,
      "loss": 0.06287643313407898,
      "num_input_tokens_seen": 2432132,
      "step": 1289,
      "train_runtime": 3507.026,
      "train_tokens_per_second": 693.503
    },
    {
      "epoch": 1.3032600454890069,
      "grad_norm": 2.0503952503204346,
      "learning_rate": 6.99746835443038e-06,
      "loss": 0.0946616381406784,
      "num_input_tokens_seen": 2433912,
      "step": 1290,
      "train_runtime": 3509.2411,
      "train_tokens_per_second": 693.572
    },
    {
      "epoch": 1.3042709123073035,
      "grad_norm": 1.9230647087097168,
      "learning_rate": 6.987341772151899e-06,
      "loss": 0.07685060054063797,
      "num_input_tokens_seen": 2435450,
      "step": 1291,
      "train_runtime": 3517.8879,
      "train_tokens_per_second": 692.305
    },
    {
      "epoch": 1.3052817791256002,
      "grad_norm": 1.7252345085144043,
      "learning_rate": 6.977215189873419e-06,
      "loss": 0.0595974363386631,
      "num_input_tokens_seen": 2437190,
      "step": 1292,
      "train_runtime": 3520.0929,
      "train_tokens_per_second": 692.365
    },
    {
      "epoch": 1.3062926459438968,
      "grad_norm": 1.6326429843902588,
      "learning_rate": 6.967088607594937e-06,
      "loss": 0.06852579861879349,
      "num_input_tokens_seen": 2439320,
      "step": 1293,
      "train_runtime": 3522.3841,
      "train_tokens_per_second": 692.52
    },
    {
      "epoch": 1.3073035127621937,
      "grad_norm": 1.673790693283081,
      "learning_rate": 6.956962025316456e-06,
      "loss": 0.06482592970132828,
      "num_input_tokens_seen": 2440856,
      "step": 1294,
      "train_runtime": 3524.5478,
      "train_tokens_per_second": 692.53
    },
    {
      "epoch": 1.3083143795804903,
      "grad_norm": 1.5817533731460571,
      "learning_rate": 6.946835443037975e-06,
      "loss": 0.060081928968429565,
      "num_input_tokens_seen": 2442592,
      "step": 1295,
      "train_runtime": 3526.7066,
      "train_tokens_per_second": 692.599
    },
    {
      "epoch": 1.309325246398787,
      "grad_norm": 1.8944607973098755,
      "learning_rate": 6.936708860759494e-06,
      "loss": 0.08795922249555588,
      "num_input_tokens_seen": 2444864,
      "step": 1296,
      "train_runtime": 3529.0064,
      "train_tokens_per_second": 692.791
    },
    {
      "epoch": 1.3103361132170837,
      "grad_norm": 2.002056360244751,
      "learning_rate": 6.926582278481013e-06,
      "loss": 0.06857173889875412,
      "num_input_tokens_seen": 2446738,
      "step": 1297,
      "train_runtime": 3531.2228,
      "train_tokens_per_second": 692.887
    },
    {
      "epoch": 1.3113469800353803,
      "grad_norm": 1.8821349143981934,
      "learning_rate": 6.916455696202532e-06,
      "loss": 0.08960074186325073,
      "num_input_tokens_seen": 2448408,
      "step": 1298,
      "train_runtime": 3533.4319,
      "train_tokens_per_second": 692.926
    },
    {
      "epoch": 1.312357846853677,
      "grad_norm": 1.921169400215149,
      "learning_rate": 6.906329113924051e-06,
      "loss": 0.0884164348244667,
      "num_input_tokens_seen": 2450152,
      "step": 1299,
      "train_runtime": 3535.6275,
      "train_tokens_per_second": 692.989
    },
    {
      "epoch": 1.3133687136719736,
      "grad_norm": 1.53725266456604,
      "learning_rate": 6.8962025316455705e-06,
      "loss": 0.07701297849416733,
      "num_input_tokens_seen": 2451912,
      "step": 1300,
      "train_runtime": 3537.8871,
      "train_tokens_per_second": 693.044
    },
    {
      "epoch": 1.3143795804902705,
      "grad_norm": 1.5910474061965942,
      "learning_rate": 6.88607594936709e-06,
      "loss": 0.05664290860295296,
      "num_input_tokens_seen": 2453816,
      "step": 1301,
      "train_runtime": 3540.0618,
      "train_tokens_per_second": 693.156
    },
    {
      "epoch": 1.3153904473085671,
      "grad_norm": 1.616972804069519,
      "learning_rate": 6.875949367088608e-06,
      "loss": 0.060027364641427994,
      "num_input_tokens_seen": 2455718,
      "step": 1302,
      "train_runtime": 3542.2769,
      "train_tokens_per_second": 693.26
    },
    {
      "epoch": 1.3164013141268638,
      "grad_norm": 1.8863670825958252,
      "learning_rate": 6.865822784810127e-06,
      "loss": 0.08367598056793213,
      "num_input_tokens_seen": 2457762,
      "step": 1303,
      "train_runtime": 3544.6403,
      "train_tokens_per_second": 693.374
    },
    {
      "epoch": 1.3174121809451604,
      "grad_norm": 1.5819474458694458,
      "learning_rate": 6.855696202531647e-06,
      "loss": 0.07555770874023438,
      "num_input_tokens_seen": 2460088,
      "step": 1304,
      "train_runtime": 3546.9737,
      "train_tokens_per_second": 693.574
    },
    {
      "epoch": 1.318423047763457,
      "grad_norm": 2.4805030822753906,
      "learning_rate": 6.845569620253165e-06,
      "loss": 0.0664629191160202,
      "num_input_tokens_seen": 2461942,
      "step": 1305,
      "train_runtime": 3549.1875,
      "train_tokens_per_second": 693.664
    },
    {
      "epoch": 1.319433914581754,
      "grad_norm": 1.8425042629241943,
      "learning_rate": 6.835443037974684e-06,
      "loss": 0.06088273227214813,
      "num_input_tokens_seen": 2463904,
      "step": 1306,
      "train_runtime": 3551.3947,
      "train_tokens_per_second": 693.785
    },
    {
      "epoch": 1.3204447814000506,
      "grad_norm": 1.5949352979660034,
      "learning_rate": 6.825316455696202e-06,
      "loss": 0.0695766881108284,
      "num_input_tokens_seen": 2465722,
      "step": 1307,
      "train_runtime": 3553.6971,
      "train_tokens_per_second": 693.847
    },
    {
      "epoch": 1.3214556482183473,
      "grad_norm": 1.667921781539917,
      "learning_rate": 6.815189873417722e-06,
      "loss": 0.07278577983379364,
      "num_input_tokens_seen": 2467266,
      "step": 1308,
      "train_runtime": 3555.8491,
      "train_tokens_per_second": 693.861
    },
    {
      "epoch": 1.322466515036644,
      "grad_norm": 1.9061810970306396,
      "learning_rate": 6.805063291139241e-06,
      "loss": 0.0660230740904808,
      "num_input_tokens_seen": 2469430,
      "step": 1309,
      "train_runtime": 3558.1521,
      "train_tokens_per_second": 694.02
    },
    {
      "epoch": 1.3234773818549406,
      "grad_norm": 1.5027179718017578,
      "learning_rate": 6.79493670886076e-06,
      "loss": 0.0616331472992897,
      "num_input_tokens_seen": 2471652,
      "step": 1310,
      "train_runtime": 3560.4394,
      "train_tokens_per_second": 694.199
    },
    {
      "epoch": 1.3244882486732372,
      "grad_norm": 2.2625632286071777,
      "learning_rate": 6.784810126582279e-06,
      "loss": 0.10223156958818436,
      "num_input_tokens_seen": 2473280,
      "step": 1311,
      "train_runtime": 3562.6105,
      "train_tokens_per_second": 694.233
    },
    {
      "epoch": 1.3254991154915339,
      "grad_norm": 1.7246503829956055,
      "learning_rate": 6.774683544303799e-06,
      "loss": 0.08304357528686523,
      "num_input_tokens_seen": 2475394,
      "step": 1312,
      "train_runtime": 3564.8679,
      "train_tokens_per_second": 694.386
    },
    {
      "epoch": 1.3265099823098307,
      "grad_norm": 1.77495276927948,
      "learning_rate": 6.764556962025317e-06,
      "loss": 0.0816446989774704,
      "num_input_tokens_seen": 2477104,
      "step": 1313,
      "train_runtime": 3567.0099,
      "train_tokens_per_second": 694.448
    },
    {
      "epoch": 1.3275208491281274,
      "grad_norm": 1.6261802911758423,
      "learning_rate": 6.754430379746836e-06,
      "loss": 0.08167357742786407,
      "num_input_tokens_seen": 2479042,
      "step": 1314,
      "train_runtime": 3569.2658,
      "train_tokens_per_second": 694.552
    },
    {
      "epoch": 1.328531715946424,
      "grad_norm": 1.394997239112854,
      "learning_rate": 6.744303797468354e-06,
      "loss": 0.049304425716400146,
      "num_input_tokens_seen": 2480676,
      "step": 1315,
      "train_runtime": 3571.4288,
      "train_tokens_per_second": 694.589
    },
    {
      "epoch": 1.3295425827647207,
      "grad_norm": 1.8155115842819214,
      "learning_rate": 6.734177215189874e-06,
      "loss": 0.08127808570861816,
      "num_input_tokens_seen": 2482858,
      "step": 1316,
      "train_runtime": 3573.7371,
      "train_tokens_per_second": 694.751
    },
    {
      "epoch": 1.3305534495830174,
      "grad_norm": 1.8929851055145264,
      "learning_rate": 6.724050632911393e-06,
      "loss": 0.09120990335941315,
      "num_input_tokens_seen": 2484364,
      "step": 1317,
      "train_runtime": 3575.8688,
      "train_tokens_per_second": 694.758
    },
    {
      "epoch": 1.3315643164013142,
      "grad_norm": 2.3512346744537354,
      "learning_rate": 6.7139240506329115e-06,
      "loss": 0.06979864835739136,
      "num_input_tokens_seen": 2486306,
      "step": 1318,
      "train_runtime": 3578.0879,
      "train_tokens_per_second": 694.87
    },
    {
      "epoch": 1.3325751832196109,
      "grad_norm": 1.641466736793518,
      "learning_rate": 6.7037974683544306e-06,
      "loss": 0.07956226170063019,
      "num_input_tokens_seen": 2488390,
      "step": 1319,
      "train_runtime": 3580.3288,
      "train_tokens_per_second": 695.017
    },
    {
      "epoch": 1.3335860500379075,
      "grad_norm": 2.02421498298645,
      "learning_rate": 6.6936708860759505e-06,
      "loss": 0.06039831042289734,
      "num_input_tokens_seen": 2490366,
      "step": 1320,
      "train_runtime": 3582.5373,
      "train_tokens_per_second": 695.14
    },
    {
      "epoch": 1.3345969168562042,
      "grad_norm": 1.466762900352478,
      "learning_rate": 6.683544303797469e-06,
      "loss": 0.0662161186337471,
      "num_input_tokens_seen": 2492082,
      "step": 1321,
      "train_runtime": 3591.314,
      "train_tokens_per_second": 693.919
    },
    {
      "epoch": 1.3356077836745008,
      "grad_norm": 1.5865466594696045,
      "learning_rate": 6.673417721518988e-06,
      "loss": 0.05934914946556091,
      "num_input_tokens_seen": 2493558,
      "step": 1322,
      "train_runtime": 3593.4424,
      "train_tokens_per_second": 693.919
    },
    {
      "epoch": 1.3366186504927975,
      "grad_norm": 1.7458940744400024,
      "learning_rate": 6.663291139240506e-06,
      "loss": 0.06001710519194603,
      "num_input_tokens_seen": 2494988,
      "step": 1323,
      "train_runtime": 3595.5422,
      "train_tokens_per_second": 693.911
    },
    {
      "epoch": 1.3376295173110941,
      "grad_norm": 2.575479745864868,
      "learning_rate": 6.653164556962026e-06,
      "loss": 0.09432195127010345,
      "num_input_tokens_seen": 2496558,
      "step": 1324,
      "train_runtime": 3597.6891,
      "train_tokens_per_second": 693.934
    },
    {
      "epoch": 1.338640384129391,
      "grad_norm": 1.7372066974639893,
      "learning_rate": 6.643037974683545e-06,
      "loss": 0.06360140442848206,
      "num_input_tokens_seen": 2498468,
      "step": 1325,
      "train_runtime": 3599.8724,
      "train_tokens_per_second": 694.043
    },
    {
      "epoch": 1.3396512509476877,
      "grad_norm": 1.9188385009765625,
      "learning_rate": 6.632911392405063e-06,
      "loss": 0.05561995506286621,
      "num_input_tokens_seen": 2500298,
      "step": 1326,
      "train_runtime": 3602.0588,
      "train_tokens_per_second": 694.13
    },
    {
      "epoch": 1.3406621177659843,
      "grad_norm": 1.9363172054290771,
      "learning_rate": 6.622784810126582e-06,
      "loss": 0.08968325704336166,
      "num_input_tokens_seen": 2502174,
      "step": 1327,
      "train_runtime": 3604.2697,
      "train_tokens_per_second": 694.225
    },
    {
      "epoch": 1.341672984584281,
      "grad_norm": 1.533034086227417,
      "learning_rate": 6.612658227848102e-06,
      "loss": 0.06440269201993942,
      "num_input_tokens_seen": 2503952,
      "step": 1328,
      "train_runtime": 3606.4535,
      "train_tokens_per_second": 694.298
    },
    {
      "epoch": 1.3426838514025778,
      "grad_norm": 1.565432071685791,
      "learning_rate": 6.6025316455696206e-06,
      "loss": 0.06231357902288437,
      "num_input_tokens_seen": 2505714,
      "step": 1329,
      "train_runtime": 3608.632,
      "train_tokens_per_second": 694.367
    },
    {
      "epoch": 1.3436947182208745,
      "grad_norm": 1.802560567855835,
      "learning_rate": 6.59240506329114e-06,
      "loss": 0.08845368027687073,
      "num_input_tokens_seen": 2507846,
      "step": 1330,
      "train_runtime": 3610.9019,
      "train_tokens_per_second": 694.521
    },
    {
      "epoch": 1.3447055850391711,
      "grad_norm": 1.2985501289367676,
      "learning_rate": 6.582278481012659e-06,
      "loss": 0.04512377083301544,
      "num_input_tokens_seen": 2510048,
      "step": 1331,
      "train_runtime": 3613.1795,
      "train_tokens_per_second": 694.692
    },
    {
      "epoch": 1.3457164518574678,
      "grad_norm": 1.8165048360824585,
      "learning_rate": 6.572151898734178e-06,
      "loss": 0.06979278475046158,
      "num_input_tokens_seen": 2511812,
      "step": 1332,
      "train_runtime": 3615.3637,
      "train_tokens_per_second": 694.761
    },
    {
      "epoch": 1.3467273186757645,
      "grad_norm": 1.9429415464401245,
      "learning_rate": 6.562025316455697e-06,
      "loss": 0.0932060182094574,
      "num_input_tokens_seen": 2514084,
      "step": 1333,
      "train_runtime": 3617.6711,
      "train_tokens_per_second": 694.945
    },
    {
      "epoch": 1.347738185494061,
      "grad_norm": 1.7487316131591797,
      "learning_rate": 6.551898734177215e-06,
      "loss": 0.05620608851313591,
      "num_input_tokens_seen": 2515838,
      "step": 1334,
      "train_runtime": 3619.8105,
      "train_tokens_per_second": 695.019
    },
    {
      "epoch": 1.3487490523123578,
      "grad_norm": 1.5573967695236206,
      "learning_rate": 6.541772151898734e-06,
      "loss": 0.0643301010131836,
      "num_input_tokens_seen": 2517778,
      "step": 1335,
      "train_runtime": 3622.0203,
      "train_tokens_per_second": 695.131
    },
    {
      "epoch": 1.3497599191306544,
      "grad_norm": 1.9148838520050049,
      "learning_rate": 6.531645569620254e-06,
      "loss": 0.08032730966806412,
      "num_input_tokens_seen": 2519282,
      "step": 1336,
      "train_runtime": 3624.163,
      "train_tokens_per_second": 695.135
    },
    {
      "epoch": 1.3507707859489513,
      "grad_norm": 1.922243356704712,
      "learning_rate": 6.521518987341772e-06,
      "loss": 0.08460190147161484,
      "num_input_tokens_seen": 2520964,
      "step": 1337,
      "train_runtime": 3626.3483,
      "train_tokens_per_second": 695.18
    },
    {
      "epoch": 1.351781652767248,
      "grad_norm": 1.753213882446289,
      "learning_rate": 6.5113924050632915e-06,
      "loss": 0.08410866558551788,
      "num_input_tokens_seen": 2523034,
      "step": 1338,
      "train_runtime": 3628.6016,
      "train_tokens_per_second": 695.319
    },
    {
      "epoch": 1.3527925195855446,
      "grad_norm": 2.136439561843872,
      "learning_rate": 6.5012658227848106e-06,
      "loss": 0.06710030138492584,
      "num_input_tokens_seen": 2525260,
      "step": 1339,
      "train_runtime": 3630.8881,
      "train_tokens_per_second": 695.494
    },
    {
      "epoch": 1.3538033864038412,
      "grad_norm": 1.7677026987075806,
      "learning_rate": 6.49113924050633e-06,
      "loss": 0.07365815341472626,
      "num_input_tokens_seen": 2527338,
      "step": 1340,
      "train_runtime": 3633.1378,
      "train_tokens_per_second": 695.635
    },
    {
      "epoch": 1.354814253222138,
      "grad_norm": 2.025024175643921,
      "learning_rate": 6.481012658227849e-06,
      "loss": 0.08970008045434952,
      "num_input_tokens_seen": 2529494,
      "step": 1341,
      "train_runtime": 3635.3922,
      "train_tokens_per_second": 695.797
    },
    {
      "epoch": 1.3558251200404348,
      "grad_norm": 1.9314972162246704,
      "learning_rate": 6.470886075949368e-06,
      "loss": 0.0951140820980072,
      "num_input_tokens_seen": 2531304,
      "step": 1342,
      "train_runtime": 3637.5845,
      "train_tokens_per_second": 695.875
    },
    {
      "epoch": 1.3568359868587314,
      "grad_norm": 1.8306108713150024,
      "learning_rate": 6.460759493670886e-06,
      "loss": 0.08341541886329651,
      "num_input_tokens_seen": 2533062,
      "step": 1343,
      "train_runtime": 3639.7713,
      "train_tokens_per_second": 695.94
    },
    {
      "epoch": 1.357846853677028,
      "grad_norm": 1.7333952188491821,
      "learning_rate": 6.450632911392406e-06,
      "loss": 0.0747184231877327,
      "num_input_tokens_seen": 2535360,
      "step": 1344,
      "train_runtime": 3642.1202,
      "train_tokens_per_second": 696.122
    },
    {
      "epoch": 1.3588577204953247,
      "grad_norm": 1.6327425241470337,
      "learning_rate": 6.440506329113925e-06,
      "loss": 0.07438857108354568,
      "num_input_tokens_seen": 2537128,
      "step": 1345,
      "train_runtime": 3644.3497,
      "train_tokens_per_second": 696.181
    },
    {
      "epoch": 1.3598685873136214,
      "grad_norm": 1.8790346384048462,
      "learning_rate": 6.430379746835443e-06,
      "loss": 0.08983723819255829,
      "num_input_tokens_seen": 2539182,
      "step": 1346,
      "train_runtime": 3646.6345,
      "train_tokens_per_second": 696.308
    },
    {
      "epoch": 1.360879454131918,
      "grad_norm": 1.5209429264068604,
      "learning_rate": 6.420253164556962e-06,
      "loss": 0.05216553434729576,
      "num_input_tokens_seen": 2541200,
      "step": 1347,
      "train_runtime": 3648.9454,
      "train_tokens_per_second": 696.42
    },
    {
      "epoch": 1.3618903209502147,
      "grad_norm": 1.7687582969665527,
      "learning_rate": 6.410126582278482e-06,
      "loss": 0.06282483041286469,
      "num_input_tokens_seen": 2543378,
      "step": 1348,
      "train_runtime": 3651.2595,
      "train_tokens_per_second": 696.576
    },
    {
      "epoch": 1.3629011877685115,
      "grad_norm": 1.3695815801620483,
      "learning_rate": 6.4000000000000006e-06,
      "loss": 0.060868073254823685,
      "num_input_tokens_seen": 2545350,
      "step": 1349,
      "train_runtime": 3653.5067,
      "train_tokens_per_second": 696.687
    },
    {
      "epoch": 1.3639120545868082,
      "grad_norm": 1.8668866157531738,
      "learning_rate": 6.38987341772152e-06,
      "loss": 0.07853592187166214,
      "num_input_tokens_seen": 2547356,
      "step": 1350,
      "train_runtime": 3655.8199,
      "train_tokens_per_second": 696.795
    },
    {
      "epoch": 1.3649229214051048,
      "grad_norm": 1.8764808177947998,
      "learning_rate": 6.379746835443038e-06,
      "loss": 0.06392718851566315,
      "num_input_tokens_seen": 2549126,
      "step": 1351,
      "train_runtime": 3664.4397,
      "train_tokens_per_second": 695.639
    },
    {
      "epoch": 1.3659337882234015,
      "grad_norm": 1.7786204814910889,
      "learning_rate": 6.369620253164558e-06,
      "loss": 0.07574222981929779,
      "num_input_tokens_seen": 2551066,
      "step": 1352,
      "train_runtime": 3666.6883,
      "train_tokens_per_second": 695.741
    },
    {
      "epoch": 1.3669446550416984,
      "grad_norm": 1.7288459539413452,
      "learning_rate": 6.359493670886077e-06,
      "loss": 0.08671493828296661,
      "num_input_tokens_seen": 2553038,
      "step": 1353,
      "train_runtime": 3668.9424,
      "train_tokens_per_second": 695.851
    },
    {
      "epoch": 1.367955521859995,
      "grad_norm": 1.5520856380462646,
      "learning_rate": 6.349367088607595e-06,
      "loss": 0.06124834716320038,
      "num_input_tokens_seen": 2555022,
      "step": 1354,
      "train_runtime": 3671.1643,
      "train_tokens_per_second": 695.97
    },
    {
      "epoch": 1.3689663886782917,
      "grad_norm": 1.3783342838287354,
      "learning_rate": 6.339240506329114e-06,
      "loss": 0.041727203875780106,
      "num_input_tokens_seen": 2556426,
      "step": 1355,
      "train_runtime": 3673.2581,
      "train_tokens_per_second": 695.956
    },
    {
      "epoch": 1.3699772554965883,
      "grad_norm": 1.722110390663147,
      "learning_rate": 6.329113924050634e-06,
      "loss": 0.0781174823641777,
      "num_input_tokens_seen": 2558122,
      "step": 1356,
      "train_runtime": 3675.4245,
      "train_tokens_per_second": 696.007
    },
    {
      "epoch": 1.370988122314885,
      "grad_norm": 1.9059115648269653,
      "learning_rate": 6.318987341772152e-06,
      "loss": 0.0744543969631195,
      "num_input_tokens_seen": 2560308,
      "step": 1357,
      "train_runtime": 3677.6823,
      "train_tokens_per_second": 696.174
    },
    {
      "epoch": 1.3719989891331816,
      "grad_norm": 1.7189818620681763,
      "learning_rate": 6.3088607594936715e-06,
      "loss": 0.06220090016722679,
      "num_input_tokens_seen": 2562338,
      "step": 1358,
      "train_runtime": 3679.8975,
      "train_tokens_per_second": 696.307
    },
    {
      "epoch": 1.3730098559514783,
      "grad_norm": 2.0385119915008545,
      "learning_rate": 6.29873417721519e-06,
      "loss": 0.07487086951732635,
      "num_input_tokens_seen": 2563842,
      "step": 1359,
      "train_runtime": 3682.0149,
      "train_tokens_per_second": 696.315
    },
    {
      "epoch": 1.3740207227697752,
      "grad_norm": 1.7844659090042114,
      "learning_rate": 6.28860759493671e-06,
      "loss": 0.06827859580516815,
      "num_input_tokens_seen": 2565870,
      "step": 1360,
      "train_runtime": 3684.262,
      "train_tokens_per_second": 696.441
    },
    {
      "epoch": 1.3750315895880718,
      "grad_norm": 1.8249586820602417,
      "learning_rate": 6.278481012658229e-06,
      "loss": 0.07555392384529114,
      "num_input_tokens_seen": 2567614,
      "step": 1361,
      "train_runtime": 3686.4585,
      "train_tokens_per_second": 696.499
    },
    {
      "epoch": 1.3760424564063685,
      "grad_norm": 1.5155221223831177,
      "learning_rate": 6.268354430379747e-06,
      "loss": 0.06685218214988708,
      "num_input_tokens_seen": 2569486,
      "step": 1362,
      "train_runtime": 3688.6845,
      "train_tokens_per_second": 696.586
    },
    {
      "epoch": 1.3770533232246651,
      "grad_norm": 1.4058446884155273,
      "learning_rate": 6.258227848101266e-06,
      "loss": 0.054429881274700165,
      "num_input_tokens_seen": 2571478,
      "step": 1363,
      "train_runtime": 3690.9048,
      "train_tokens_per_second": 696.707
    },
    {
      "epoch": 1.3780641900429618,
      "grad_norm": 1.9558351039886475,
      "learning_rate": 6.248101265822786e-06,
      "loss": 0.07545759528875351,
      "num_input_tokens_seen": 2573162,
      "step": 1364,
      "train_runtime": 3693.1254,
      "train_tokens_per_second": 696.744
    },
    {
      "epoch": 1.3790750568612586,
      "grad_norm": 1.8815573453903198,
      "learning_rate": 6.237974683544304e-06,
      "loss": 0.05953256040811539,
      "num_input_tokens_seen": 2574666,
      "step": 1365,
      "train_runtime": 3695.2583,
      "train_tokens_per_second": 696.749
    },
    {
      "epoch": 1.3800859236795553,
      "grad_norm": 1.5381945371627808,
      "learning_rate": 6.227848101265823e-06,
      "loss": 0.06057102233171463,
      "num_input_tokens_seen": 2576094,
      "step": 1366,
      "train_runtime": 3697.36,
      "train_tokens_per_second": 696.739
    },
    {
      "epoch": 1.381096790497852,
      "grad_norm": 2.591580867767334,
      "learning_rate": 6.2177215189873415e-06,
      "loss": 0.05954180657863617,
      "num_input_tokens_seen": 2577802,
      "step": 1367,
      "train_runtime": 3699.5495,
      "train_tokens_per_second": 696.788
    },
    {
      "epoch": 1.3821076573161486,
      "grad_norm": 1.715431809425354,
      "learning_rate": 6.2075949367088615e-06,
      "loss": 0.07338304817676544,
      "num_input_tokens_seen": 2580078,
      "step": 1368,
      "train_runtime": 3701.8612,
      "train_tokens_per_second": 696.968
    },
    {
      "epoch": 1.3831185241344452,
      "grad_norm": 1.794058918952942,
      "learning_rate": 6.1974683544303805e-06,
      "loss": 0.06370040029287338,
      "num_input_tokens_seen": 2581822,
      "step": 1369,
      "train_runtime": 3704.0847,
      "train_tokens_per_second": 697.02
    },
    {
      "epoch": 1.384129390952742,
      "grad_norm": 1.4786248207092285,
      "learning_rate": 6.187341772151899e-06,
      "loss": 0.06929203867912292,
      "num_input_tokens_seen": 2584068,
      "step": 1370,
      "train_runtime": 3706.4385,
      "train_tokens_per_second": 697.184
    },
    {
      "epoch": 1.3851402577710386,
      "grad_norm": 1.4867762327194214,
      "learning_rate": 6.177215189873418e-06,
      "loss": 0.04443085193634033,
      "num_input_tokens_seen": 2585860,
      "step": 1371,
      "train_runtime": 3708.6663,
      "train_tokens_per_second": 697.248
    },
    {
      "epoch": 1.3861511245893354,
      "grad_norm": 1.640121579170227,
      "learning_rate": 6.167088607594938e-06,
      "loss": 0.07376250624656677,
      "num_input_tokens_seen": 2587634,
      "step": 1372,
      "train_runtime": 3710.8687,
      "train_tokens_per_second": 697.312
    },
    {
      "epoch": 1.387161991407632,
      "grad_norm": 1.3839681148529053,
      "learning_rate": 6.156962025316456e-06,
      "loss": 0.0596800222992897,
      "num_input_tokens_seen": 2589498,
      "step": 1373,
      "train_runtime": 3713.1099,
      "train_tokens_per_second": 697.393
    },
    {
      "epoch": 1.3881728582259287,
      "grad_norm": 1.494248867034912,
      "learning_rate": 6.146835443037975e-06,
      "loss": 0.05772831290960312,
      "num_input_tokens_seen": 2591676,
      "step": 1374,
      "train_runtime": 3715.3991,
      "train_tokens_per_second": 697.55
    },
    {
      "epoch": 1.3891837250442254,
      "grad_norm": 1.8152834177017212,
      "learning_rate": 6.136708860759494e-06,
      "loss": 0.05584399402141571,
      "num_input_tokens_seen": 2593544,
      "step": 1375,
      "train_runtime": 3717.6242,
      "train_tokens_per_second": 697.635
    },
    {
      "epoch": 1.390194591862522,
      "grad_norm": 1.642403483390808,
      "learning_rate": 6.126582278481013e-06,
      "loss": 0.06559552252292633,
      "num_input_tokens_seen": 2595692,
      "step": 1376,
      "train_runtime": 3719.913,
      "train_tokens_per_second": 697.783
    },
    {
      "epoch": 1.391205458680819,
      "grad_norm": 2.095770835876465,
      "learning_rate": 6.116455696202532e-06,
      "loss": 0.10935665667057037,
      "num_input_tokens_seen": 2597534,
      "step": 1377,
      "train_runtime": 3722.1085,
      "train_tokens_per_second": 697.866
    },
    {
      "epoch": 1.3922163254991156,
      "grad_norm": 1.8721085786819458,
      "learning_rate": 6.106329113924051e-06,
      "loss": 0.08921398222446442,
      "num_input_tokens_seen": 2600018,
      "step": 1378,
      "train_runtime": 3724.4617,
      "train_tokens_per_second": 698.092
    },
    {
      "epoch": 1.3932271923174122,
      "grad_norm": 1.6955461502075195,
      "learning_rate": 6.09620253164557e-06,
      "loss": 0.06785810738801956,
      "num_input_tokens_seen": 2602052,
      "step": 1379,
      "train_runtime": 3726.6847,
      "train_tokens_per_second": 698.222
    },
    {
      "epoch": 1.3942380591357089,
      "grad_norm": 1.7655616998672485,
      "learning_rate": 6.08607594936709e-06,
      "loss": 0.06537257879972458,
      "num_input_tokens_seen": 2603758,
      "step": 1380,
      "train_runtime": 3728.8324,
      "train_tokens_per_second": 698.277
    },
    {
      "epoch": 1.3952489259540055,
      "grad_norm": 1.4429453611373901,
      "learning_rate": 6.075949367088608e-06,
      "loss": 0.052360452711582184,
      "num_input_tokens_seen": 2605644,
      "step": 1381,
      "train_runtime": 3737.6123,
      "train_tokens_per_second": 697.141
    },
    {
      "epoch": 1.3962597927723022,
      "grad_norm": 1.580642580986023,
      "learning_rate": 6.065822784810127e-06,
      "loss": 0.0596926175057888,
      "num_input_tokens_seen": 2607354,
      "step": 1382,
      "train_runtime": 3739.8068,
      "train_tokens_per_second": 697.189
    },
    {
      "epoch": 1.3972706595905988,
      "grad_norm": 1.4305837154388428,
      "learning_rate": 6.055696202531646e-06,
      "loss": 0.050742533057928085,
      "num_input_tokens_seen": 2609280,
      "step": 1383,
      "train_runtime": 3742.0436,
      "train_tokens_per_second": 697.287
    },
    {
      "epoch": 1.3982815264088957,
      "grad_norm": 2.0289719104766846,
      "learning_rate": 6.045569620253165e-06,
      "loss": 0.10896015167236328,
      "num_input_tokens_seen": 2611402,
      "step": 1384,
      "train_runtime": 3744.3117,
      "train_tokens_per_second": 697.432
    },
    {
      "epoch": 1.3992923932271923,
      "grad_norm": 2.4022905826568604,
      "learning_rate": 6.035443037974684e-06,
      "loss": 0.07937447726726532,
      "num_input_tokens_seen": 2612776,
      "step": 1385,
      "train_runtime": 3746.3747,
      "train_tokens_per_second": 697.414
    },
    {
      "epoch": 1.400303260045489,
      "grad_norm": 1.754387617111206,
      "learning_rate": 6.025316455696203e-06,
      "loss": 0.08007083833217621,
      "num_input_tokens_seen": 2614562,
      "step": 1386,
      "train_runtime": 3748.5569,
      "train_tokens_per_second": 697.485
    },
    {
      "epoch": 1.400303260045489,
      "eval_loss": 0.3937505781650543,
      "eval_runtime": 52.57,
      "eval_samples_per_second": 16.721,
      "eval_steps_per_second": 8.37,
      "num_input_tokens_seen": 2614562,
      "step": 1386
    },
    {
      "epoch": 1.4013141268637856,
      "grad_norm": 1.4994029998779297,
      "learning_rate": 6.0151898734177215e-06,
      "loss": 0.05978060141205788,
      "num_input_tokens_seen": 2616790,
      "step": 1387,
      "train_runtime": 3803.4318,
      "train_tokens_per_second": 688.008
    },
    {
      "epoch": 1.4023249936820825,
      "grad_norm": 1.4958637952804565,
      "learning_rate": 6.0050632911392415e-06,
      "loss": 0.056719157844781876,
      "num_input_tokens_seen": 2618498,
      "step": 1388,
      "train_runtime": 3805.6325,
      "train_tokens_per_second": 688.059
    },
    {
      "epoch": 1.4033358605003792,
      "grad_norm": 1.4207665920257568,
      "learning_rate": 5.9949367088607605e-06,
      "loss": 0.06697021424770355,
      "num_input_tokens_seen": 2620730,
      "step": 1389,
      "train_runtime": 3807.9158,
      "train_tokens_per_second": 688.232
    },
    {
      "epoch": 1.4043467273186758,
      "grad_norm": 1.6702016592025757,
      "learning_rate": 5.984810126582279e-06,
      "loss": 0.06788940727710724,
      "num_input_tokens_seen": 2622210,
      "step": 1390,
      "train_runtime": 3810.0515,
      "train_tokens_per_second": 688.235
    },
    {
      "epoch": 1.4053575941369725,
      "grad_norm": 1.6265910863876343,
      "learning_rate": 5.974683544303798e-06,
      "loss": 0.06827136129140854,
      "num_input_tokens_seen": 2623708,
      "step": 1391,
      "train_runtime": 3812.1554,
      "train_tokens_per_second": 688.248
    },
    {
      "epoch": 1.4063684609552691,
      "grad_norm": 1.6982901096343994,
      "learning_rate": 5.964556962025318e-06,
      "loss": 0.08324596285820007,
      "num_input_tokens_seen": 2625760,
      "step": 1392,
      "train_runtime": 3814.4253,
      "train_tokens_per_second": 688.376
    },
    {
      "epoch": 1.4073793277735658,
      "grad_norm": 1.9044620990753174,
      "learning_rate": 5.954430379746836e-06,
      "loss": 0.07309594750404358,
      "num_input_tokens_seen": 2627750,
      "step": 1393,
      "train_runtime": 3816.6739,
      "train_tokens_per_second": 688.492
    },
    {
      "epoch": 1.4083901945918624,
      "grad_norm": 1.7940040826797485,
      "learning_rate": 5.944303797468355e-06,
      "loss": 0.07906541228294373,
      "num_input_tokens_seen": 2630024,
      "step": 1394,
      "train_runtime": 3819.0094,
      "train_tokens_per_second": 688.667
    },
    {
      "epoch": 1.409401061410159,
      "grad_norm": 1.5503826141357422,
      "learning_rate": 5.934177215189873e-06,
      "loss": 0.055062007158994675,
      "num_input_tokens_seen": 2631576,
      "step": 1395,
      "train_runtime": 3821.1292,
      "train_tokens_per_second": 688.691
    },
    {
      "epoch": 1.410411928228456,
      "grad_norm": 1.3000690937042236,
      "learning_rate": 5.924050632911393e-06,
      "loss": 0.06002321466803551,
      "num_input_tokens_seen": 2633758,
      "step": 1396,
      "train_runtime": 3823.3741,
      "train_tokens_per_second": 688.857
    },
    {
      "epoch": 1.4114227950467526,
      "grad_norm": 1.722181797027588,
      "learning_rate": 5.913924050632912e-06,
      "loss": 0.06877709925174713,
      "num_input_tokens_seen": 2635928,
      "step": 1397,
      "train_runtime": 3825.6248,
      "train_tokens_per_second": 689.019
    },
    {
      "epoch": 1.4124336618650493,
      "grad_norm": 1.6934517621994019,
      "learning_rate": 5.903797468354431e-06,
      "loss": 0.06084431707859039,
      "num_input_tokens_seen": 2637284,
      "step": 1398,
      "train_runtime": 3827.7022,
      "train_tokens_per_second": 688.999
    },
    {
      "epoch": 1.413444528683346,
      "grad_norm": 2.1336758136749268,
      "learning_rate": 5.89367088607595e-06,
      "loss": 0.08002732694149017,
      "num_input_tokens_seen": 2639002,
      "step": 1399,
      "train_runtime": 3829.9005,
      "train_tokens_per_second": 689.052
    },
    {
      "epoch": 1.4144553955016428,
      "grad_norm": 1.7232083082199097,
      "learning_rate": 5.88354430379747e-06,
      "loss": 0.06252460181713104,
      "num_input_tokens_seen": 2640572,
      "step": 1400,
      "train_runtime": 3832.0309,
      "train_tokens_per_second": 689.079
    },
    {
      "epoch": 1.4154662623199394,
      "grad_norm": 1.6357442140579224,
      "learning_rate": 5.873417721518988e-06,
      "loss": 0.07521212100982666,
      "num_input_tokens_seen": 2642784,
      "step": 1401,
      "train_runtime": 3834.3821,
      "train_tokens_per_second": 689.233
    },
    {
      "epoch": 1.416477129138236,
      "grad_norm": 1.3101296424865723,
      "learning_rate": 5.863291139240507e-06,
      "loss": 0.04918498545885086,
      "num_input_tokens_seen": 2645018,
      "step": 1402,
      "train_runtime": 3836.6691,
      "train_tokens_per_second": 689.405
    },
    {
      "epoch": 1.4174879959565327,
      "grad_norm": 2.26466703414917,
      "learning_rate": 5.853164556962025e-06,
      "loss": 0.06629481911659241,
      "num_input_tokens_seen": 2646966,
      "step": 1403,
      "train_runtime": 3838.9682,
      "train_tokens_per_second": 689.499
    },
    {
      "epoch": 1.4184988627748294,
      "grad_norm": 1.6053820848464966,
      "learning_rate": 5.843037974683545e-06,
      "loss": 0.0732828676700592,
      "num_input_tokens_seen": 2649610,
      "step": 1404,
      "train_runtime": 3841.3732,
      "train_tokens_per_second": 689.756
    },
    {
      "epoch": 1.419509729593126,
      "grad_norm": 1.8432995080947876,
      "learning_rate": 5.832911392405064e-06,
      "loss": 0.07649111747741699,
      "num_input_tokens_seen": 2651580,
      "step": 1405,
      "train_runtime": 3843.6649,
      "train_tokens_per_second": 689.857
    },
    {
      "epoch": 1.4205205964114227,
      "grad_norm": 1.5591950416564941,
      "learning_rate": 5.8227848101265824e-06,
      "loss": 0.06862519681453705,
      "num_input_tokens_seen": 2653682,
      "step": 1406,
      "train_runtime": 3845.9135,
      "train_tokens_per_second": 690.0
    },
    {
      "epoch": 1.4215314632297193,
      "grad_norm": 1.5466194152832031,
      "learning_rate": 5.8126582278481015e-06,
      "loss": 0.05616164579987526,
      "num_input_tokens_seen": 2655742,
      "step": 1407,
      "train_runtime": 3848.1953,
      "train_tokens_per_second": 690.127
    },
    {
      "epoch": 1.4225423300480162,
      "grad_norm": 1.8269500732421875,
      "learning_rate": 5.8025316455696214e-06,
      "loss": 0.0843963474035263,
      "num_input_tokens_seen": 2658112,
      "step": 1408,
      "train_runtime": 3850.5147,
      "train_tokens_per_second": 690.326
    },
    {
      "epoch": 1.4235531968663129,
      "grad_norm": 1.3943942785263062,
      "learning_rate": 5.79240506329114e-06,
      "loss": 0.04282160475850105,
      "num_input_tokens_seen": 2659640,
      "step": 1409,
      "train_runtime": 3852.6641,
      "train_tokens_per_second": 690.338
    },
    {
      "epoch": 1.4245640636846095,
      "grad_norm": 1.5564720630645752,
      "learning_rate": 5.782278481012659e-06,
      "loss": 0.06570545583963394,
      "num_input_tokens_seen": 2661468,
      "step": 1410,
      "train_runtime": 3854.9127,
      "train_tokens_per_second": 690.409
    },
    {
      "epoch": 1.4255749305029062,
      "grad_norm": 1.5751513242721558,
      "learning_rate": 5.772151898734177e-06,
      "loss": 0.059860941022634506,
      "num_input_tokens_seen": 2663680,
      "step": 1411,
      "train_runtime": 3863.6955,
      "train_tokens_per_second": 689.413
    },
    {
      "epoch": 1.426585797321203,
      "grad_norm": 1.4742987155914307,
      "learning_rate": 5.762025316455697e-06,
      "loss": 0.07407383620738983,
      "num_input_tokens_seen": 2666480,
      "step": 1412,
      "train_runtime": 3866.1418,
      "train_tokens_per_second": 689.701
    },
    {
      "epoch": 1.4275966641394997,
      "grad_norm": 1.8826227188110352,
      "learning_rate": 5.751898734177216e-06,
      "loss": 0.08610516041517258,
      "num_input_tokens_seen": 2668490,
      "step": 1413,
      "train_runtime": 3868.3845,
      "train_tokens_per_second": 689.82
    },
    {
      "epoch": 1.4286075309577964,
      "grad_norm": 2.008103847503662,
      "learning_rate": 5.741772151898734e-06,
      "loss": 0.0816655233502388,
      "num_input_tokens_seen": 2670236,
      "step": 1414,
      "train_runtime": 3870.5652,
      "train_tokens_per_second": 689.883
    },
    {
      "epoch": 1.429618397776093,
      "grad_norm": 1.6061736345291138,
      "learning_rate": 5.731645569620253e-06,
      "loss": 0.08215966075658798,
      "num_input_tokens_seen": 2672264,
      "step": 1415,
      "train_runtime": 3872.8982,
      "train_tokens_per_second": 689.991
    },
    {
      "epoch": 1.4306292645943897,
      "grad_norm": 1.6165075302124023,
      "learning_rate": 5.721518987341773e-06,
      "loss": 0.06657277792692184,
      "num_input_tokens_seen": 2674302,
      "step": 1416,
      "train_runtime": 3875.2026,
      "train_tokens_per_second": 690.106
    },
    {
      "epoch": 1.4316401314126863,
      "grad_norm": 1.482926845550537,
      "learning_rate": 5.7113924050632915e-06,
      "loss": 0.0545504130423069,
      "num_input_tokens_seen": 2676036,
      "step": 1417,
      "train_runtime": 3877.386,
      "train_tokens_per_second": 690.165
    },
    {
      "epoch": 1.432650998230983,
      "grad_norm": 1.6906814575195312,
      "learning_rate": 5.701265822784811e-06,
      "loss": 0.09162657707929611,
      "num_input_tokens_seen": 2678014,
      "step": 1418,
      "train_runtime": 3879.8251,
      "train_tokens_per_second": 690.241
    },
    {
      "epoch": 1.4336618650492798,
      "grad_norm": 1.653327465057373,
      "learning_rate": 5.69113924050633e-06,
      "loss": 0.061128970235586166,
      "num_input_tokens_seen": 2679970,
      "step": 1419,
      "train_runtime": 3882.0345,
      "train_tokens_per_second": 690.352
    },
    {
      "epoch": 1.4346727318675765,
      "grad_norm": 1.8733450174331665,
      "learning_rate": 5.681012658227849e-06,
      "loss": 0.08715441077947617,
      "num_input_tokens_seen": 2681874,
      "step": 1420,
      "train_runtime": 3884.3075,
      "train_tokens_per_second": 690.438
    },
    {
      "epoch": 1.4356835986858731,
      "grad_norm": 1.596731185913086,
      "learning_rate": 5.670886075949368e-06,
      "loss": 0.058932047337293625,
      "num_input_tokens_seen": 2683802,
      "step": 1421,
      "train_runtime": 3886.5173,
      "train_tokens_per_second": 690.542
    },
    {
      "epoch": 1.4366944655041698,
      "grad_norm": 2.342883586883545,
      "learning_rate": 5.660759493670886e-06,
      "loss": 0.08989928662776947,
      "num_input_tokens_seen": 2685204,
      "step": 1422,
      "train_runtime": 3888.6005,
      "train_tokens_per_second": 690.532
    },
    {
      "epoch": 1.4377053323224664,
      "grad_norm": 1.5533877611160278,
      "learning_rate": 5.650632911392405e-06,
      "loss": 0.08014171570539474,
      "num_input_tokens_seen": 2687734,
      "step": 1423,
      "train_runtime": 3891.0027,
      "train_tokens_per_second": 690.756
    },
    {
      "epoch": 1.4387161991407633,
      "grad_norm": 1.75212824344635,
      "learning_rate": 5.640506329113925e-06,
      "loss": 0.074334055185318,
      "num_input_tokens_seen": 2689396,
      "step": 1424,
      "train_runtime": 3893.2317,
      "train_tokens_per_second": 690.788
    },
    {
      "epoch": 1.43972706595906,
      "grad_norm": 1.780698537826538,
      "learning_rate": 5.630379746835443e-06,
      "loss": 0.09270895272493362,
      "num_input_tokens_seen": 2691060,
      "step": 1425,
      "train_runtime": 3895.4469,
      "train_tokens_per_second": 690.822
    },
    {
      "epoch": 1.4407379327773566,
      "grad_norm": 1.399304986000061,
      "learning_rate": 5.6202531645569624e-06,
      "loss": 0.04954908788204193,
      "num_input_tokens_seen": 2692586,
      "step": 1426,
      "train_runtime": 3897.6832,
      "train_tokens_per_second": 690.817
    },
    {
      "epoch": 1.4417487995956533,
      "grad_norm": 1.6596004962921143,
      "learning_rate": 5.6101265822784815e-06,
      "loss": 0.08171042054891586,
      "num_input_tokens_seen": 2694948,
      "step": 1427,
      "train_runtime": 3900.0356,
      "train_tokens_per_second": 691.006
    },
    {
      "epoch": 1.44275966641395,
      "grad_norm": 1.5269056558609009,
      "learning_rate": 5.600000000000001e-06,
      "loss": 0.07041347026824951,
      "num_input_tokens_seen": 2697050,
      "step": 1428,
      "train_runtime": 3902.2756,
      "train_tokens_per_second": 691.148
    },
    {
      "epoch": 1.4437705332322466,
      "grad_norm": 1.4953557252883911,
      "learning_rate": 5.58987341772152e-06,
      "loss": 0.06814124435186386,
      "num_input_tokens_seen": 2699474,
      "step": 1429,
      "train_runtime": 3904.6004,
      "train_tokens_per_second": 691.357
    },
    {
      "epoch": 1.4447814000505432,
      "grad_norm": 2.0349628925323486,
      "learning_rate": 5.579746835443039e-06,
      "loss": 0.08178409934043884,
      "num_input_tokens_seen": 2701090,
      "step": 1430,
      "train_runtime": 3906.7502,
      "train_tokens_per_second": 691.391
    },
    {
      "epoch": 1.44579226686884,
      "grad_norm": 1.6497156620025635,
      "learning_rate": 5.569620253164557e-06,
      "loss": 0.06791052967309952,
      "num_input_tokens_seen": 2702828,
      "step": 1431,
      "train_runtime": 3908.9029,
      "train_tokens_per_second": 691.454
    },
    {
      "epoch": 1.4468031336871368,
      "grad_norm": 1.7280336618423462,
      "learning_rate": 5.559493670886077e-06,
      "loss": 0.07457365095615387,
      "num_input_tokens_seen": 2704850,
      "step": 1432,
      "train_runtime": 3911.1117,
      "train_tokens_per_second": 691.581
    },
    {
      "epoch": 1.4478140005054334,
      "grad_norm": 1.5326509475708008,
      "learning_rate": 5.549367088607596e-06,
      "loss": 0.050631217658519745,
      "num_input_tokens_seen": 2706554,
      "step": 1433,
      "train_runtime": 3913.2965,
      "train_tokens_per_second": 691.63
    },
    {
      "epoch": 1.44882486732373,
      "grad_norm": 2.0489110946655273,
      "learning_rate": 5.539240506329114e-06,
      "loss": 0.08721400797367096,
      "num_input_tokens_seen": 2708422,
      "step": 1434,
      "train_runtime": 3915.5314,
      "train_tokens_per_second": 691.713
    },
    {
      "epoch": 1.4498357341420267,
      "grad_norm": 2.0255892276763916,
      "learning_rate": 5.529113924050633e-06,
      "loss": 0.08184416592121124,
      "num_input_tokens_seen": 2710138,
      "step": 1435,
      "train_runtime": 3917.7093,
      "train_tokens_per_second": 691.766
    },
    {
      "epoch": 1.4508466009603236,
      "grad_norm": 1.8512778282165527,
      "learning_rate": 5.518987341772153e-06,
      "loss": 0.10774432867765427,
      "num_input_tokens_seen": 2712362,
      "step": 1436,
      "train_runtime": 3920.0078,
      "train_tokens_per_second": 691.928
    },
    {
      "epoch": 1.4518574677786202,
      "grad_norm": 1.8451319932937622,
      "learning_rate": 5.5088607594936715e-06,
      "loss": 0.09010189026594162,
      "num_input_tokens_seen": 2714324,
      "step": 1437,
      "train_runtime": 3922.2595,
      "train_tokens_per_second": 692.031
    },
    {
      "epoch": 1.4528683345969169,
      "grad_norm": 1.6371424198150635,
      "learning_rate": 5.498734177215191e-06,
      "loss": 0.05427632853388786,
      "num_input_tokens_seen": 2716354,
      "step": 1438,
      "train_runtime": 3924.5002,
      "train_tokens_per_second": 692.153
    },
    {
      "epoch": 1.4538792014152135,
      "grad_norm": 1.239633321762085,
      "learning_rate": 5.488607594936709e-06,
      "loss": 0.043502919375896454,
      "num_input_tokens_seen": 2718476,
      "step": 1439,
      "train_runtime": 3926.7668,
      "train_tokens_per_second": 692.294
    },
    {
      "epoch": 1.4548900682335102,
      "grad_norm": 1.7125214338302612,
      "learning_rate": 5.478481012658229e-06,
      "loss": 0.060112956911325455,
      "num_input_tokens_seen": 2720324,
      "step": 1440,
      "train_runtime": 3928.9493,
      "train_tokens_per_second": 692.38
    },
    {
      "epoch": 1.4559009350518068,
      "grad_norm": 1.7435688972473145,
      "learning_rate": 5.468354430379748e-06,
      "loss": 0.06201600283384323,
      "num_input_tokens_seen": 2721974,
      "step": 1441,
      "train_runtime": 3937.5219,
      "train_tokens_per_second": 691.291
    },
    {
      "epoch": 1.4569118018701035,
      "grad_norm": 1.6306309700012207,
      "learning_rate": 5.458227848101266e-06,
      "loss": 0.07703009992837906,
      "num_input_tokens_seen": 2724214,
      "step": 1442,
      "train_runtime": 3939.8263,
      "train_tokens_per_second": 691.455
    },
    {
      "epoch": 1.4579226686884004,
      "grad_norm": 2.3203349113464355,
      "learning_rate": 5.448101265822785e-06,
      "loss": 0.11020687222480774,
      "num_input_tokens_seen": 2725932,
      "step": 1443,
      "train_runtime": 3941.9816,
      "train_tokens_per_second": 691.513
    },
    {
      "epoch": 1.458933535506697,
      "grad_norm": 1.8823951482772827,
      "learning_rate": 5.437974683544303e-06,
      "loss": 0.06746494770050049,
      "num_input_tokens_seen": 2727498,
      "step": 1444,
      "train_runtime": 3944.1288,
      "train_tokens_per_second": 691.534
    },
    {
      "epoch": 1.4599444023249937,
      "grad_norm": 1.5358587503433228,
      "learning_rate": 5.427848101265823e-06,
      "loss": 0.05518173426389694,
      "num_input_tokens_seen": 2729276,
      "step": 1445,
      "train_runtime": 3946.2782,
      "train_tokens_per_second": 691.608
    },
    {
      "epoch": 1.4609552691432903,
      "grad_norm": 1.3288687467575073,
      "learning_rate": 5.417721518987342e-06,
      "loss": 0.04659959301352501,
      "num_input_tokens_seen": 2731094,
      "step": 1446,
      "train_runtime": 3948.4866,
      "train_tokens_per_second": 691.681
    },
    {
      "epoch": 1.4619661359615872,
      "grad_norm": 1.4637774229049683,
      "learning_rate": 5.407594936708861e-06,
      "loss": 0.054771788418293,
      "num_input_tokens_seen": 2732632,
      "step": 1447,
      "train_runtime": 3950.6033,
      "train_tokens_per_second": 691.7
    },
    {
      "epoch": 1.4629770027798839,
      "grad_norm": 1.5719244480133057,
      "learning_rate": 5.39746835443038e-06,
      "loss": 0.07881119847297668,
      "num_input_tokens_seen": 2734660,
      "step": 1448,
      "train_runtime": 3952.8335,
      "train_tokens_per_second": 691.823
    },
    {
      "epoch": 1.4639878695981805,
      "grad_norm": 1.8760321140289307,
      "learning_rate": 5.3873417721519e-06,
      "loss": 0.08335188776254654,
      "num_input_tokens_seen": 2736088,
      "step": 1449,
      "train_runtime": 3954.9171,
      "train_tokens_per_second": 691.819
    },
    {
      "epoch": 1.4649987364164772,
      "grad_norm": 1.4611618518829346,
      "learning_rate": 5.377215189873418e-06,
      "loss": 0.0509427934885025,
      "num_input_tokens_seen": 2737516,
      "step": 1450,
      "train_runtime": 3956.974,
      "train_tokens_per_second": 691.821
    },
    {
      "epoch": 1.4660096032347738,
      "grad_norm": 1.4783076047897339,
      "learning_rate": 5.367088607594937e-06,
      "loss": 0.06264941394329071,
      "num_input_tokens_seen": 2740122,
      "step": 1451,
      "train_runtime": 3959.313,
      "train_tokens_per_second": 692.07
    },
    {
      "epoch": 1.4670204700530705,
      "grad_norm": 1.5367320775985718,
      "learning_rate": 5.356962025316455e-06,
      "loss": 0.06179020553827286,
      "num_input_tokens_seen": 2741914,
      "step": 1452,
      "train_runtime": 3961.5109,
      "train_tokens_per_second": 692.138
    },
    {
      "epoch": 1.468031336871367,
      "grad_norm": 1.5037331581115723,
      "learning_rate": 5.346835443037975e-06,
      "loss": 0.05475204810500145,
      "num_input_tokens_seen": 2743786,
      "step": 1453,
      "train_runtime": 3963.7768,
      "train_tokens_per_second": 692.215
    },
    {
      "epoch": 1.4690422036896638,
      "grad_norm": 1.5973622798919678,
      "learning_rate": 5.336708860759494e-06,
      "loss": 0.057547181844711304,
      "num_input_tokens_seen": 2745376,
      "step": 1454,
      "train_runtime": 3965.8905,
      "train_tokens_per_second": 692.247
    },
    {
      "epoch": 1.4700530705079606,
      "grad_norm": 1.4456324577331543,
      "learning_rate": 5.3265822784810125e-06,
      "loss": 0.056840915232896805,
      "num_input_tokens_seen": 2747474,
      "step": 1455,
      "train_runtime": 3968.1596,
      "train_tokens_per_second": 692.38
    },
    {
      "epoch": 1.4710639373262573,
      "grad_norm": 1.569125771522522,
      "learning_rate": 5.3164556962025316e-06,
      "loss": 0.06012757495045662,
      "num_input_tokens_seen": 2749496,
      "step": 1456,
      "train_runtime": 3970.434,
      "train_tokens_per_second": 692.493
    },
    {
      "epoch": 1.472074804144554,
      "grad_norm": 1.5364798307418823,
      "learning_rate": 5.3063291139240515e-06,
      "loss": 0.07710444182157516,
      "num_input_tokens_seen": 2751386,
      "step": 1457,
      "train_runtime": 3972.644,
      "train_tokens_per_second": 692.583
    },
    {
      "epoch": 1.4730856709628506,
      "grad_norm": 1.6436177492141724,
      "learning_rate": 5.29620253164557e-06,
      "loss": 0.06534931808710098,
      "num_input_tokens_seen": 2753164,
      "step": 1458,
      "train_runtime": 3974.8311,
      "train_tokens_per_second": 692.649
    },
    {
      "epoch": 1.4740965377811475,
      "grad_norm": 1.5389630794525146,
      "learning_rate": 5.286075949367089e-06,
      "loss": 0.06264296174049377,
      "num_input_tokens_seen": 2755032,
      "step": 1459,
      "train_runtime": 3977.0651,
      "train_tokens_per_second": 692.73
    },
    {
      "epoch": 1.4751074045994441,
      "grad_norm": 1.715559720993042,
      "learning_rate": 5.275949367088608e-06,
      "loss": 0.05479292944073677,
      "num_input_tokens_seen": 2756680,
      "step": 1460,
      "train_runtime": 3979.2082,
      "train_tokens_per_second": 692.771
    },
    {
      "epoch": 1.4761182714177408,
      "grad_norm": 1.5218862295150757,
      "learning_rate": 5.265822784810127e-06,
      "loss": 0.05441203713417053,
      "num_input_tokens_seen": 2758496,
      "step": 1461,
      "train_runtime": 3981.3814,
      "train_tokens_per_second": 692.849
    },
    {
      "epoch": 1.4771291382360374,
      "grad_norm": 1.719048261642456,
      "learning_rate": 5.255696202531646e-06,
      "loss": 0.06784654408693314,
      "num_input_tokens_seen": 2760264,
      "step": 1462,
      "train_runtime": 3983.5277,
      "train_tokens_per_second": 692.919
    },
    {
      "epoch": 1.478140005054334,
      "grad_norm": 1.5282814502716064,
      "learning_rate": 5.245569620253165e-06,
      "loss": 0.06325586140155792,
      "num_input_tokens_seen": 2762474,
      "step": 1463,
      "train_runtime": 3985.8251,
      "train_tokens_per_second": 693.075
    },
    {
      "epoch": 1.4791508718726307,
      "grad_norm": 1.684726357460022,
      "learning_rate": 5.235443037974683e-06,
      "loss": 0.09345227479934692,
      "num_input_tokens_seen": 2764950,
      "step": 1464,
      "train_runtime": 3988.1829,
      "train_tokens_per_second": 693.286
    },
    {
      "epoch": 1.4801617386909274,
      "grad_norm": 1.6782022714614868,
      "learning_rate": 5.225316455696203e-06,
      "loss": 0.05765301361680031,
      "num_input_tokens_seen": 2766558,
      "step": 1465,
      "train_runtime": 3990.2964,
      "train_tokens_per_second": 693.321
    },
    {
      "epoch": 1.481172605509224,
      "grad_norm": 1.8614157438278198,
      "learning_rate": 5.2151898734177216e-06,
      "loss": 0.05889071151614189,
      "num_input_tokens_seen": 2768498,
      "step": 1466,
      "train_runtime": 3992.5301,
      "train_tokens_per_second": 693.419
    },
    {
      "epoch": 1.482183472327521,
      "grad_norm": 2.2837564945220947,
      "learning_rate": 5.205063291139241e-06,
      "loss": 0.08481335639953613,
      "num_input_tokens_seen": 2770106,
      "step": 1467,
      "train_runtime": 3994.6696,
      "train_tokens_per_second": 693.451
    },
    {
      "epoch": 1.4831943391458176,
      "grad_norm": 1.754288673400879,
      "learning_rate": 5.19493670886076e-06,
      "loss": 0.08612794429063797,
      "num_input_tokens_seen": 2771954,
      "step": 1468,
      "train_runtime": 3996.8334,
      "train_tokens_per_second": 693.538
    },
    {
      "epoch": 1.4842052059641142,
      "grad_norm": 2.616823196411133,
      "learning_rate": 5.184810126582279e-06,
      "loss": 0.060365933924913406,
      "num_input_tokens_seen": 2773760,
      "step": 1469,
      "train_runtime": 3999.0241,
      "train_tokens_per_second": 693.609
    },
    {
      "epoch": 1.4852160727824109,
      "grad_norm": 1.5449236631393433,
      "learning_rate": 5.174683544303798e-06,
      "loss": 0.07536165416240692,
      "num_input_tokens_seen": 2776132,
      "step": 1470,
      "train_runtime": 4001.3085,
      "train_tokens_per_second": 693.806
    },
    {
      "epoch": 1.4862269396007077,
      "grad_norm": 1.879141926765442,
      "learning_rate": 5.164556962025317e-06,
      "loss": 0.056780196726322174,
      "num_input_tokens_seen": 2777870,
      "step": 1471,
      "train_runtime": 4009.5867,
      "train_tokens_per_second": 692.807
    },
    {
      "epoch": 1.4872378064190044,
      "grad_norm": 1.9275370836257935,
      "learning_rate": 5.154430379746835e-06,
      "loss": 0.07316268980503082,
      "num_input_tokens_seen": 2779626,
      "step": 1472,
      "train_runtime": 4011.8369,
      "train_tokens_per_second": 692.856
    },
    {
      "epoch": 1.488248673237301,
      "grad_norm": 1.7731972932815552,
      "learning_rate": 5.144303797468355e-06,
      "loss": 0.06532439589500427,
      "num_input_tokens_seen": 2780798,
      "step": 1473,
      "train_runtime": 4013.9479,
      "train_tokens_per_second": 692.784
    },
    {
      "epoch": 1.4892595400555977,
      "grad_norm": 1.7615017890930176,
      "learning_rate": 5.134177215189874e-06,
      "loss": 0.07552695274353027,
      "num_input_tokens_seen": 2782748,
      "step": 1474,
      "train_runtime": 4016.1319,
      "train_tokens_per_second": 692.893
    },
    {
      "epoch": 1.4902704068738943,
      "grad_norm": 2.1878254413604736,
      "learning_rate": 5.1240506329113925e-06,
      "loss": 0.05805474519729614,
      "num_input_tokens_seen": 2784756,
      "step": 1475,
      "train_runtime": 4018.3974,
      "train_tokens_per_second": 693.002
    },
    {
      "epoch": 1.491281273692191,
      "grad_norm": 1.5653693675994873,
      "learning_rate": 5.1139240506329116e-06,
      "loss": 0.05668792873620987,
      "num_input_tokens_seen": 2786760,
      "step": 1476,
      "train_runtime": 4020.6413,
      "train_tokens_per_second": 693.113
    },
    {
      "epoch": 1.4922921405104876,
      "grad_norm": 1.833175539970398,
      "learning_rate": 5.1037974683544315e-06,
      "loss": 0.08333535492420197,
      "num_input_tokens_seen": 2788472,
      "step": 1477,
      "train_runtime": 4022.7867,
      "train_tokens_per_second": 693.169
    },
    {
      "epoch": 1.4933030073287845,
      "grad_norm": 1.3737703561782837,
      "learning_rate": 5.09367088607595e-06,
      "loss": 0.05029531568288803,
      "num_input_tokens_seen": 2790172,
      "step": 1478,
      "train_runtime": 4024.956,
      "train_tokens_per_second": 693.218
    },
    {
      "epoch": 1.4943138741470812,
      "grad_norm": 1.6074210405349731,
      "learning_rate": 5.083544303797469e-06,
      "loss": 0.05395309627056122,
      "num_input_tokens_seen": 2791846,
      "step": 1479,
      "train_runtime": 4027.1609,
      "train_tokens_per_second": 693.254
    },
    {
      "epoch": 1.4953247409653778,
      "grad_norm": 1.667372465133667,
      "learning_rate": 5.073417721518987e-06,
      "loss": 0.0809595137834549,
      "num_input_tokens_seen": 2793818,
      "step": 1480,
      "train_runtime": 4029.3947,
      "train_tokens_per_second": 693.359
    },
    {
      "epoch": 1.4963356077836745,
      "grad_norm": 1.7782727479934692,
      "learning_rate": 5.063291139240507e-06,
      "loss": 0.060248326510190964,
      "num_input_tokens_seen": 2795626,
      "step": 1481,
      "train_runtime": 4031.5828,
      "train_tokens_per_second": 693.431
    },
    {
      "epoch": 1.4973464746019711,
      "grad_norm": 2.026197671890259,
      "learning_rate": 5.053164556962026e-06,
      "loss": 0.08361667394638062,
      "num_input_tokens_seen": 2797286,
      "step": 1482,
      "train_runtime": 4033.733,
      "train_tokens_per_second": 693.473
    },
    {
      "epoch": 1.498357341420268,
      "grad_norm": 2.1825647354125977,
      "learning_rate": 5.043037974683544e-06,
      "loss": 0.07168866693973541,
      "num_input_tokens_seen": 2798946,
      "step": 1483,
      "train_runtime": 4035.8867,
      "train_tokens_per_second": 693.515
    },
    {
      "epoch": 1.4993682082385646,
      "grad_norm": 1.9255553483963013,
      "learning_rate": 5.032911392405063e-06,
      "loss": 0.08370400965213776,
      "num_input_tokens_seen": 2800864,
      "step": 1484,
      "train_runtime": 4038.1086,
      "train_tokens_per_second": 693.608
    },
    {
      "epoch": 1.5003790750568613,
      "grad_norm": 1.5774973630905151,
      "learning_rate": 5.022784810126583e-06,
      "loss": 0.06282275170087814,
      "num_input_tokens_seen": 2802846,
      "step": 1485,
      "train_runtime": 4040.3976,
      "train_tokens_per_second": 693.705
    },
    {
      "epoch": 1.501389941875158,
      "grad_norm": 2.2015116214752197,
      "learning_rate": 5.0126582278481016e-06,
      "loss": 0.0942520946264267,
      "num_input_tokens_seen": 2804606,
      "step": 1486,
      "train_runtime": 4042.5864,
      "train_tokens_per_second": 693.765
    },
    {
      "epoch": 1.5024008086934546,
      "grad_norm": 1.7354520559310913,
      "learning_rate": 5.002531645569621e-06,
      "loss": 0.06485570967197418,
      "num_input_tokens_seen": 2806254,
      "step": 1487,
      "train_runtime": 4044.7431,
      "train_tokens_per_second": 693.803
    },
    {
      "epoch": 1.5034116755117513,
      "grad_norm": 1.662319540977478,
      "learning_rate": 4.99240506329114e-06,
      "loss": 0.06699532270431519,
      "num_input_tokens_seen": 2807812,
      "step": 1488,
      "train_runtime": 4046.8851,
      "train_tokens_per_second": 693.821
    },
    {
      "epoch": 1.504422542330048,
      "grad_norm": 1.3281782865524292,
      "learning_rate": 4.982278481012659e-06,
      "loss": 0.052544791251420975,
      "num_input_tokens_seen": 2809866,
      "step": 1489,
      "train_runtime": 4049.0936,
      "train_tokens_per_second": 693.949
    },
    {
      "epoch": 1.5054334091483446,
      "grad_norm": 1.7143620252609253,
      "learning_rate": 4.972151898734178e-06,
      "loss": 0.06422758847475052,
      "num_input_tokens_seen": 2811744,
      "step": 1490,
      "train_runtime": 4051.2866,
      "train_tokens_per_second": 694.037
    },
    {
      "epoch": 1.5064442759666414,
      "grad_norm": 1.552291750907898,
      "learning_rate": 4.962025316455696e-06,
      "loss": 0.07418136298656464,
      "num_input_tokens_seen": 2814384,
      "step": 1491,
      "train_runtime": 4053.6381,
      "train_tokens_per_second": 694.286
    },
    {
      "epoch": 1.507455142784938,
      "grad_norm": 1.4776345491409302,
      "learning_rate": 4.951898734177216e-06,
      "loss": 0.06440728902816772,
      "num_input_tokens_seen": 2816184,
      "step": 1492,
      "train_runtime": 4055.8897,
      "train_tokens_per_second": 694.344
    },
    {
      "epoch": 1.5084660096032347,
      "grad_norm": 1.6045451164245605,
      "learning_rate": 4.941772151898734e-06,
      "loss": 0.06612151861190796,
      "num_input_tokens_seen": 2818180,
      "step": 1493,
      "train_runtime": 4058.1291,
      "train_tokens_per_second": 694.453
    },
    {
      "epoch": 1.5094768764215316,
      "grad_norm": 1.7005598545074463,
      "learning_rate": 4.931645569620253e-06,
      "loss": 0.07076660543680191,
      "num_input_tokens_seen": 2820102,
      "step": 1494,
      "train_runtime": 4060.3338,
      "train_tokens_per_second": 694.549
    },
    {
      "epoch": 1.5104877432398283,
      "grad_norm": 1.6715127229690552,
      "learning_rate": 4.9215189873417725e-06,
      "loss": 0.05914783850312233,
      "num_input_tokens_seen": 2821580,
      "step": 1495,
      "train_runtime": 4062.4516,
      "train_tokens_per_second": 694.551
    },
    {
      "epoch": 1.511498610058125,
      "grad_norm": 1.5523298978805542,
      "learning_rate": 4.9113924050632915e-06,
      "loss": 0.057301413267850876,
      "num_input_tokens_seen": 2824202,
      "step": 1496,
      "train_runtime": 4064.866,
      "train_tokens_per_second": 694.784
    },
    {
      "epoch": 1.5125094768764216,
      "grad_norm": 1.474369764328003,
      "learning_rate": 4.901265822784811e-06,
      "loss": 0.06793700903654099,
      "num_input_tokens_seen": 2826432,
      "step": 1497,
      "train_runtime": 4067.1398,
      "train_tokens_per_second": 694.943
    },
    {
      "epoch": 1.5135203436947182,
      "grad_norm": 1.3737822771072388,
      "learning_rate": 4.89113924050633e-06,
      "loss": 0.05869380757212639,
      "num_input_tokens_seen": 2828690,
      "step": 1498,
      "train_runtime": 4069.436,
      "train_tokens_per_second": 695.106
    },
    {
      "epoch": 1.5145312105130149,
      "grad_norm": 1.8771203756332397,
      "learning_rate": 4.881012658227848e-06,
      "loss": 0.07134260982275009,
      "num_input_tokens_seen": 2830248,
      "step": 1499,
      "train_runtime": 4071.5527,
      "train_tokens_per_second": 695.127
    },
    {
      "epoch": 1.5155420773313115,
      "grad_norm": 1.5327411890029907,
      "learning_rate": 4.870886075949368e-06,
      "loss": 0.06758476048707962,
      "num_input_tokens_seen": 2832522,
      "step": 1500,
      "train_runtime": 4073.8616,
      "train_tokens_per_second": 695.292
    },
    {
      "epoch": 1.5165529441496082,
      "grad_norm": 1.4994726181030273,
      "learning_rate": 4.860759493670886e-06,
      "loss": 0.06690610945224762,
      "num_input_tokens_seen": 2834288,
      "step": 1501,
      "train_runtime": 4082.459,
      "train_tokens_per_second": 694.26
    },
    {
      "epoch": 1.5175638109679048,
      "grad_norm": 1.600295066833496,
      "learning_rate": 4.850632911392405e-06,
      "loss": 0.06315764784812927,
      "num_input_tokens_seen": 2836114,
      "step": 1502,
      "train_runtime": 4084.7241,
      "train_tokens_per_second": 694.322
    },
    {
      "epoch": 1.5185746777862017,
      "grad_norm": 1.817538857460022,
      "learning_rate": 4.840506329113924e-06,
      "loss": 0.08383622020483017,
      "num_input_tokens_seen": 2838072,
      "step": 1503,
      "train_runtime": 4087.1024,
      "train_tokens_per_second": 694.397
    },
    {
      "epoch": 1.5195855446044984,
      "grad_norm": 1.6323297023773193,
      "learning_rate": 4.830379746835443e-06,
      "loss": 0.05724768713116646,
      "num_input_tokens_seen": 2840010,
      "step": 1504,
      "train_runtime": 4089.3052,
      "train_tokens_per_second": 694.497
    },
    {
      "epoch": 1.520596411422795,
      "grad_norm": 1.6135390996932983,
      "learning_rate": 4.8202531645569625e-06,
      "loss": 0.05224788933992386,
      "num_input_tokens_seen": 2841990,
      "step": 1505,
      "train_runtime": 4091.5168,
      "train_tokens_per_second": 694.605
    },
    {
      "epoch": 1.5216072782410919,
      "grad_norm": 1.7678437232971191,
      "learning_rate": 4.8101265822784815e-06,
      "loss": 0.06881388276815414,
      "num_input_tokens_seen": 2843498,
      "step": 1506,
      "train_runtime": 4093.6093,
      "train_tokens_per_second": 694.619
    },
    {
      "epoch": 1.5226181450593885,
      "grad_norm": 1.913795828819275,
      "learning_rate": 4.800000000000001e-06,
      "loss": 0.0925389900803566,
      "num_input_tokens_seen": 2845090,
      "step": 1507,
      "train_runtime": 4095.7249,
      "train_tokens_per_second": 694.649
    },
    {
      "epoch": 1.5236290118776852,
      "grad_norm": 1.9918967485427856,
      "learning_rate": 4.78987341772152e-06,
      "loss": 0.08005580306053162,
      "num_input_tokens_seen": 2846700,
      "step": 1508,
      "train_runtime": 4098.0559,
      "train_tokens_per_second": 694.646
    },
    {
      "epoch": 1.5246398786959818,
      "grad_norm": 1.7330108880996704,
      "learning_rate": 4.779746835443038e-06,
      "loss": 0.05722786486148834,
      "num_input_tokens_seen": 2848114,
      "step": 1509,
      "train_runtime": 4100.1429,
      "train_tokens_per_second": 694.638
    },
    {
      "epoch": 1.5256507455142785,
      "grad_norm": 1.9117228984832764,
      "learning_rate": 4.769620253164557e-06,
      "loss": 0.09200979769229889,
      "num_input_tokens_seen": 2849846,
      "step": 1510,
      "train_runtime": 4102.3037,
      "train_tokens_per_second": 694.694
    },
    {
      "epoch": 1.5266616123325751,
      "grad_norm": 1.6707384586334229,
      "learning_rate": 4.759493670886076e-06,
      "loss": 0.07151450216770172,
      "num_input_tokens_seen": 2851468,
      "step": 1511,
      "train_runtime": 4104.4497,
      "train_tokens_per_second": 694.726
    },
    {
      "epoch": 1.5276724791508718,
      "grad_norm": 1.771851658821106,
      "learning_rate": 4.749367088607595e-06,
      "loss": 0.08380179107189178,
      "num_input_tokens_seen": 2853370,
      "step": 1512,
      "train_runtime": 4106.6563,
      "train_tokens_per_second": 694.816
    },
    {
      "epoch": 1.5286833459691684,
      "grad_norm": 1.8305950164794922,
      "learning_rate": 4.739240506329114e-06,
      "loss": 0.08426954597234726,
      "num_input_tokens_seen": 2855182,
      "step": 1513,
      "train_runtime": 4108.8379,
      "train_tokens_per_second": 694.888
    },
    {
      "epoch": 1.529694212787465,
      "grad_norm": 2.3114895820617676,
      "learning_rate": 4.729113924050633e-06,
      "loss": 0.08215209096670151,
      "num_input_tokens_seen": 2857214,
      "step": 1514,
      "train_runtime": 4111.0775,
      "train_tokens_per_second": 695.004
    },
    {
      "epoch": 1.530705079605762,
      "grad_norm": 1.875800609588623,
      "learning_rate": 4.7189873417721525e-06,
      "loss": 0.07074695080518723,
      "num_input_tokens_seen": 2859046,
      "step": 1515,
      "train_runtime": 4113.2816,
      "train_tokens_per_second": 695.077
    },
    {
      "epoch": 1.5317159464240586,
      "grad_norm": 1.830914855003357,
      "learning_rate": 4.7088607594936715e-06,
      "loss": 0.07563849538564682,
      "num_input_tokens_seen": 2861014,
      "step": 1516,
      "train_runtime": 4115.5223,
      "train_tokens_per_second": 695.176
    },
    {
      "epoch": 1.5327268132423553,
      "grad_norm": 1.6784169673919678,
      "learning_rate": 4.69873417721519e-06,
      "loss": 0.07204784452915192,
      "num_input_tokens_seen": 2862968,
      "step": 1517,
      "train_runtime": 4117.7335,
      "train_tokens_per_second": 695.278
    },
    {
      "epoch": 1.5337376800606521,
      "grad_norm": 1.5273574590682983,
      "learning_rate": 4.68860759493671e-06,
      "loss": 0.05450946092605591,
      "num_input_tokens_seen": 2864930,
      "step": 1518,
      "train_runtime": 4119.9373,
      "train_tokens_per_second": 695.382
    },
    {
      "epoch": 1.5347485468789488,
      "grad_norm": 1.5135360956192017,
      "learning_rate": 4.678481012658228e-06,
      "loss": 0.059337448328733444,
      "num_input_tokens_seen": 2867000,
      "step": 1519,
      "train_runtime": 4122.1856,
      "train_tokens_per_second": 695.505
    },
    {
      "epoch": 1.5357594136972454,
      "grad_norm": 1.509782314300537,
      "learning_rate": 4.668354430379747e-06,
      "loss": 0.0536489337682724,
      "num_input_tokens_seen": 2869152,
      "step": 1520,
      "train_runtime": 4124.4506,
      "train_tokens_per_second": 695.645
    },
    {
      "epoch": 1.536770280515542,
      "grad_norm": 2.1016454696655273,
      "learning_rate": 4.658227848101266e-06,
      "loss": 0.07627323269844055,
      "num_input_tokens_seen": 2870778,
      "step": 1521,
      "train_runtime": 4126.6768,
      "train_tokens_per_second": 695.663
    },
    {
      "epoch": 1.5377811473338387,
      "grad_norm": 1.5608389377593994,
      "learning_rate": 4.648101265822785e-06,
      "loss": 0.07444711029529572,
      "num_input_tokens_seen": 2872896,
      "step": 1522,
      "train_runtime": 4128.9836,
      "train_tokens_per_second": 695.788
    },
    {
      "epoch": 1.5387920141521354,
      "grad_norm": 1.8106434345245361,
      "learning_rate": 4.637974683544304e-06,
      "loss": 0.06342519074678421,
      "num_input_tokens_seen": 2874902,
      "step": 1523,
      "train_runtime": 4131.2116,
      "train_tokens_per_second": 695.898
    },
    {
      "epoch": 1.539802880970432,
      "grad_norm": 1.6673105955123901,
      "learning_rate": 4.627848101265823e-06,
      "loss": 0.06298794597387314,
      "num_input_tokens_seen": 2876462,
      "step": 1524,
      "train_runtime": 4133.3725,
      "train_tokens_per_second": 695.912
    },
    {
      "epoch": 1.5408137477887287,
      "grad_norm": 1.5623661279678345,
      "learning_rate": 4.617721518987342e-06,
      "loss": 0.07657312601804733,
      "num_input_tokens_seen": 2878202,
      "step": 1525,
      "train_runtime": 4135.5681,
      "train_tokens_per_second": 695.963
    },
    {
      "epoch": 1.5418246146070256,
      "grad_norm": 1.5652337074279785,
      "learning_rate": 4.6075949367088615e-06,
      "loss": 0.07076507806777954,
      "num_input_tokens_seen": 2880316,
      "step": 1526,
      "train_runtime": 4137.8502,
      "train_tokens_per_second": 696.09
    },
    {
      "epoch": 1.5428354814253222,
      "grad_norm": 1.7186670303344727,
      "learning_rate": 4.59746835443038e-06,
      "loss": 0.07635582238435745,
      "num_input_tokens_seen": 2881992,
      "step": 1527,
      "train_runtime": 4140.0128,
      "train_tokens_per_second": 696.131
    },
    {
      "epoch": 1.5438463482436189,
      "grad_norm": 1.3687337636947632,
      "learning_rate": 4.587341772151899e-06,
      "loss": 0.05123043805360794,
      "num_input_tokens_seen": 2883762,
      "step": 1528,
      "train_runtime": 4142.2782,
      "train_tokens_per_second": 696.178
    },
    {
      "epoch": 1.5448572150619158,
      "grad_norm": 2.1513876914978027,
      "learning_rate": 4.577215189873418e-06,
      "loss": 0.06788188219070435,
      "num_input_tokens_seen": 2886030,
      "step": 1529,
      "train_runtime": 4144.7615,
      "train_tokens_per_second": 696.308
    },
    {
      "epoch": 1.5458680818802124,
      "grad_norm": 1.595686435699463,
      "learning_rate": 4.567088607594937e-06,
      "loss": 0.0664450153708458,
      "num_input_tokens_seen": 2887852,
      "step": 1530,
      "train_runtime": 4146.9505,
      "train_tokens_per_second": 696.38
    },
    {
      "epoch": 1.546878948698509,
      "grad_norm": 1.803052306175232,
      "learning_rate": 4.556962025316456e-06,
      "loss": 0.07884696125984192,
      "num_input_tokens_seen": 2889610,
      "step": 1531,
      "train_runtime": 4155.6233,
      "train_tokens_per_second": 695.349
    },
    {
      "epoch": 1.5478898155168057,
      "grad_norm": 1.6938878297805786,
      "learning_rate": 4.546835443037975e-06,
      "loss": 0.06919347494840622,
      "num_input_tokens_seen": 2891706,
      "step": 1532,
      "train_runtime": 4157.8654,
      "train_tokens_per_second": 695.479
    },
    {
      "epoch": 1.5489006823351024,
      "grad_norm": 1.9805231094360352,
      "learning_rate": 4.536708860759494e-06,
      "loss": 0.08436493575572968,
      "num_input_tokens_seen": 2893848,
      "step": 1533,
      "train_runtime": 4160.1334,
      "train_tokens_per_second": 695.614
    },
    {
      "epoch": 1.549911549153399,
      "grad_norm": 1.4778332710266113,
      "learning_rate": 4.526582278481013e-06,
      "loss": 0.05188169330358505,
      "num_input_tokens_seen": 2895582,
      "step": 1534,
      "train_runtime": 4162.3948,
      "train_tokens_per_second": 695.653
    },
    {
      "epoch": 1.5509224159716957,
      "grad_norm": 1.6248661279678345,
      "learning_rate": 4.516455696202532e-06,
      "loss": 0.0605023130774498,
      "num_input_tokens_seen": 2897422,
      "step": 1535,
      "train_runtime": 4164.5631,
      "train_tokens_per_second": 695.733
    },
    {
      "epoch": 1.5519332827899923,
      "grad_norm": 1.6457431316375732,
      "learning_rate": 4.5063291139240515e-06,
      "loss": 0.048475056886672974,
      "num_input_tokens_seen": 2898578,
      "step": 1536,
      "train_runtime": 4166.5701,
      "train_tokens_per_second": 695.675
    },
    {
      "epoch": 1.552944149608289,
      "grad_norm": 1.3036322593688965,
      "learning_rate": 4.49620253164557e-06,
      "loss": 0.061688873916864395,
      "num_input_tokens_seen": 2900830,
      "step": 1537,
      "train_runtime": 4168.908,
      "train_tokens_per_second": 695.825
    },
    {
      "epoch": 1.5539550164265858,
      "grad_norm": 1.6548831462860107,
      "learning_rate": 4.486075949367089e-06,
      "loss": 0.06499501317739487,
      "num_input_tokens_seen": 2902758,
      "step": 1538,
      "train_runtime": 4171.1733,
      "train_tokens_per_second": 695.909
    },
    {
      "epoch": 1.5549658832448825,
      "grad_norm": 1.465171456336975,
      "learning_rate": 4.475949367088608e-06,
      "loss": 0.06906162202358246,
      "num_input_tokens_seen": 2904738,
      "step": 1539,
      "train_runtime": 4173.3907,
      "train_tokens_per_second": 696.014
    },
    {
      "epoch": 1.5559767500631791,
      "grad_norm": 1.5636467933654785,
      "learning_rate": 4.465822784810127e-06,
      "loss": 0.0834270566701889,
      "num_input_tokens_seen": 2907096,
      "step": 1540,
      "train_runtime": 4175.7625,
      "train_tokens_per_second": 696.183
    },
    {
      "epoch": 1.556987616881476,
      "grad_norm": 1.7297980785369873,
      "learning_rate": 4.455696202531646e-06,
      "loss": 0.07621173560619354,
      "num_input_tokens_seen": 2908836,
      "step": 1541,
      "train_runtime": 4177.9427,
      "train_tokens_per_second": 696.236
    },
    {
      "epoch": 1.5579984836997727,
      "grad_norm": 1.9271103143692017,
      "learning_rate": 4.445569620253165e-06,
      "loss": 0.07846500724554062,
      "num_input_tokens_seen": 2910510,
      "step": 1542,
      "train_runtime": 4180.1229,
      "train_tokens_per_second": 696.274
    },
    {
      "epoch": 1.5590093505180693,
      "grad_norm": 1.575373888015747,
      "learning_rate": 4.4354430379746834e-06,
      "loss": 0.05334926396608353,
      "num_input_tokens_seen": 2912316,
      "step": 1543,
      "train_runtime": 4182.3641,
      "train_tokens_per_second": 696.332
    },
    {
      "epoch": 1.560020217336366,
      "grad_norm": 1.826368808746338,
      "learning_rate": 4.425316455696203e-06,
      "loss": 0.09432559460401535,
      "num_input_tokens_seen": 2914070,
      "step": 1544,
      "train_runtime": 4184.5555,
      "train_tokens_per_second": 696.387
    },
    {
      "epoch": 1.5610310841546626,
      "grad_norm": 1.9886971712112427,
      "learning_rate": 4.415189873417722e-06,
      "loss": 0.10036467760801315,
      "num_input_tokens_seen": 2916386,
      "step": 1545,
      "train_runtime": 4186.8955,
      "train_tokens_per_second": 696.551
    },
    {
      "epoch": 1.5620419509729593,
      "grad_norm": 1.8595750331878662,
      "learning_rate": 4.405063291139241e-06,
      "loss": 0.07906724512577057,
      "num_input_tokens_seen": 2918148,
      "step": 1546,
      "train_runtime": 4189.0786,
      "train_tokens_per_second": 696.609
    },
    {
      "epoch": 1.563052817791256,
      "grad_norm": 1.4908061027526855,
      "learning_rate": 4.39493670886076e-06,
      "loss": 0.07332147657871246,
      "num_input_tokens_seen": 2920642,
      "step": 1547,
      "train_runtime": 4191.4473,
      "train_tokens_per_second": 696.81
    },
    {
      "epoch": 1.5640636846095526,
      "grad_norm": 1.8409886360168457,
      "learning_rate": 4.384810126582279e-06,
      "loss": 0.07795832306146622,
      "num_input_tokens_seen": 2922254,
      "step": 1548,
      "train_runtime": 4193.6605,
      "train_tokens_per_second": 696.827
    },
    {
      "epoch": 1.5650745514278492,
      "grad_norm": 1.5730971097946167,
      "learning_rate": 4.374683544303798e-06,
      "loss": 0.0563131645321846,
      "num_input_tokens_seen": 2924156,
      "step": 1549,
      "train_runtime": 4195.9226,
      "train_tokens_per_second": 696.904
    },
    {
      "epoch": 1.566085418246146,
      "grad_norm": 2.096383810043335,
      "learning_rate": 4.364556962025317e-06,
      "loss": 0.09253260493278503,
      "num_input_tokens_seen": 2926290,
      "step": 1550,
      "train_runtime": 4198.1958,
      "train_tokens_per_second": 697.035
    },
    {
      "epoch": 1.5670962850644428,
      "grad_norm": 1.662731409072876,
      "learning_rate": 4.354430379746836e-06,
      "loss": 0.06166689097881317,
      "num_input_tokens_seen": 2928344,
      "step": 1551,
      "train_runtime": 4200.4667,
      "train_tokens_per_second": 697.147
    },
    {
      "epoch": 1.5681071518827394,
      "grad_norm": 1.193656325340271,
      "learning_rate": 4.344303797468355e-06,
      "loss": 0.04309312254190445,
      "num_input_tokens_seen": 2930198,
      "step": 1552,
      "train_runtime": 4202.6776,
      "train_tokens_per_second": 697.222
    },
    {
      "epoch": 1.5691180187010363,
      "grad_norm": 1.6584079265594482,
      "learning_rate": 4.3341772151898734e-06,
      "loss": 0.0468953512609005,
      "num_input_tokens_seen": 2931694,
      "step": 1553,
      "train_runtime": 4204.7669,
      "train_tokens_per_second": 697.231
    },
    {
      "epoch": 1.570128885519333,
      "grad_norm": 1.9389419555664062,
      "learning_rate": 4.3240506329113925e-06,
      "loss": 0.09267403930425644,
      "num_input_tokens_seen": 2933774,
      "step": 1554,
      "train_runtime": 4207.0499,
      "train_tokens_per_second": 697.347
    },
    {
      "epoch": 1.5711397523376296,
      "grad_norm": 1.7845357656478882,
      "learning_rate": 4.313924050632912e-06,
      "loss": 0.08263450860977173,
      "num_input_tokens_seen": 2935584,
      "step": 1555,
      "train_runtime": 4209.2304,
      "train_tokens_per_second": 697.416
    },
    {
      "epoch": 1.5721506191559262,
      "grad_norm": 2.0508620738983154,
      "learning_rate": 4.303797468354431e-06,
      "loss": 0.07426123321056366,
      "num_input_tokens_seen": 2937110,
      "step": 1556,
      "train_runtime": 4211.4416,
      "train_tokens_per_second": 697.412
    },
    {
      "epoch": 1.573161485974223,
      "grad_norm": 1.685733437538147,
      "learning_rate": 4.29367088607595e-06,
      "loss": 0.05370761454105377,
      "num_input_tokens_seen": 2938708,
      "step": 1557,
      "train_runtime": 4213.6474,
      "train_tokens_per_second": 697.426
    },
    {
      "epoch": 1.5741723527925195,
      "grad_norm": 1.7071869373321533,
      "learning_rate": 4.283544303797469e-06,
      "loss": 0.0799432173371315,
      "num_input_tokens_seen": 2941278,
      "step": 1558,
      "train_runtime": 4216.021,
      "train_tokens_per_second": 697.643
    },
    {
      "epoch": 1.5751832196108162,
      "grad_norm": 1.7251238822937012,
      "learning_rate": 4.273417721518988e-06,
      "loss": 0.07716953009366989,
      "num_input_tokens_seen": 2942738,
      "step": 1559,
      "train_runtime": 4218.1611,
      "train_tokens_per_second": 697.635
    },
    {
      "epoch": 1.5761940864291129,
      "grad_norm": 1.7671867609024048,
      "learning_rate": 4.263291139240507e-06,
      "loss": 0.07282590121030807,
      "num_input_tokens_seen": 2944420,
      "step": 1560,
      "train_runtime": 4220.2998,
      "train_tokens_per_second": 697.68
    },
    {
      "epoch": 1.5772049532474095,
      "grad_norm": 1.8977595567703247,
      "learning_rate": 4.253164556962025e-06,
      "loss": 0.08582794666290283,
      "num_input_tokens_seen": 2946336,
      "step": 1561,
      "train_runtime": 4229.2752,
      "train_tokens_per_second": 696.653
    },
    {
      "epoch": 1.5782158200657064,
      "grad_norm": 1.7306803464889526,
      "learning_rate": 4.243037974683545e-06,
      "loss": 0.07706762105226517,
      "num_input_tokens_seen": 2947836,
      "step": 1562,
      "train_runtime": 4231.482,
      "train_tokens_per_second": 696.644
    },
    {
      "epoch": 1.579226686884003,
      "grad_norm": 1.5827447175979614,
      "learning_rate": 4.232911392405063e-06,
      "loss": 0.07217445969581604,
      "num_input_tokens_seen": 2950112,
      "step": 1563,
      "train_runtime": 4233.7765,
      "train_tokens_per_second": 696.804
    },
    {
      "epoch": 1.5802375537022997,
      "grad_norm": 2.0294995307922363,
      "learning_rate": 4.2227848101265825e-06,
      "loss": 0.08813823014497757,
      "num_input_tokens_seen": 2952126,
      "step": 1564,
      "train_runtime": 4236.0123,
      "train_tokens_per_second": 696.912
    },
    {
      "epoch": 1.5812484205205966,
      "grad_norm": 1.4940035343170166,
      "learning_rate": 4.212658227848102e-06,
      "loss": 0.0729675441980362,
      "num_input_tokens_seen": 2953936,
      "step": 1565,
      "train_runtime": 4238.2175,
      "train_tokens_per_second": 696.976
    },
    {
      "epoch": 1.5822592873388932,
      "grad_norm": 1.6191201210021973,
      "learning_rate": 4.202531645569621e-06,
      "loss": 0.07353523373603821,
      "num_input_tokens_seen": 2955706,
      "step": 1566,
      "train_runtime": 4240.402,
      "train_tokens_per_second": 697.034
    },
    {
      "epoch": 1.5832701541571899,
      "grad_norm": 1.9050995111465454,
      "learning_rate": 4.19240506329114e-06,
      "loss": 0.08147936314344406,
      "num_input_tokens_seen": 2957292,
      "step": 1567,
      "train_runtime": 4242.5713,
      "train_tokens_per_second": 697.052
    },
    {
      "epoch": 1.5842810209754865,
      "grad_norm": 1.478128433227539,
      "learning_rate": 4.182278481012659e-06,
      "loss": 0.06900140643119812,
      "num_input_tokens_seen": 2959516,
      "step": 1568,
      "train_runtime": 4244.9261,
      "train_tokens_per_second": 697.189
    },
    {
      "epoch": 1.5852918877937832,
      "grad_norm": 1.4391906261444092,
      "learning_rate": 4.172151898734177e-06,
      "loss": 0.06270866096019745,
      "num_input_tokens_seen": 2961496,
      "step": 1569,
      "train_runtime": 4247.1968,
      "train_tokens_per_second": 697.282
    },
    {
      "epoch": 1.5863027546120798,
      "grad_norm": 1.5638253688812256,
      "learning_rate": 4.162025316455697e-06,
      "loss": 0.05708848685026169,
      "num_input_tokens_seen": 2963170,
      "step": 1570,
      "train_runtime": 4249.383,
      "train_tokens_per_second": 697.318
    },
    {
      "epoch": 1.5873136214303765,
      "grad_norm": 1.6085665225982666,
      "learning_rate": 4.151898734177215e-06,
      "loss": 0.06098378449678421,
      "num_input_tokens_seen": 2965072,
      "step": 1571,
      "train_runtime": 4251.6007,
      "train_tokens_per_second": 697.401
    },
    {
      "epoch": 1.5883244882486731,
      "grad_norm": 1.5476840734481812,
      "learning_rate": 4.141772151898734e-06,
      "loss": 0.07203972339630127,
      "num_input_tokens_seen": 2967136,
      "step": 1572,
      "train_runtime": 4253.8611,
      "train_tokens_per_second": 697.516
    },
    {
      "epoch": 1.5893353550669698,
      "grad_norm": 1.5056819915771484,
      "learning_rate": 4.131645569620253e-06,
      "loss": 0.05791059881448746,
      "num_input_tokens_seen": 2968980,
      "step": 1573,
      "train_runtime": 4256.0918,
      "train_tokens_per_second": 697.584
    },
    {
      "epoch": 1.5903462218852666,
      "grad_norm": 1.7315961122512817,
      "learning_rate": 4.1215189873417725e-06,
      "loss": 0.07439381629228592,
      "num_input_tokens_seen": 2970630,
      "step": 1574,
      "train_runtime": 4258.2375,
      "train_tokens_per_second": 697.62
    },
    {
      "epoch": 1.5913570887035633,
      "grad_norm": 1.4721879959106445,
      "learning_rate": 4.111392405063292e-06,
      "loss": 0.060391560196876526,
      "num_input_tokens_seen": 2972686,
      "step": 1575,
      "train_runtime": 4260.5192,
      "train_tokens_per_second": 697.729
    },
    {
      "epoch": 1.59236795552186,
      "grad_norm": 1.6085623502731323,
      "learning_rate": 4.101265822784811e-06,
      "loss": 0.0684574544429779,
      "num_input_tokens_seen": 2974526,
      "step": 1576,
      "train_runtime": 4262.7479,
      "train_tokens_per_second": 697.795
    },
    {
      "epoch": 1.5933788223401568,
      "grad_norm": 1.4107952117919922,
      "learning_rate": 4.09113924050633e-06,
      "loss": 0.05432877317070961,
      "num_input_tokens_seen": 2976356,
      "step": 1577,
      "train_runtime": 4264.9605,
      "train_tokens_per_second": 697.862
    },
    {
      "epoch": 1.5943896891584535,
      "grad_norm": 1.934281587600708,
      "learning_rate": 4.081012658227849e-06,
      "loss": 0.07048742473125458,
      "num_input_tokens_seen": 2977998,
      "step": 1578,
      "train_runtime": 4267.1758,
      "train_tokens_per_second": 697.885
    },
    {
      "epoch": 1.5954005559767501,
      "grad_norm": 1.5795378684997559,
      "learning_rate": 4.070886075949367e-06,
      "loss": 0.06708382070064545,
      "num_input_tokens_seen": 2979636,
      "step": 1579,
      "train_runtime": 4269.2968,
      "train_tokens_per_second": 697.922
    },
    {
      "epoch": 1.5964114227950468,
      "grad_norm": 1.7154045104980469,
      "learning_rate": 4.060759493670887e-06,
      "loss": 0.07803279906511307,
      "num_input_tokens_seen": 2981328,
      "step": 1580,
      "train_runtime": 4271.5225,
      "train_tokens_per_second": 697.954
    },
    {
      "epoch": 1.5974222896133434,
      "grad_norm": 1.827027440071106,
      "learning_rate": 4.050632911392405e-06,
      "loss": 0.07495563477277756,
      "num_input_tokens_seen": 2982932,
      "step": 1581,
      "train_runtime": 4273.6507,
      "train_tokens_per_second": 697.982
    },
    {
      "epoch": 1.59843315643164,
      "grad_norm": 1.8193178176879883,
      "learning_rate": 4.040506329113924e-06,
      "loss": 0.07988477498292923,
      "num_input_tokens_seen": 2985058,
      "step": 1582,
      "train_runtime": 4275.92,
      "train_tokens_per_second": 698.109
    },
    {
      "epoch": 1.5994440232499367,
      "grad_norm": 1.5274325609207153,
      "learning_rate": 4.030379746835443e-06,
      "loss": 0.057682231068611145,
      "num_input_tokens_seen": 2987304,
      "step": 1583,
      "train_runtime": 4278.2365,
      "train_tokens_per_second": 698.256
    },
    {
      "epoch": 1.6004548900682334,
      "grad_norm": 1.6576076745986938,
      "learning_rate": 4.0202531645569625e-06,
      "loss": 0.07831859588623047,
      "num_input_tokens_seen": 2989272,
      "step": 1584,
      "train_runtime": 4280.5089,
      "train_tokens_per_second": 698.345
    },
    {
      "epoch": 1.6004548900682334,
      "eval_loss": 0.39059677720069885,
      "eval_runtime": 52.2929,
      "eval_samples_per_second": 16.809,
      "eval_steps_per_second": 8.414,
      "num_input_tokens_seen": 2989272,
      "step": 1584
    },
    {
      "epoch": 1.60146575688653,
      "grad_norm": 1.6437644958496094,
      "learning_rate": 4.010126582278482e-06,
      "loss": 0.05386246368288994,
      "num_input_tokens_seen": 2991142,
      "step": 1585,
      "train_runtime": 4335.0405,
      "train_tokens_per_second": 689.992
    },
    {
      "epoch": 1.602476623704827,
      "grad_norm": 1.8074007034301758,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.0719929188489914,
      "num_input_tokens_seen": 2992704,
      "step": 1586,
      "train_runtime": 4337.1586,
      "train_tokens_per_second": 690.015
    },
    {
      "epoch": 1.6034874905231236,
      "grad_norm": 1.7193204164505005,
      "learning_rate": 3.989873417721519e-06,
      "loss": 0.069422647356987,
      "num_input_tokens_seen": 2994422,
      "step": 1587,
      "train_runtime": 4339.3055,
      "train_tokens_per_second": 690.069
    },
    {
      "epoch": 1.6044983573414204,
      "grad_norm": 2.030726909637451,
      "learning_rate": 3.979746835443039e-06,
      "loss": 0.09575729072093964,
      "num_input_tokens_seen": 2996456,
      "step": 1588,
      "train_runtime": 4341.5585,
      "train_tokens_per_second": 690.18
    },
    {
      "epoch": 1.605509224159717,
      "grad_norm": 1.33951735496521,
      "learning_rate": 3.969620253164557e-06,
      "loss": 0.05319521203637123,
      "num_input_tokens_seen": 2998580,
      "step": 1589,
      "train_runtime": 4343.8351,
      "train_tokens_per_second": 690.307
    },
    {
      "epoch": 1.6065200909780137,
      "grad_norm": 1.704698920249939,
      "learning_rate": 3.959493670886076e-06,
      "loss": 0.06892026215791702,
      "num_input_tokens_seen": 3000296,
      "step": 1590,
      "train_runtime": 4346.0234,
      "train_tokens_per_second": 690.354
    },
    {
      "epoch": 1.6075309577963104,
      "grad_norm": 1.8045384883880615,
      "learning_rate": 3.949367088607595e-06,
      "loss": 0.07704678177833557,
      "num_input_tokens_seen": 3002560,
      "step": 1591,
      "train_runtime": 4354.8159,
      "train_tokens_per_second": 689.48
    },
    {
      "epoch": 1.608541824614607,
      "grad_norm": 1.6605801582336426,
      "learning_rate": 3.939240506329114e-06,
      "loss": 0.059775568544864655,
      "num_input_tokens_seen": 3004104,
      "step": 1592,
      "train_runtime": 4356.9849,
      "train_tokens_per_second": 689.491
    },
    {
      "epoch": 1.6095526914329037,
      "grad_norm": 1.8942264318466187,
      "learning_rate": 3.929113924050633e-06,
      "loss": 0.05962919443845749,
      "num_input_tokens_seen": 3005420,
      "step": 1593,
      "train_runtime": 4359.0408,
      "train_tokens_per_second": 689.468
    },
    {
      "epoch": 1.6105635582512003,
      "grad_norm": 2.2421839237213135,
      "learning_rate": 3.9189873417721525e-06,
      "loss": 0.11422710865736008,
      "num_input_tokens_seen": 3006926,
      "step": 1594,
      "train_runtime": 4361.1646,
      "train_tokens_per_second": 689.478
    },
    {
      "epoch": 1.611574425069497,
      "grad_norm": 1.8324493169784546,
      "learning_rate": 3.908860759493672e-06,
      "loss": 0.06773149222135544,
      "num_input_tokens_seen": 3008578,
      "step": 1595,
      "train_runtime": 4363.313,
      "train_tokens_per_second": 689.517
    },
    {
      "epoch": 1.6125852918877936,
      "grad_norm": 1.7596731185913086,
      "learning_rate": 3.898734177215191e-06,
      "loss": 0.08236124366521835,
      "num_input_tokens_seen": 3010530,
      "step": 1596,
      "train_runtime": 4365.5348,
      "train_tokens_per_second": 689.613
    },
    {
      "epoch": 1.6135961587060905,
      "grad_norm": 1.3826698064804077,
      "learning_rate": 3.888607594936709e-06,
      "loss": 0.06851591914892197,
      "num_input_tokens_seen": 3012808,
      "step": 1597,
      "train_runtime": 4367.8733,
      "train_tokens_per_second": 689.765
    },
    {
      "epoch": 1.6146070255243872,
      "grad_norm": 1.873353362083435,
      "learning_rate": 3.878481012658228e-06,
      "loss": 0.09259966760873795,
      "num_input_tokens_seen": 3014468,
      "step": 1598,
      "train_runtime": 4369.9937,
      "train_tokens_per_second": 689.811
    },
    {
      "epoch": 1.6156178923426838,
      "grad_norm": 1.746298909187317,
      "learning_rate": 3.868354430379747e-06,
      "loss": 0.06972785294055939,
      "num_input_tokens_seen": 3015760,
      "step": 1599,
      "train_runtime": 4372.0915,
      "train_tokens_per_second": 689.775
    },
    {
      "epoch": 1.6166287591609807,
      "grad_norm": 1.7015008926391602,
      "learning_rate": 3.858227848101266e-06,
      "loss": 0.06295757740736008,
      "num_input_tokens_seen": 3017286,
      "step": 1600,
      "train_runtime": 4374.2064,
      "train_tokens_per_second": 689.79
    },
    {
      "epoch": 1.6176396259792774,
      "grad_norm": 2.12400221824646,
      "learning_rate": 3.848101265822785e-06,
      "loss": 0.10723674297332764,
      "num_input_tokens_seen": 3019240,
      "step": 1601,
      "train_runtime": 4376.504,
      "train_tokens_per_second": 689.875
    },
    {
      "epoch": 1.618650492797574,
      "grad_norm": 1.5593434572219849,
      "learning_rate": 3.837974683544304e-06,
      "loss": 0.07614826411008835,
      "num_input_tokens_seen": 3021418,
      "step": 1602,
      "train_runtime": 4378.8848,
      "train_tokens_per_second": 689.997
    },
    {
      "epoch": 1.6196613596158707,
      "grad_norm": 1.7512574195861816,
      "learning_rate": 3.827848101265823e-06,
      "loss": 0.06531091034412384,
      "num_input_tokens_seen": 3023596,
      "step": 1603,
      "train_runtime": 4381.2168,
      "train_tokens_per_second": 690.127
    },
    {
      "epoch": 1.6206722264341673,
      "grad_norm": 1.5757083892822266,
      "learning_rate": 3.8177215189873425e-06,
      "loss": 0.06421954929828644,
      "num_input_tokens_seen": 3025522,
      "step": 1604,
      "train_runtime": 4383.5067,
      "train_tokens_per_second": 690.206
    },
    {
      "epoch": 1.621683093252464,
      "grad_norm": 1.5954736471176147,
      "learning_rate": 3.807594936708861e-06,
      "loss": 0.06996570527553558,
      "num_input_tokens_seen": 3027840,
      "step": 1605,
      "train_runtime": 4385.8104,
      "train_tokens_per_second": 690.372
    },
    {
      "epoch": 1.6226939600707606,
      "grad_norm": 1.5947481393814087,
      "learning_rate": 3.7974683544303802e-06,
      "loss": 0.07503378391265869,
      "num_input_tokens_seen": 3029608,
      "step": 1606,
      "train_runtime": 4387.99,
      "train_tokens_per_second": 690.432
    },
    {
      "epoch": 1.6237048268890573,
      "grad_norm": 1.5519132614135742,
      "learning_rate": 3.787341772151899e-06,
      "loss": 0.05764038860797882,
      "num_input_tokens_seen": 3031396,
      "step": 1607,
      "train_runtime": 4390.2205,
      "train_tokens_per_second": 690.488
    },
    {
      "epoch": 1.624715693707354,
      "grad_norm": 1.7704308032989502,
      "learning_rate": 3.7772151898734184e-06,
      "loss": 0.07108024507761002,
      "num_input_tokens_seen": 3033048,
      "step": 1608,
      "train_runtime": 4392.4485,
      "train_tokens_per_second": 690.514
    },
    {
      "epoch": 1.6257265605256508,
      "grad_norm": 1.584835410118103,
      "learning_rate": 3.767088607594937e-06,
      "loss": 0.07015027105808258,
      "num_input_tokens_seen": 3035034,
      "step": 1609,
      "train_runtime": 4394.7266,
      "train_tokens_per_second": 690.608
    },
    {
      "epoch": 1.6267374273439474,
      "grad_norm": 1.7231712341308594,
      "learning_rate": 3.756962025316456e-06,
      "loss": 0.07047292590141296,
      "num_input_tokens_seen": 3036868,
      "step": 1610,
      "train_runtime": 4396.9995,
      "train_tokens_per_second": 690.668
    },
    {
      "epoch": 1.627748294162244,
      "grad_norm": 1.724179744720459,
      "learning_rate": 3.746835443037975e-06,
      "loss": 0.07304114103317261,
      "num_input_tokens_seen": 3038720,
      "step": 1611,
      "train_runtime": 4399.2715,
      "train_tokens_per_second": 690.733
    },
    {
      "epoch": 1.628759160980541,
      "grad_norm": 2.4093263149261475,
      "learning_rate": 3.7367088607594943e-06,
      "loss": 0.09261861443519592,
      "num_input_tokens_seen": 3040610,
      "step": 1612,
      "train_runtime": 4401.5036,
      "train_tokens_per_second": 690.812
    },
    {
      "epoch": 1.6297700277988376,
      "grad_norm": 1.445542573928833,
      "learning_rate": 3.726582278481013e-06,
      "loss": 0.057139504700899124,
      "num_input_tokens_seen": 3042618,
      "step": 1613,
      "train_runtime": 4403.7229,
      "train_tokens_per_second": 690.919
    },
    {
      "epoch": 1.6307808946171343,
      "grad_norm": 1.726773738861084,
      "learning_rate": 3.716455696202532e-06,
      "loss": 0.07488808780908585,
      "num_input_tokens_seen": 3044814,
      "step": 1614,
      "train_runtime": 4406.0114,
      "train_tokens_per_second": 691.059
    },
    {
      "epoch": 1.631791761435431,
      "grad_norm": 2.161237955093384,
      "learning_rate": 3.7063291139240507e-06,
      "loss": 0.07743383944034576,
      "num_input_tokens_seen": 3046482,
      "step": 1615,
      "train_runtime": 4408.1552,
      "train_tokens_per_second": 691.101
    },
    {
      "epoch": 1.6328026282537276,
      "grad_norm": 1.8333536386489868,
      "learning_rate": 3.6962025316455702e-06,
      "loss": 0.07610010355710983,
      "num_input_tokens_seen": 3048240,
      "step": 1616,
      "train_runtime": 4410.3365,
      "train_tokens_per_second": 691.158
    },
    {
      "epoch": 1.6338134950720242,
      "grad_norm": 1.3004517555236816,
      "learning_rate": 3.686075949367089e-06,
      "loss": 0.0713384598493576,
      "num_input_tokens_seen": 3050738,
      "step": 1617,
      "train_runtime": 4412.6996,
      "train_tokens_per_second": 691.354
    },
    {
      "epoch": 1.6348243618903209,
      "grad_norm": 1.4625111818313599,
      "learning_rate": 3.675949367088608e-06,
      "loss": 0.06168478727340698,
      "num_input_tokens_seen": 3052798,
      "step": 1618,
      "train_runtime": 4414.9383,
      "train_tokens_per_second": 691.47
    },
    {
      "epoch": 1.6358352287086175,
      "grad_norm": 1.6302261352539062,
      "learning_rate": 3.665822784810127e-06,
      "loss": 0.05223661661148071,
      "num_input_tokens_seen": 3054350,
      "step": 1619,
      "train_runtime": 4417.153,
      "train_tokens_per_second": 691.475
    },
    {
      "epoch": 1.6368460955269142,
      "grad_norm": 1.4504172801971436,
      "learning_rate": 3.655696202531646e-06,
      "loss": 0.05241897702217102,
      "num_input_tokens_seen": 3056636,
      "step": 1620,
      "train_runtime": 4419.4538,
      "train_tokens_per_second": 691.632
    },
    {
      "epoch": 1.637856962345211,
      "grad_norm": 1.685484766960144,
      "learning_rate": 3.645569620253165e-06,
      "loss": 0.06576771289110184,
      "num_input_tokens_seen": 3058180,
      "step": 1621,
      "train_runtime": 4428.1924,
      "train_tokens_per_second": 690.616
    },
    {
      "epoch": 1.6388678291635077,
      "grad_norm": 1.4740047454833984,
      "learning_rate": 3.6354430379746835e-06,
      "loss": 0.05852600559592247,
      "num_input_tokens_seen": 3059906,
      "step": 1622,
      "train_runtime": 4430.4169,
      "train_tokens_per_second": 690.659
    },
    {
      "epoch": 1.6398786959818044,
      "grad_norm": 1.201348900794983,
      "learning_rate": 3.625316455696203e-06,
      "loss": 0.042996060103178024,
      "num_input_tokens_seen": 3062134,
      "step": 1623,
      "train_runtime": 4432.7694,
      "train_tokens_per_second": 690.795
    },
    {
      "epoch": 1.6408895628001012,
      "grad_norm": 1.4870134592056274,
      "learning_rate": 3.6151898734177216e-06,
      "loss": 0.059534698724746704,
      "num_input_tokens_seen": 3063896,
      "step": 1624,
      "train_runtime": 4435.0148,
      "train_tokens_per_second": 690.842
    },
    {
      "epoch": 1.6419004296183979,
      "grad_norm": 1.9917396306991577,
      "learning_rate": 3.6050632911392407e-06,
      "loss": 0.08953356742858887,
      "num_input_tokens_seen": 3065622,
      "step": 1625,
      "train_runtime": 4437.2786,
      "train_tokens_per_second": 690.879
    },
    {
      "epoch": 1.6429112964366945,
      "grad_norm": 1.9785447120666504,
      "learning_rate": 3.5949367088607594e-06,
      "loss": 0.06318928301334381,
      "num_input_tokens_seen": 3067090,
      "step": 1626,
      "train_runtime": 4439.3804,
      "train_tokens_per_second": 690.882
    },
    {
      "epoch": 1.6439221632549912,
      "grad_norm": 1.4886456727981567,
      "learning_rate": 3.584810126582279e-06,
      "loss": 0.06752782315015793,
      "num_input_tokens_seen": 3069452,
      "step": 1627,
      "train_runtime": 4441.6769,
      "train_tokens_per_second": 691.057
    },
    {
      "epoch": 1.6449330300732878,
      "grad_norm": 1.9452000856399536,
      "learning_rate": 3.5746835443037975e-06,
      "loss": 0.06755784898996353,
      "num_input_tokens_seen": 3071050,
      "step": 1628,
      "train_runtime": 4443.8489,
      "train_tokens_per_second": 691.079
    },
    {
      "epoch": 1.6459438968915845,
      "grad_norm": 2.1163554191589355,
      "learning_rate": 3.5645569620253166e-06,
      "loss": 0.08850133419036865,
      "num_input_tokens_seen": 3072606,
      "step": 1629,
      "train_runtime": 4445.9685,
      "train_tokens_per_second": 691.099
    },
    {
      "epoch": 1.6469547637098811,
      "grad_norm": 2.0030453205108643,
      "learning_rate": 3.5544303797468353e-06,
      "loss": 0.06304336339235306,
      "num_input_tokens_seen": 3074244,
      "step": 1630,
      "train_runtime": 4448.1109,
      "train_tokens_per_second": 691.135
    },
    {
      "epoch": 1.6479656305281778,
      "grad_norm": 1.8989089727401733,
      "learning_rate": 3.544303797468355e-06,
      "loss": 0.08454688638448715,
      "num_input_tokens_seen": 3075800,
      "step": 1631,
      "train_runtime": 4450.3088,
      "train_tokens_per_second": 691.143
    },
    {
      "epoch": 1.6489764973464744,
      "grad_norm": 1.5028738975524902,
      "learning_rate": 3.5341772151898735e-06,
      "loss": 0.04924144223332405,
      "num_input_tokens_seen": 3077748,
      "step": 1632,
      "train_runtime": 4452.5601,
      "train_tokens_per_second": 691.231
    },
    {
      "epoch": 1.6499873641647713,
      "grad_norm": 1.6033862829208374,
      "learning_rate": 3.5240506329113925e-06,
      "loss": 0.06796707212924957,
      "num_input_tokens_seen": 3079672,
      "step": 1633,
      "train_runtime": 4454.7921,
      "train_tokens_per_second": 691.317
    },
    {
      "epoch": 1.650998230983068,
      "grad_norm": 1.9249001741409302,
      "learning_rate": 3.5139240506329116e-06,
      "loss": 0.08959044516086578,
      "num_input_tokens_seen": 3081838,
      "step": 1634,
      "train_runtime": 4457.0619,
      "train_tokens_per_second": 691.451
    },
    {
      "epoch": 1.6520090978013646,
      "grad_norm": 2.064527750015259,
      "learning_rate": 3.5037974683544307e-06,
      "loss": 0.09661225974559784,
      "num_input_tokens_seen": 3084034,
      "step": 1635,
      "train_runtime": 4459.3265,
      "train_tokens_per_second": 691.592
    },
    {
      "epoch": 1.6530199646196615,
      "grad_norm": 1.7836114168167114,
      "learning_rate": 3.4936708860759494e-06,
      "loss": 0.08342694491147995,
      "num_input_tokens_seen": 3085876,
      "step": 1636,
      "train_runtime": 4461.5649,
      "train_tokens_per_second": 691.658
    },
    {
      "epoch": 1.6540308314379581,
      "grad_norm": 1.934723138809204,
      "learning_rate": 3.4835443037974685e-06,
      "loss": 0.09056703746318817,
      "num_input_tokens_seen": 3087968,
      "step": 1637,
      "train_runtime": 4463.8005,
      "train_tokens_per_second": 691.78
    },
    {
      "epoch": 1.6550416982562548,
      "grad_norm": 1.387448787689209,
      "learning_rate": 3.4734177215189875e-06,
      "loss": 0.0635710060596466,
      "num_input_tokens_seen": 3089862,
      "step": 1638,
      "train_runtime": 4466.0482,
      "train_tokens_per_second": 691.856
    },
    {
      "epoch": 1.6560525650745515,
      "grad_norm": 1.956931233406067,
      "learning_rate": 3.4632911392405066e-06,
      "loss": 0.08530846238136292,
      "num_input_tokens_seen": 3091530,
      "step": 1639,
      "train_runtime": 4468.1598,
      "train_tokens_per_second": 691.902
    },
    {
      "epoch": 1.657063431892848,
      "grad_norm": 1.7151684761047363,
      "learning_rate": 3.4531645569620253e-06,
      "loss": 0.06122661754488945,
      "num_input_tokens_seen": 3093300,
      "step": 1640,
      "train_runtime": 4470.3079,
      "train_tokens_per_second": 691.966
    },
    {
      "epoch": 1.6580742987111448,
      "grad_norm": 1.5635318756103516,
      "learning_rate": 3.443037974683545e-06,
      "loss": 0.08052826672792435,
      "num_input_tokens_seen": 3095484,
      "step": 1641,
      "train_runtime": 4472.5779,
      "train_tokens_per_second": 692.103
    },
    {
      "epoch": 1.6590851655294414,
      "grad_norm": 1.4197895526885986,
      "learning_rate": 3.4329113924050635e-06,
      "loss": 0.05699260160326958,
      "num_input_tokens_seen": 3097440,
      "step": 1642,
      "train_runtime": 4474.7916,
      "train_tokens_per_second": 692.198
    },
    {
      "epoch": 1.660096032347738,
      "grad_norm": 1.5599781274795532,
      "learning_rate": 3.4227848101265825e-06,
      "loss": 0.062324777245521545,
      "num_input_tokens_seen": 3099356,
      "step": 1643,
      "train_runtime": 4476.9946,
      "train_tokens_per_second": 692.285
    },
    {
      "epoch": 1.6611068991660347,
      "grad_norm": 1.7807284593582153,
      "learning_rate": 3.412658227848101e-06,
      "loss": 0.08422867953777313,
      "num_input_tokens_seen": 3101190,
      "step": 1644,
      "train_runtime": 4479.1684,
      "train_tokens_per_second": 692.358
    },
    {
      "epoch": 1.6621177659843316,
      "grad_norm": 1.3435440063476562,
      "learning_rate": 3.4025316455696207e-06,
      "loss": 0.06103749945759773,
      "num_input_tokens_seen": 3103408,
      "step": 1645,
      "train_runtime": 4481.4558,
      "train_tokens_per_second": 692.5
    },
    {
      "epoch": 1.6631286328026282,
      "grad_norm": 1.5479793548583984,
      "learning_rate": 3.3924050632911394e-06,
      "loss": 0.0730970948934555,
      "num_input_tokens_seen": 3105644,
      "step": 1646,
      "train_runtime": 4483.7199,
      "train_tokens_per_second": 692.649
    },
    {
      "epoch": 1.664139499620925,
      "grad_norm": 1.6312743425369263,
      "learning_rate": 3.3822784810126585e-06,
      "loss": 0.06483624875545502,
      "num_input_tokens_seen": 3107626,
      "step": 1647,
      "train_runtime": 4485.9265,
      "train_tokens_per_second": 692.75
    },
    {
      "epoch": 1.6651503664392218,
      "grad_norm": 1.7495090961456299,
      "learning_rate": 3.372151898734177e-06,
      "loss": 0.07979393005371094,
      "num_input_tokens_seen": 3109686,
      "step": 1648,
      "train_runtime": 4488.168,
      "train_tokens_per_second": 692.863
    },
    {
      "epoch": 1.6661612332575184,
      "grad_norm": 1.8143082857131958,
      "learning_rate": 3.3620253164556966e-06,
      "loss": 0.08356431871652603,
      "num_input_tokens_seen": 3111442,
      "step": 1649,
      "train_runtime": 4490.3453,
      "train_tokens_per_second": 692.918
    },
    {
      "epoch": 1.667172100075815,
      "grad_norm": 1.535099744796753,
      "learning_rate": 3.3518987341772153e-06,
      "loss": 0.06810334324836731,
      "num_input_tokens_seen": 3113542,
      "step": 1650,
      "train_runtime": 4492.6074,
      "train_tokens_per_second": 693.037
    },
    {
      "epoch": 1.6681829668941117,
      "grad_norm": 1.5244773626327515,
      "learning_rate": 3.3417721518987344e-06,
      "loss": 0.06630674004554749,
      "num_input_tokens_seen": 3115364,
      "step": 1651,
      "train_runtime": 4501.4263,
      "train_tokens_per_second": 692.084
    },
    {
      "epoch": 1.6691938337124084,
      "grad_norm": 1.589932918548584,
      "learning_rate": 3.331645569620253e-06,
      "loss": 0.06367498636245728,
      "num_input_tokens_seen": 3117738,
      "step": 1652,
      "train_runtime": 4503.7675,
      "train_tokens_per_second": 692.251
    },
    {
      "epoch": 1.670204700530705,
      "grad_norm": 1.299252986907959,
      "learning_rate": 3.3215189873417725e-06,
      "loss": 0.05266419053077698,
      "num_input_tokens_seen": 3119876,
      "step": 1653,
      "train_runtime": 4506.0313,
      "train_tokens_per_second": 692.378
    },
    {
      "epoch": 1.6712155673490017,
      "grad_norm": 1.5991474390029907,
      "learning_rate": 3.311392405063291e-06,
      "loss": 0.059744007885456085,
      "num_input_tokens_seen": 3121986,
      "step": 1654,
      "train_runtime": 4508.2753,
      "train_tokens_per_second": 692.501
    },
    {
      "epoch": 1.6722264341672983,
      "grad_norm": 1.7557860612869263,
      "learning_rate": 3.3012658227848103e-06,
      "loss": 0.09023163467645645,
      "num_input_tokens_seen": 3123822,
      "step": 1655,
      "train_runtime": 4510.4883,
      "train_tokens_per_second": 692.568
    },
    {
      "epoch": 1.6732373009855952,
      "grad_norm": 1.4985078573226929,
      "learning_rate": 3.2911392405063294e-06,
      "loss": 0.049209535121917725,
      "num_input_tokens_seen": 3125864,
      "step": 1656,
      "train_runtime": 4512.7028,
      "train_tokens_per_second": 692.681
    },
    {
      "epoch": 1.6742481678038919,
      "grad_norm": 1.7229870557785034,
      "learning_rate": 3.2810126582278485e-06,
      "loss": 0.04763895273208618,
      "num_input_tokens_seen": 3127242,
      "step": 1657,
      "train_runtime": 4514.7677,
      "train_tokens_per_second": 692.67
    },
    {
      "epoch": 1.6752590346221885,
      "grad_norm": 1.9979041814804077,
      "learning_rate": 3.270886075949367e-06,
      "loss": 0.07513368129730225,
      "num_input_tokens_seen": 3128716,
      "step": 1658,
      "train_runtime": 4516.8916,
      "train_tokens_per_second": 692.67
    },
    {
      "epoch": 1.6762699014404854,
      "grad_norm": 2.3191444873809814,
      "learning_rate": 3.260759493670886e-06,
      "loss": 0.0995558649301529,
      "num_input_tokens_seen": 3130486,
      "step": 1659,
      "train_runtime": 4519.0803,
      "train_tokens_per_second": 692.726
    },
    {
      "epoch": 1.677280768258782,
      "grad_norm": 1.8650801181793213,
      "learning_rate": 3.2506329113924053e-06,
      "loss": 0.05107400566339493,
      "num_input_tokens_seen": 3132030,
      "step": 1660,
      "train_runtime": 4521.2069,
      "train_tokens_per_second": 692.742
    },
    {
      "epoch": 1.6782916350770787,
      "grad_norm": 1.7372885942459106,
      "learning_rate": 3.2405063291139244e-06,
      "loss": 0.06642575562000275,
      "num_input_tokens_seen": 3133638,
      "step": 1661,
      "train_runtime": 4523.358,
      "train_tokens_per_second": 692.768
    },
    {
      "epoch": 1.6793025018953753,
      "grad_norm": 2.25758957862854,
      "learning_rate": 3.230379746835443e-06,
      "loss": 0.0813010036945343,
      "num_input_tokens_seen": 3135314,
      "step": 1662,
      "train_runtime": 4525.5291,
      "train_tokens_per_second": 692.806
    },
    {
      "epoch": 1.680313368713672,
      "grad_norm": 1.727011799812317,
      "learning_rate": 3.2202531645569625e-06,
      "loss": 0.08289699256420135,
      "num_input_tokens_seen": 3137340,
      "step": 1663,
      "train_runtime": 4527.7888,
      "train_tokens_per_second": 692.908
    },
    {
      "epoch": 1.6813242355319686,
      "grad_norm": 1.6658594608306885,
      "learning_rate": 3.210126582278481e-06,
      "loss": 0.0790754035115242,
      "num_input_tokens_seen": 3139916,
      "step": 1664,
      "train_runtime": 4530.1336,
      "train_tokens_per_second": 693.118
    },
    {
      "epoch": 1.6823351023502653,
      "grad_norm": 1.7198365926742554,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 0.07097277790307999,
      "num_input_tokens_seen": 3141946,
      "step": 1665,
      "train_runtime": 4532.3675,
      "train_tokens_per_second": 693.224
    },
    {
      "epoch": 1.683345969168562,
      "grad_norm": 1.7861515283584595,
      "learning_rate": 3.189873417721519e-06,
      "loss": 0.06614462286233902,
      "num_input_tokens_seen": 3143520,
      "step": 1666,
      "train_runtime": 4534.4948,
      "train_tokens_per_second": 693.246
    },
    {
      "epoch": 1.6843568359868586,
      "grad_norm": 1.7967497110366821,
      "learning_rate": 3.1797468354430384e-06,
      "loss": 0.05091815069317818,
      "num_input_tokens_seen": 3145024,
      "step": 1667,
      "train_runtime": 4536.6193,
      "train_tokens_per_second": 693.253
    },
    {
      "epoch": 1.6853677028051555,
      "grad_norm": 2.064927101135254,
      "learning_rate": 3.169620253164557e-06,
      "loss": 0.09674140065908432,
      "num_input_tokens_seen": 3146530,
      "step": 1668,
      "train_runtime": 4538.7116,
      "train_tokens_per_second": 693.265
    },
    {
      "epoch": 1.6863785696234521,
      "grad_norm": 1.603570818901062,
      "learning_rate": 3.159493670886076e-06,
      "loss": 0.06721428036689758,
      "num_input_tokens_seen": 3148500,
      "step": 1669,
      "train_runtime": 4540.9123,
      "train_tokens_per_second": 693.363
    },
    {
      "epoch": 1.6873894364417488,
      "grad_norm": 1.5717071294784546,
      "learning_rate": 3.149367088607595e-06,
      "loss": 0.07083986699581146,
      "num_input_tokens_seen": 3150822,
      "step": 1670,
      "train_runtime": 4543.2038,
      "train_tokens_per_second": 693.524
    },
    {
      "epoch": 1.6884003032600456,
      "grad_norm": 2.09252667427063,
      "learning_rate": 3.1392405063291144e-06,
      "loss": 0.07435572892427444,
      "num_input_tokens_seen": 3152442,
      "step": 1671,
      "train_runtime": 4545.3557,
      "train_tokens_per_second": 693.552
    },
    {
      "epoch": 1.6894111700783423,
      "grad_norm": 1.8658822774887085,
      "learning_rate": 3.129113924050633e-06,
      "loss": 0.08821466565132141,
      "num_input_tokens_seen": 3154712,
      "step": 1672,
      "train_runtime": 4547.6472,
      "train_tokens_per_second": 693.702
    },
    {
      "epoch": 1.690422036896639,
      "grad_norm": 2.1665990352630615,
      "learning_rate": 3.118987341772152e-06,
      "loss": 0.057355549186468124,
      "num_input_tokens_seen": 3156586,
      "step": 1673,
      "train_runtime": 4549.8184,
      "train_tokens_per_second": 693.783
    },
    {
      "epoch": 1.6914329037149356,
      "grad_norm": 1.779031753540039,
      "learning_rate": 3.1088607594936708e-06,
      "loss": 0.08638505637645721,
      "num_input_tokens_seen": 3158526,
      "step": 1674,
      "train_runtime": 4552.0236,
      "train_tokens_per_second": 693.873
    },
    {
      "epoch": 1.6924437705332323,
      "grad_norm": 1.9613770246505737,
      "learning_rate": 3.0987341772151903e-06,
      "loss": 0.08512400090694427,
      "num_input_tokens_seen": 3160372,
      "step": 1675,
      "train_runtime": 4554.2366,
      "train_tokens_per_second": 693.941
    },
    {
      "epoch": 1.693454637351529,
      "grad_norm": 1.8615342378616333,
      "learning_rate": 3.088607594936709e-06,
      "loss": 0.07008308172225952,
      "num_input_tokens_seen": 3162204,
      "step": 1676,
      "train_runtime": 4556.4188,
      "train_tokens_per_second": 694.011
    },
    {
      "epoch": 1.6944655041698256,
      "grad_norm": 1.9587600231170654,
      "learning_rate": 3.078481012658228e-06,
      "loss": 0.07925236970186234,
      "num_input_tokens_seen": 3164144,
      "step": 1677,
      "train_runtime": 4558.6777,
      "train_tokens_per_second": 694.092
    },
    {
      "epoch": 1.6954763709881222,
      "grad_norm": 2.0435128211975098,
      "learning_rate": 3.068354430379747e-06,
      "loss": 0.05458016321063042,
      "num_input_tokens_seen": 3165916,
      "step": 1678,
      "train_runtime": 4560.8526,
      "train_tokens_per_second": 694.15
    },
    {
      "epoch": 1.6964872378064189,
      "grad_norm": 1.8596713542938232,
      "learning_rate": 3.058227848101266e-06,
      "loss": 0.0700383260846138,
      "num_input_tokens_seen": 3167602,
      "step": 1679,
      "train_runtime": 4562.9963,
      "train_tokens_per_second": 694.193
    },
    {
      "epoch": 1.6974981046247157,
      "grad_norm": 1.6042331457138062,
      "learning_rate": 3.048101265822785e-06,
      "loss": 0.07046147435903549,
      "num_input_tokens_seen": 3169826,
      "step": 1680,
      "train_runtime": 4565.2614,
      "train_tokens_per_second": 694.336
    },
    {
      "epoch": 1.6985089714430124,
      "grad_norm": 1.3264026641845703,
      "learning_rate": 3.037974683544304e-06,
      "loss": 0.043173111975193024,
      "num_input_tokens_seen": 3171628,
      "step": 1681,
      "train_runtime": 4574.0294,
      "train_tokens_per_second": 693.399
    },
    {
      "epoch": 1.699519838261309,
      "grad_norm": 1.794228434562683,
      "learning_rate": 3.027848101265823e-06,
      "loss": 0.07822631299495697,
      "num_input_tokens_seen": 3173482,
      "step": 1682,
      "train_runtime": 4576.2654,
      "train_tokens_per_second": 693.465
    },
    {
      "epoch": 1.700530705079606,
      "grad_norm": 1.9968194961547852,
      "learning_rate": 3.017721518987342e-06,
      "loss": 0.08686088770627975,
      "num_input_tokens_seen": 3175088,
      "step": 1683,
      "train_runtime": 4578.4004,
      "train_tokens_per_second": 693.493
    },
    {
      "epoch": 1.7015415718979026,
      "grad_norm": 1.5558940172195435,
      "learning_rate": 3.0075949367088608e-06,
      "loss": 0.07532624900341034,
      "num_input_tokens_seen": 3177244,
      "step": 1684,
      "train_runtime": 4580.6711,
      "train_tokens_per_second": 693.62
    },
    {
      "epoch": 1.7025524387161992,
      "grad_norm": 1.4439938068389893,
      "learning_rate": 2.9974683544303803e-06,
      "loss": 0.051425449550151825,
      "num_input_tokens_seen": 3178878,
      "step": 1685,
      "train_runtime": 4582.8741,
      "train_tokens_per_second": 693.643
    },
    {
      "epoch": 1.7035633055344959,
      "grad_norm": 1.6569932699203491,
      "learning_rate": 2.987341772151899e-06,
      "loss": 0.07870493829250336,
      "num_input_tokens_seen": 3181054,
      "step": 1686,
      "train_runtime": 4585.1513,
      "train_tokens_per_second": 693.773
    },
    {
      "epoch": 1.7045741723527925,
      "grad_norm": 1.6693202257156372,
      "learning_rate": 2.977215189873418e-06,
      "loss": 0.08688843250274658,
      "num_input_tokens_seen": 3182894,
      "step": 1687,
      "train_runtime": 4587.3665,
      "train_tokens_per_second": 693.839
    },
    {
      "epoch": 1.7055850391710892,
      "grad_norm": 1.738955020904541,
      "learning_rate": 2.9670886075949367e-06,
      "loss": 0.07897424697875977,
      "num_input_tokens_seen": 3185000,
      "step": 1688,
      "train_runtime": 4589.6253,
      "train_tokens_per_second": 693.956
    },
    {
      "epoch": 1.7065959059893858,
      "grad_norm": 1.9710723161697388,
      "learning_rate": 2.956962025316456e-06,
      "loss": 0.06945981830358505,
      "num_input_tokens_seen": 3186718,
      "step": 1689,
      "train_runtime": 4591.8036,
      "train_tokens_per_second": 694.001
    },
    {
      "epoch": 1.7076067728076825,
      "grad_norm": 2.2010416984558105,
      "learning_rate": 2.946835443037975e-06,
      "loss": 0.077201247215271,
      "num_input_tokens_seen": 3188424,
      "step": 1690,
      "train_runtime": 4593.9774,
      "train_tokens_per_second": 694.044
    },
    {
      "epoch": 1.7086176396259791,
      "grad_norm": 1.2872505187988281,
      "learning_rate": 2.936708860759494e-06,
      "loss": 0.050885725766420364,
      "num_input_tokens_seen": 3190914,
      "step": 1691,
      "train_runtime": 4596.3587,
      "train_tokens_per_second": 694.226
    },
    {
      "epoch": 1.709628506444276,
      "grad_norm": 1.5386303663253784,
      "learning_rate": 2.9265822784810126e-06,
      "loss": 0.06214267387986183,
      "num_input_tokens_seen": 3192874,
      "step": 1692,
      "train_runtime": 4598.6118,
      "train_tokens_per_second": 694.313
    },
    {
      "epoch": 1.7106393732625726,
      "grad_norm": 1.4623299837112427,
      "learning_rate": 2.916455696202532e-06,
      "loss": 0.06471185386180878,
      "num_input_tokens_seen": 3194340,
      "step": 1693,
      "train_runtime": 4600.727,
      "train_tokens_per_second": 694.312
    },
    {
      "epoch": 1.7116502400808693,
      "grad_norm": 1.3713880777359009,
      "learning_rate": 2.9063291139240508e-06,
      "loss": 0.052813608199357986,
      "num_input_tokens_seen": 3196300,
      "step": 1694,
      "train_runtime": 4602.9378,
      "train_tokens_per_second": 694.404
    },
    {
      "epoch": 1.7126611068991662,
      "grad_norm": 1.6384897232055664,
      "learning_rate": 2.89620253164557e-06,
      "loss": 0.07733742892742157,
      "num_input_tokens_seen": 3198282,
      "step": 1695,
      "train_runtime": 4605.1528,
      "train_tokens_per_second": 694.501
    },
    {
      "epoch": 1.7136719737174628,
      "grad_norm": 2.3512582778930664,
      "learning_rate": 2.8860759493670885e-06,
      "loss": 0.09398391842842102,
      "num_input_tokens_seen": 3199958,
      "step": 1696,
      "train_runtime": 4607.2801,
      "train_tokens_per_second": 694.544
    },
    {
      "epoch": 1.7146828405357595,
      "grad_norm": 1.9150465726852417,
      "learning_rate": 2.875949367088608e-06,
      "loss": 0.0935385674238205,
      "num_input_tokens_seen": 3201796,
      "step": 1697,
      "train_runtime": 4609.4598,
      "train_tokens_per_second": 694.614
    },
    {
      "epoch": 1.7156937073540561,
      "grad_norm": 1.784387230873108,
      "learning_rate": 2.8658227848101267e-06,
      "loss": 0.07881203293800354,
      "num_input_tokens_seen": 3203628,
      "step": 1698,
      "train_runtime": 4611.7597,
      "train_tokens_per_second": 694.665
    },
    {
      "epoch": 1.7167045741723528,
      "grad_norm": 1.836754322052002,
      "learning_rate": 2.8556962025316458e-06,
      "loss": 0.08432598412036896,
      "num_input_tokens_seen": 3205424,
      "step": 1699,
      "train_runtime": 4613.9619,
      "train_tokens_per_second": 694.723
    },
    {
      "epoch": 1.7177154409906494,
      "grad_norm": 1.6243864297866821,
      "learning_rate": 2.845569620253165e-06,
      "loss": 0.07003642618656158,
      "num_input_tokens_seen": 3207514,
      "step": 1700,
      "train_runtime": 4616.1911,
      "train_tokens_per_second": 694.84
    },
    {
      "epoch": 1.718726307808946,
      "grad_norm": 2.060173988342285,
      "learning_rate": 2.835443037974684e-06,
      "loss": 0.09881354123353958,
      "num_input_tokens_seen": 3209430,
      "step": 1701,
      "train_runtime": 4618.3895,
      "train_tokens_per_second": 694.924
    },
    {
      "epoch": 1.7197371746272427,
      "grad_norm": 1.5945155620574951,
      "learning_rate": 2.8253164556962026e-06,
      "loss": 0.06285209953784943,
      "num_input_tokens_seen": 3211230,
      "step": 1702,
      "train_runtime": 4620.5583,
      "train_tokens_per_second": 694.987
    },
    {
      "epoch": 1.7207480414455394,
      "grad_norm": 1.4761182069778442,
      "learning_rate": 2.8151898734177217e-06,
      "loss": 0.07465441524982452,
      "num_input_tokens_seen": 3213296,
      "step": 1703,
      "train_runtime": 4622.7663,
      "train_tokens_per_second": 695.102
    },
    {
      "epoch": 1.7217589082638363,
      "grad_norm": 1.878750205039978,
      "learning_rate": 2.8050632911392408e-06,
      "loss": 0.08725883066654205,
      "num_input_tokens_seen": 3215426,
      "step": 1704,
      "train_runtime": 4625.0287,
      "train_tokens_per_second": 695.223
    },
    {
      "epoch": 1.722769775082133,
      "grad_norm": 1.9042385816574097,
      "learning_rate": 2.79493670886076e-06,
      "loss": 0.07090901583433151,
      "num_input_tokens_seen": 3217030,
      "step": 1705,
      "train_runtime": 4627.15,
      "train_tokens_per_second": 695.251
    },
    {
      "epoch": 1.7237806419004296,
      "grad_norm": 1.5664784908294678,
      "learning_rate": 2.7848101265822785e-06,
      "loss": 0.06554043292999268,
      "num_input_tokens_seen": 3219282,
      "step": 1706,
      "train_runtime": 4629.4175,
      "train_tokens_per_second": 695.397
    },
    {
      "epoch": 1.7247915087187264,
      "grad_norm": 2.254207134246826,
      "learning_rate": 2.774683544303798e-06,
      "loss": 0.06345508992671967,
      "num_input_tokens_seen": 3221324,
      "step": 1707,
      "train_runtime": 4631.7041,
      "train_tokens_per_second": 695.494
    },
    {
      "epoch": 1.725802375537023,
      "grad_norm": 1.6412838697433472,
      "learning_rate": 2.7645569620253167e-06,
      "loss": 0.05187619850039482,
      "num_input_tokens_seen": 3223002,
      "step": 1708,
      "train_runtime": 4633.8926,
      "train_tokens_per_second": 695.528
    },
    {
      "epoch": 1.7268132423553197,
      "grad_norm": 1.8688623905181885,
      "learning_rate": 2.7544303797468358e-06,
      "loss": 0.08315165340900421,
      "num_input_tokens_seen": 3224970,
      "step": 1709,
      "train_runtime": 4636.1099,
      "train_tokens_per_second": 695.62
    },
    {
      "epoch": 1.7278241091736164,
      "grad_norm": 1.593793511390686,
      "learning_rate": 2.7443037974683544e-06,
      "loss": 0.05407699942588806,
      "num_input_tokens_seen": 3226402,
      "step": 1710,
      "train_runtime": 4638.2018,
      "train_tokens_per_second": 695.615
    },
    {
      "epoch": 1.728834975991913,
      "grad_norm": 1.5109833478927612,
      "learning_rate": 2.734177215189874e-06,
      "loss": 0.05802329257130623,
      "num_input_tokens_seen": 3228294,
      "step": 1711,
      "train_runtime": 4646.8163,
      "train_tokens_per_second": 694.732
    },
    {
      "epoch": 1.7298458428102097,
      "grad_norm": 1.7301241159439087,
      "learning_rate": 2.7240506329113926e-06,
      "loss": 0.08108571171760559,
      "num_input_tokens_seen": 3230478,
      "step": 1712,
      "train_runtime": 4649.2404,
      "train_tokens_per_second": 694.84
    },
    {
      "epoch": 1.7308567096285064,
      "grad_norm": 1.6756778955459595,
      "learning_rate": 2.7139240506329117e-06,
      "loss": 0.060624442994594574,
      "num_input_tokens_seen": 3232430,
      "step": 1713,
      "train_runtime": 4651.4335,
      "train_tokens_per_second": 694.932
    },
    {
      "epoch": 1.731867576446803,
      "grad_norm": 1.9304794073104858,
      "learning_rate": 2.7037974683544303e-06,
      "loss": 0.09782436490058899,
      "num_input_tokens_seen": 3234332,
      "step": 1714,
      "train_runtime": 4653.6634,
      "train_tokens_per_second": 695.008
    },
    {
      "epoch": 1.7328784432650999,
      "grad_norm": 1.450645923614502,
      "learning_rate": 2.69367088607595e-06,
      "loss": 0.06790439039468765,
      "num_input_tokens_seen": 3236438,
      "step": 1715,
      "train_runtime": 4655.8586,
      "train_tokens_per_second": 695.132
    },
    {
      "epoch": 1.7338893100833965,
      "grad_norm": 1.8678185939788818,
      "learning_rate": 2.6835443037974685e-06,
      "loss": 0.061199259012937546,
      "num_input_tokens_seen": 3238184,
      "step": 1716,
      "train_runtime": 4658.0395,
      "train_tokens_per_second": 695.182
    },
    {
      "epoch": 1.7349001769016932,
      "grad_norm": 1.601326823234558,
      "learning_rate": 2.6734177215189876e-06,
      "loss": 0.052972547709941864,
      "num_input_tokens_seen": 3239792,
      "step": 1717,
      "train_runtime": 4660.2039,
      "train_tokens_per_second": 695.204
    },
    {
      "epoch": 1.73591104371999,
      "grad_norm": 1.6289334297180176,
      "learning_rate": 2.6632911392405062e-06,
      "loss": 0.07092973589897156,
      "num_input_tokens_seen": 3241846,
      "step": 1718,
      "train_runtime": 4662.4344,
      "train_tokens_per_second": 695.312
    },
    {
      "epoch": 1.7369219105382867,
      "grad_norm": 1.7682119607925415,
      "learning_rate": 2.6531645569620257e-06,
      "loss": 0.059664398431777954,
      "num_input_tokens_seen": 3243330,
      "step": 1719,
      "train_runtime": 4664.5205,
      "train_tokens_per_second": 695.319
    },
    {
      "epoch": 1.7379327773565834,
      "grad_norm": 1.7986894845962524,
      "learning_rate": 2.6430379746835444e-06,
      "loss": 0.05744286626577377,
      "num_input_tokens_seen": 3244718,
      "step": 1720,
      "train_runtime": 4666.6681,
      "train_tokens_per_second": 695.297
    },
    {
      "epoch": 1.73894364417488,
      "grad_norm": 1.805665373802185,
      "learning_rate": 2.6329113924050635e-06,
      "loss": 0.07611143589019775,
      "num_input_tokens_seen": 3246278,
      "step": 1721,
      "train_runtime": 4668.7493,
      "train_tokens_per_second": 695.321
    },
    {
      "epoch": 1.7399545109931767,
      "grad_norm": 2.240658760070801,
      "learning_rate": 2.6227848101265826e-06,
      "loss": 0.10359307378530502,
      "num_input_tokens_seen": 3248164,
      "step": 1722,
      "train_runtime": 4670.9573,
      "train_tokens_per_second": 695.396
    },
    {
      "epoch": 1.7409653778114733,
      "grad_norm": 1.5116084814071655,
      "learning_rate": 2.6126582278481017e-06,
      "loss": 0.06377623975276947,
      "num_input_tokens_seen": 3249786,
      "step": 1723,
      "train_runtime": 4673.1407,
      "train_tokens_per_second": 695.418
    },
    {
      "epoch": 1.74197624462977,
      "grad_norm": 1.8039082288742065,
      "learning_rate": 2.6025316455696203e-06,
      "loss": 0.06861607730388641,
      "num_input_tokens_seen": 3251246,
      "step": 1724,
      "train_runtime": 4675.2583,
      "train_tokens_per_second": 695.415
    },
    {
      "epoch": 1.7429871114480666,
      "grad_norm": 1.8272242546081543,
      "learning_rate": 2.5924050632911394e-06,
      "loss": 0.07011061906814575,
      "num_input_tokens_seen": 3252628,
      "step": 1725,
      "train_runtime": 4677.3422,
      "train_tokens_per_second": 695.401
    },
    {
      "epoch": 1.7439979782663633,
      "grad_norm": 1.44402015209198,
      "learning_rate": 2.5822784810126585e-06,
      "loss": 0.05340193957090378,
      "num_input_tokens_seen": 3254830,
      "step": 1726,
      "train_runtime": 4679.6964,
      "train_tokens_per_second": 695.522
    },
    {
      "epoch": 1.7450088450846601,
      "grad_norm": 1.5405900478363037,
      "learning_rate": 2.5721518987341776e-06,
      "loss": 0.06300503760576248,
      "num_input_tokens_seen": 3256576,
      "step": 1727,
      "train_runtime": 4681.8873,
      "train_tokens_per_second": 695.569
    },
    {
      "epoch": 1.7460197119029568,
      "grad_norm": 1.3495336771011353,
      "learning_rate": 2.5620253164556962e-06,
      "loss": 0.05233839899301529,
      "num_input_tokens_seen": 3258778,
      "step": 1728,
      "train_runtime": 4684.1474,
      "train_tokens_per_second": 695.704
    },
    {
      "epoch": 1.7470305787212534,
      "grad_norm": 1.8360068798065186,
      "learning_rate": 2.5518987341772157e-06,
      "loss": 0.07781828939914703,
      "num_input_tokens_seen": 3260680,
      "step": 1729,
      "train_runtime": 4686.3573,
      "train_tokens_per_second": 695.781
    },
    {
      "epoch": 1.7480414455395503,
      "grad_norm": 1.5859555006027222,
      "learning_rate": 2.5417721518987344e-06,
      "loss": 0.053702931851148605,
      "num_input_tokens_seen": 3262762,
      "step": 1730,
      "train_runtime": 4688.6701,
      "train_tokens_per_second": 695.882
    },
    {
      "epoch": 1.749052312357847,
      "grad_norm": 1.8131262063980103,
      "learning_rate": 2.5316455696202535e-06,
      "loss": 0.07879688590765,
      "num_input_tokens_seen": 3264636,
      "step": 1731,
      "train_runtime": 4690.8507,
      "train_tokens_per_second": 695.958
    },
    {
      "epoch": 1.7500631791761436,
      "grad_norm": 1.6486413478851318,
      "learning_rate": 2.521518987341772e-06,
      "loss": 0.06214247643947601,
      "num_input_tokens_seen": 3266290,
      "step": 1732,
      "train_runtime": 4693.1265,
      "train_tokens_per_second": 695.973
    },
    {
      "epoch": 1.7510740459944403,
      "grad_norm": 1.5044211149215698,
      "learning_rate": 2.5113924050632917e-06,
      "loss": 0.05706506222486496,
      "num_input_tokens_seen": 3268124,
      "step": 1733,
      "train_runtime": 4695.3399,
      "train_tokens_per_second": 696.036
    },
    {
      "epoch": 1.752084912812737,
      "grad_norm": 1.5830748081207275,
      "learning_rate": 2.5012658227848103e-06,
      "loss": 0.04672452062368393,
      "num_input_tokens_seen": 3269868,
      "step": 1734,
      "train_runtime": 4697.4884,
      "train_tokens_per_second": 696.089
    },
    {
      "epoch": 1.7530957796310336,
      "grad_norm": 1.5254648923873901,
      "learning_rate": 2.4911392405063294e-06,
      "loss": 0.04982186108827591,
      "num_input_tokens_seen": 3271730,
      "step": 1735,
      "train_runtime": 4699.6983,
      "train_tokens_per_second": 696.157
    },
    {
      "epoch": 1.7541066464493302,
      "grad_norm": 1.5697605609893799,
      "learning_rate": 2.481012658227848e-06,
      "loss": 0.058300647884607315,
      "num_input_tokens_seen": 3273410,
      "step": 1736,
      "train_runtime": 4701.8422,
      "train_tokens_per_second": 696.197
    },
    {
      "epoch": 1.7551175132676269,
      "grad_norm": 1.98488187789917,
      "learning_rate": 2.470886075949367e-06,
      "loss": 0.05246957391500473,
      "num_input_tokens_seen": 3274634,
      "step": 1737,
      "train_runtime": 4703.9654,
      "train_tokens_per_second": 696.143
    },
    {
      "epoch": 1.7561283800859235,
      "grad_norm": 1.7348859310150146,
      "learning_rate": 2.4607594936708862e-06,
      "loss": 0.06733670830726624,
      "num_input_tokens_seen": 3276240,
      "step": 1738,
      "train_runtime": 4706.1358,
      "train_tokens_per_second": 696.164
    },
    {
      "epoch": 1.7571392469042204,
      "grad_norm": 2.027111768722534,
      "learning_rate": 2.4506329113924053e-06,
      "loss": 0.08496301621198654,
      "num_input_tokens_seen": 3278222,
      "step": 1739,
      "train_runtime": 4708.3428,
      "train_tokens_per_second": 696.258
    },
    {
      "epoch": 1.758150113722517,
      "grad_norm": 1.8352309465408325,
      "learning_rate": 2.440506329113924e-06,
      "loss": 0.07484448701143265,
      "num_input_tokens_seen": 3279978,
      "step": 1740,
      "train_runtime": 4710.5559,
      "train_tokens_per_second": 696.304
    },
    {
      "epoch": 1.7591609805408137,
      "grad_norm": 1.4469865560531616,
      "learning_rate": 2.430379746835443e-06,
      "loss": 0.056457456201314926,
      "num_input_tokens_seen": 3282112,
      "step": 1741,
      "train_runtime": 4719.5913,
      "train_tokens_per_second": 695.423
    },
    {
      "epoch": 1.7601718473591106,
      "grad_norm": 1.6029822826385498,
      "learning_rate": 2.420253164556962e-06,
      "loss": 0.06924290210008621,
      "num_input_tokens_seen": 3284090,
      "step": 1742,
      "train_runtime": 4721.8856,
      "train_tokens_per_second": 695.504
    },
    {
      "epoch": 1.7611827141774072,
      "grad_norm": 1.4457576274871826,
      "learning_rate": 2.4101265822784812e-06,
      "loss": 0.05837494134902954,
      "num_input_tokens_seen": 3285966,
      "step": 1743,
      "train_runtime": 4724.1435,
      "train_tokens_per_second": 695.569
    },
    {
      "epoch": 1.762193580995704,
      "grad_norm": 1.6344085931777954,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 0.07796967774629593,
      "num_input_tokens_seen": 3287836,
      "step": 1744,
      "train_runtime": 4726.4068,
      "train_tokens_per_second": 695.631
    },
    {
      "epoch": 1.7632044478140005,
      "grad_norm": 1.641486406326294,
      "learning_rate": 2.389873417721519e-06,
      "loss": 0.07068021595478058,
      "num_input_tokens_seen": 3289794,
      "step": 1745,
      "train_runtime": 4728.6524,
      "train_tokens_per_second": 695.715
    },
    {
      "epoch": 1.7642153146322972,
      "grad_norm": 1.5976099967956543,
      "learning_rate": 2.379746835443038e-06,
      "loss": 0.05329452455043793,
      "num_input_tokens_seen": 3291814,
      "step": 1746,
      "train_runtime": 4730.9261,
      "train_tokens_per_second": 695.808
    },
    {
      "epoch": 1.7652261814505938,
      "grad_norm": 1.4337458610534668,
      "learning_rate": 2.369620253164557e-06,
      "loss": 0.06847281754016876,
      "num_input_tokens_seen": 3293728,
      "step": 1747,
      "train_runtime": 4733.1414,
      "train_tokens_per_second": 695.886
    },
    {
      "epoch": 1.7662370482688905,
      "grad_norm": 1.8719747066497803,
      "learning_rate": 2.3594936708860762e-06,
      "loss": 0.07044140994548798,
      "num_input_tokens_seen": 3295612,
      "step": 1748,
      "train_runtime": 4735.3261,
      "train_tokens_per_second": 695.963
    },
    {
      "epoch": 1.7672479150871871,
      "grad_norm": 1.7119461297988892,
      "learning_rate": 2.349367088607595e-06,
      "loss": 0.08578880876302719,
      "num_input_tokens_seen": 3297530,
      "step": 1749,
      "train_runtime": 4737.5433,
      "train_tokens_per_second": 696.042
    },
    {
      "epoch": 1.7682587819054838,
      "grad_norm": 1.3110840320587158,
      "learning_rate": 2.339240506329114e-06,
      "loss": 0.051098622381687164,
      "num_input_tokens_seen": 3299672,
      "step": 1750,
      "train_runtime": 4739.8124,
      "train_tokens_per_second": 696.161
    },
    {
      "epoch": 1.7692696487237807,
      "grad_norm": 2.289757251739502,
      "learning_rate": 2.329113924050633e-06,
      "loss": 0.08187605440616608,
      "num_input_tokens_seen": 3301324,
      "step": 1751,
      "train_runtime": 4742.0031,
      "train_tokens_per_second": 696.188
    },
    {
      "epoch": 1.7702805155420773,
      "grad_norm": 1.4145066738128662,
      "learning_rate": 2.318987341772152e-06,
      "loss": 0.06139783933758736,
      "num_input_tokens_seen": 3303024,
      "step": 1752,
      "train_runtime": 4744.1954,
      "train_tokens_per_second": 696.224
    },
    {
      "epoch": 1.771291382360374,
      "grad_norm": 1.6471421718597412,
      "learning_rate": 2.308860759493671e-06,
      "loss": 0.05509607493877411,
      "num_input_tokens_seen": 3304390,
      "step": 1753,
      "train_runtime": 4746.2525,
      "train_tokens_per_second": 696.21
    },
    {
      "epoch": 1.7723022491786709,
      "grad_norm": 1.8534189462661743,
      "learning_rate": 2.29873417721519e-06,
      "loss": 0.07779213786125183,
      "num_input_tokens_seen": 3306056,
      "step": 1754,
      "train_runtime": 4748.4236,
      "train_tokens_per_second": 696.243
    },
    {
      "epoch": 1.7733131159969675,
      "grad_norm": 1.6584595441818237,
      "learning_rate": 2.288607594936709e-06,
      "loss": 0.05281475931406021,
      "num_input_tokens_seen": 3308636,
      "step": 1755,
      "train_runtime": 4750.8198,
      "train_tokens_per_second": 696.435
    },
    {
      "epoch": 1.7743239828152642,
      "grad_norm": 1.6571482419967651,
      "learning_rate": 2.278481012658228e-06,
      "loss": 0.056524455547332764,
      "num_input_tokens_seen": 3310440,
      "step": 1756,
      "train_runtime": 4753.0233,
      "train_tokens_per_second": 696.491
    },
    {
      "epoch": 1.7753348496335608,
      "grad_norm": 1.9512845277786255,
      "learning_rate": 2.268354430379747e-06,
      "loss": 0.06879110634326935,
      "num_input_tokens_seen": 3312156,
      "step": 1757,
      "train_runtime": 4755.2561,
      "train_tokens_per_second": 696.525
    },
    {
      "epoch": 1.7763457164518575,
      "grad_norm": 1.4111546277999878,
      "learning_rate": 2.258227848101266e-06,
      "loss": 0.04614604637026787,
      "num_input_tokens_seen": 3314202,
      "step": 1758,
      "train_runtime": 4757.5207,
      "train_tokens_per_second": 696.624
    },
    {
      "epoch": 1.7773565832701541,
      "grad_norm": 1.5941832065582275,
      "learning_rate": 2.248101265822785e-06,
      "loss": 0.0545719638466835,
      "num_input_tokens_seen": 3315796,
      "step": 1759,
      "train_runtime": 4759.6459,
      "train_tokens_per_second": 696.648
    },
    {
      "epoch": 1.7783674500884508,
      "grad_norm": 1.647290825843811,
      "learning_rate": 2.237974683544304e-06,
      "loss": 0.06566182523965836,
      "num_input_tokens_seen": 3317610,
      "step": 1760,
      "train_runtime": 4761.8744,
      "train_tokens_per_second": 696.703
    },
    {
      "epoch": 1.7793783169067474,
      "grad_norm": 1.487747311592102,
      "learning_rate": 2.227848101265823e-06,
      "loss": 0.07319916784763336,
      "num_input_tokens_seen": 3319370,
      "step": 1761,
      "train_runtime": 4764.1199,
      "train_tokens_per_second": 696.744
    },
    {
      "epoch": 1.780389183725044,
      "grad_norm": 1.6492506265640259,
      "learning_rate": 2.2177215189873417e-06,
      "loss": 0.06596957892179489,
      "num_input_tokens_seen": 3321328,
      "step": 1762,
      "train_runtime": 4766.368,
      "train_tokens_per_second": 696.826
    },
    {
      "epoch": 1.781400050543341,
      "grad_norm": 1.9240431785583496,
      "learning_rate": 2.207594936708861e-06,
      "loss": 0.0817854031920433,
      "num_input_tokens_seen": 3322902,
      "step": 1763,
      "train_runtime": 4768.4628,
      "train_tokens_per_second": 696.85
    },
    {
      "epoch": 1.7824109173616376,
      "grad_norm": 1.530824065208435,
      "learning_rate": 2.19746835443038e-06,
      "loss": 0.07161606103181839,
      "num_input_tokens_seen": 3324686,
      "step": 1764,
      "train_runtime": 4770.6171,
      "train_tokens_per_second": 696.909
    },
    {
      "epoch": 1.7834217841799342,
      "grad_norm": 1.5622555017471313,
      "learning_rate": 2.187341772151899e-06,
      "loss": 0.07548850774765015,
      "num_input_tokens_seen": 3327138,
      "step": 1765,
      "train_runtime": 4772.9491,
      "train_tokens_per_second": 697.082
    },
    {
      "epoch": 1.7844326509982311,
      "grad_norm": 1.4703495502471924,
      "learning_rate": 2.177215189873418e-06,
      "loss": 0.07107003033161163,
      "num_input_tokens_seen": 3329122,
      "step": 1766,
      "train_runtime": 4775.1946,
      "train_tokens_per_second": 697.17
    },
    {
      "epoch": 1.7854435178165278,
      "grad_norm": 1.4259532690048218,
      "learning_rate": 2.1670886075949367e-06,
      "loss": 0.05702789127826691,
      "num_input_tokens_seen": 3330766,
      "step": 1767,
      "train_runtime": 4777.391,
      "train_tokens_per_second": 697.194
    },
    {
      "epoch": 1.7864543846348244,
      "grad_norm": 1.676520586013794,
      "learning_rate": 2.156962025316456e-06,
      "loss": 0.06703896820545197,
      "num_input_tokens_seen": 3332672,
      "step": 1768,
      "train_runtime": 4779.572,
      "train_tokens_per_second": 697.274
    },
    {
      "epoch": 1.787465251453121,
      "grad_norm": 1.47948157787323,
      "learning_rate": 2.146835443037975e-06,
      "loss": 0.06247759610414505,
      "num_input_tokens_seen": 3334644,
      "step": 1769,
      "train_runtime": 4781.8088,
      "train_tokens_per_second": 697.36
    },
    {
      "epoch": 1.7884761182714177,
      "grad_norm": 1.456027865409851,
      "learning_rate": 2.136708860759494e-06,
      "loss": 0.05956772714853287,
      "num_input_tokens_seen": 3336452,
      "step": 1770,
      "train_runtime": 4783.9683,
      "train_tokens_per_second": 697.424
    },
    {
      "epoch": 1.7894869850897144,
      "grad_norm": 1.5981473922729492,
      "learning_rate": 2.1265822784810126e-06,
      "loss": 0.07002440094947815,
      "num_input_tokens_seen": 3338456,
      "step": 1771,
      "train_runtime": 4792.7152,
      "train_tokens_per_second": 696.569
    },
    {
      "epoch": 1.790497851908011,
      "grad_norm": 2.0481436252593994,
      "learning_rate": 2.1164556962025317e-06,
      "loss": 0.08195114880800247,
      "num_input_tokens_seen": 3340018,
      "step": 1772,
      "train_runtime": 4794.9511,
      "train_tokens_per_second": 696.57
    },
    {
      "epoch": 1.7915087187263077,
      "grad_norm": 1.4928512573242188,
      "learning_rate": 2.106329113924051e-06,
      "loss": 0.05744026601314545,
      "num_input_tokens_seen": 3342170,
      "step": 1773,
      "train_runtime": 4797.3743,
      "train_tokens_per_second": 696.667
    },
    {
      "epoch": 1.7925195855446043,
      "grad_norm": 1.6322563886642456,
      "learning_rate": 2.09620253164557e-06,
      "loss": 0.06046850234270096,
      "num_input_tokens_seen": 3343816,
      "step": 1774,
      "train_runtime": 4799.7204,
      "train_tokens_per_second": 696.669
    },
    {
      "epoch": 1.7935304523629012,
      "grad_norm": 1.6751173734664917,
      "learning_rate": 2.0860759493670885e-06,
      "loss": 0.057147130370140076,
      "num_input_tokens_seen": 3345358,
      "step": 1775,
      "train_runtime": 4802.1459,
      "train_tokens_per_second": 696.638
    },
    {
      "epoch": 1.7945413191811979,
      "grad_norm": 1.7018221616744995,
      "learning_rate": 2.0759493670886076e-06,
      "loss": 0.07538791000843048,
      "num_input_tokens_seen": 3347366,
      "step": 1776,
      "train_runtime": 4804.6966,
      "train_tokens_per_second": 696.686
    },
    {
      "epoch": 1.7955521859994947,
      "grad_norm": 1.5675290822982788,
      "learning_rate": 2.0658227848101267e-06,
      "loss": 0.06920869648456573,
      "num_input_tokens_seen": 3349976,
      "step": 1777,
      "train_runtime": 4807.2503,
      "train_tokens_per_second": 696.859
    },
    {
      "epoch": 1.7965630528177914,
      "grad_norm": 1.8330793380737305,
      "learning_rate": 2.055696202531646e-06,
      "loss": 0.08451933413743973,
      "num_input_tokens_seen": 3351784,
      "step": 1778,
      "train_runtime": 4809.6388,
      "train_tokens_per_second": 696.889
    },
    {
      "epoch": 1.797573919636088,
      "grad_norm": 1.486541509628296,
      "learning_rate": 2.045569620253165e-06,
      "loss": 0.07506123930215836,
      "num_input_tokens_seen": 3353932,
      "step": 1779,
      "train_runtime": 4812.0519,
      "train_tokens_per_second": 696.986
    },
    {
      "epoch": 1.7985847864543847,
      "grad_norm": 2.7654199600219727,
      "learning_rate": 2.0354430379746835e-06,
      "loss": 0.0834086686372757,
      "num_input_tokens_seen": 3355726,
      "step": 1780,
      "train_runtime": 4814.2263,
      "train_tokens_per_second": 697.044
    },
    {
      "epoch": 1.7995956532726813,
      "grad_norm": 1.4466283321380615,
      "learning_rate": 2.0253164556962026e-06,
      "loss": 0.06209476292133331,
      "num_input_tokens_seen": 3358448,
      "step": 1781,
      "train_runtime": 4816.6352,
      "train_tokens_per_second": 697.26
    },
    {
      "epoch": 1.800606520090978,
      "grad_norm": 1.5030364990234375,
      "learning_rate": 2.0151898734177217e-06,
      "loss": 0.07335429638624191,
      "num_input_tokens_seen": 3360568,
      "step": 1782,
      "train_runtime": 4818.8755,
      "train_tokens_per_second": 697.376
    },
    {
      "epoch": 1.800606520090978,
      "eval_loss": 0.3904574513435364,
      "eval_runtime": 52.5525,
      "eval_samples_per_second": 16.726,
      "eval_steps_per_second": 8.373,
      "num_input_tokens_seen": 3360568,
      "step": 1782
    },
    {
      "epoch": 1.8016173869092746,
      "grad_norm": 1.8171679973602295,
      "learning_rate": 2.005063291139241e-06,
      "loss": 0.08095956593751907,
      "num_input_tokens_seen": 3362796,
      "step": 1783,
      "train_runtime": 4873.7509,
      "train_tokens_per_second": 689.981
    },
    {
      "epoch": 1.8026282537275713,
      "grad_norm": 1.6126306056976318,
      "learning_rate": 1.9949367088607595e-06,
      "loss": 0.07605712860822678,
      "num_input_tokens_seen": 3364652,
      "step": 1784,
      "train_runtime": 4876.0926,
      "train_tokens_per_second": 690.03
    },
    {
      "epoch": 1.803639120545868,
      "grad_norm": 2.00726580619812,
      "learning_rate": 1.9848101265822785e-06,
      "loss": 0.0651559978723526,
      "num_input_tokens_seen": 3366096,
      "step": 1785,
      "train_runtime": 4878.2623,
      "train_tokens_per_second": 690.019
    },
    {
      "epoch": 1.8046499873641648,
      "grad_norm": 1.4736859798431396,
      "learning_rate": 1.9746835443037976e-06,
      "loss": 0.0557970367372036,
      "num_input_tokens_seen": 3368244,
      "step": 1786,
      "train_runtime": 4880.6843,
      "train_tokens_per_second": 690.117
    },
    {
      "epoch": 1.8056608541824615,
      "grad_norm": 1.845827579498291,
      "learning_rate": 1.9645569620253167e-06,
      "loss": 0.09264189004898071,
      "num_input_tokens_seen": 3369938,
      "step": 1787,
      "train_runtime": 4882.8735,
      "train_tokens_per_second": 690.155
    },
    {
      "epoch": 1.8066717210007581,
      "grad_norm": 1.808158278465271,
      "learning_rate": 1.954430379746836e-06,
      "loss": 0.06721871346235275,
      "num_input_tokens_seen": 3371544,
      "step": 1788,
      "train_runtime": 4884.9635,
      "train_tokens_per_second": 690.188
    },
    {
      "epoch": 1.807682587819055,
      "grad_norm": 1.875516653060913,
      "learning_rate": 1.9443037974683544e-06,
      "loss": 0.05865068733692169,
      "num_input_tokens_seen": 3372914,
      "step": 1789,
      "train_runtime": 4887.0357,
      "train_tokens_per_second": 690.176
    },
    {
      "epoch": 1.8086934546373516,
      "grad_norm": 1.557678461074829,
      "learning_rate": 1.9341772151898735e-06,
      "loss": 0.06890945136547089,
      "num_input_tokens_seen": 3374998,
      "step": 1790,
      "train_runtime": 4889.2849,
      "train_tokens_per_second": 690.285
    },
    {
      "epoch": 1.8097043214556483,
      "grad_norm": 1.3954472541809082,
      "learning_rate": 1.9240506329113926e-06,
      "loss": 0.062456563115119934,
      "num_input_tokens_seen": 3376898,
      "step": 1791,
      "train_runtime": 4891.5463,
      "train_tokens_per_second": 690.354
    },
    {
      "epoch": 1.810715188273945,
      "grad_norm": 1.6728461980819702,
      "learning_rate": 1.9139240506329117e-06,
      "loss": 0.070929154753685,
      "num_input_tokens_seen": 3379128,
      "step": 1792,
      "train_runtime": 4893.8304,
      "train_tokens_per_second": 690.487
    },
    {
      "epoch": 1.8117260550922416,
      "grad_norm": 1.968773365020752,
      "learning_rate": 1.9037974683544306e-06,
      "loss": 0.06813179701566696,
      "num_input_tokens_seen": 3380466,
      "step": 1793,
      "train_runtime": 4895.9043,
      "train_tokens_per_second": 690.468
    },
    {
      "epoch": 1.8127369219105383,
      "grad_norm": 1.569763422012329,
      "learning_rate": 1.8936708860759494e-06,
      "loss": 0.06483528763055801,
      "num_input_tokens_seen": 3381916,
      "step": 1794,
      "train_runtime": 4898.0221,
      "train_tokens_per_second": 690.466
    },
    {
      "epoch": 1.813747788728835,
      "grad_norm": 1.5195773839950562,
      "learning_rate": 1.8835443037974685e-06,
      "loss": 0.06212960183620453,
      "num_input_tokens_seen": 3384378,
      "step": 1795,
      "train_runtime": 4900.3322,
      "train_tokens_per_second": 690.643
    },
    {
      "epoch": 1.8147586555471316,
      "grad_norm": 1.8203085660934448,
      "learning_rate": 1.8734177215189874e-06,
      "loss": 0.09631272405385971,
      "num_input_tokens_seen": 3386200,
      "step": 1796,
      "train_runtime": 4902.4921,
      "train_tokens_per_second": 690.71
    },
    {
      "epoch": 1.8157695223654282,
      "grad_norm": 1.3569339513778687,
      "learning_rate": 1.8632911392405065e-06,
      "loss": 0.04996223375201225,
      "num_input_tokens_seen": 3388380,
      "step": 1797,
      "train_runtime": 4904.7428,
      "train_tokens_per_second": 690.837
    },
    {
      "epoch": 1.816780389183725,
      "grad_norm": 1.7530453205108643,
      "learning_rate": 1.8531645569620254e-06,
      "loss": 0.049221381545066833,
      "num_input_tokens_seen": 3390244,
      "step": 1798,
      "train_runtime": 4906.9329,
      "train_tokens_per_second": 690.909
    },
    {
      "epoch": 1.8177912560020217,
      "grad_norm": 2.0620510578155518,
      "learning_rate": 1.8430379746835444e-06,
      "loss": 0.08164380490779877,
      "num_input_tokens_seen": 3392044,
      "step": 1799,
      "train_runtime": 4909.1442,
      "train_tokens_per_second": 690.964
    },
    {
      "epoch": 1.8188021228203184,
      "grad_norm": 1.41123628616333,
      "learning_rate": 1.8329113924050635e-06,
      "loss": 0.052159473299980164,
      "num_input_tokens_seen": 3394136,
      "step": 1800,
      "train_runtime": 4911.3918,
      "train_tokens_per_second": 691.074
    },
    {
      "epoch": 1.8198129896386153,
      "grad_norm": 1.3010928630828857,
      "learning_rate": 1.8227848101265824e-06,
      "loss": 0.06299445778131485,
      "num_input_tokens_seen": 3396196,
      "step": 1801,
      "train_runtime": 4920.1333,
      "train_tokens_per_second": 690.265
    },
    {
      "epoch": 1.820823856456912,
      "grad_norm": 1.6793879270553589,
      "learning_rate": 1.8126582278481015e-06,
      "loss": 0.05772136151790619,
      "num_input_tokens_seen": 3398414,
      "step": 1802,
      "train_runtime": 4922.3972,
      "train_tokens_per_second": 690.398
    },
    {
      "epoch": 1.8218347232752086,
      "grad_norm": 1.858668565750122,
      "learning_rate": 1.8025316455696204e-06,
      "loss": 0.07672533392906189,
      "num_input_tokens_seen": 3400244,
      "step": 1803,
      "train_runtime": 4924.5831,
      "train_tokens_per_second": 690.463
    },
    {
      "epoch": 1.8228455900935052,
      "grad_norm": 1.902227520942688,
      "learning_rate": 1.7924050632911394e-06,
      "loss": 0.08474107831716537,
      "num_input_tokens_seen": 3401862,
      "step": 1804,
      "train_runtime": 4926.7568,
      "train_tokens_per_second": 690.487
    },
    {
      "epoch": 1.8238564569118019,
      "grad_norm": 1.6770691871643066,
      "learning_rate": 1.7822784810126583e-06,
      "loss": 0.0702476054430008,
      "num_input_tokens_seen": 3403396,
      "step": 1805,
      "train_runtime": 4928.8713,
      "train_tokens_per_second": 690.502
    },
    {
      "epoch": 1.8248673237300985,
      "grad_norm": 1.501683235168457,
      "learning_rate": 1.7721518987341774e-06,
      "loss": 0.058430809527635574,
      "num_input_tokens_seen": 3405348,
      "step": 1806,
      "train_runtime": 4931.1975,
      "train_tokens_per_second": 690.572
    },
    {
      "epoch": 1.8258781905483952,
      "grad_norm": 1.3843590021133423,
      "learning_rate": 1.7620253164556963e-06,
      "loss": 0.053882379084825516,
      "num_input_tokens_seen": 3407222,
      "step": 1807,
      "train_runtime": 4933.4259,
      "train_tokens_per_second": 690.64
    },
    {
      "epoch": 1.8268890573666918,
      "grad_norm": 1.9313164949417114,
      "learning_rate": 1.7518987341772154e-06,
      "loss": 0.060391172766685486,
      "num_input_tokens_seen": 3408828,
      "step": 1808,
      "train_runtime": 4935.5799,
      "train_tokens_per_second": 690.664
    },
    {
      "epoch": 1.8278999241849885,
      "grad_norm": 1.171385645866394,
      "learning_rate": 1.7417721518987342e-06,
      "loss": 0.039266109466552734,
      "num_input_tokens_seen": 3410766,
      "step": 1809,
      "train_runtime": 4937.7982,
      "train_tokens_per_second": 690.746
    },
    {
      "epoch": 1.8289107910032854,
      "grad_norm": 2.3054921627044678,
      "learning_rate": 1.7316455696202533e-06,
      "loss": 0.04750610888004303,
      "num_input_tokens_seen": 3412244,
      "step": 1810,
      "train_runtime": 4939.9203,
      "train_tokens_per_second": 690.749
    },
    {
      "epoch": 1.829921657821582,
      "grad_norm": 1.761833667755127,
      "learning_rate": 1.7215189873417724e-06,
      "loss": 0.06808142364025116,
      "num_input_tokens_seen": 3413644,
      "step": 1811,
      "train_runtime": 4942.0062,
      "train_tokens_per_second": 690.741
    },
    {
      "epoch": 1.8309325246398787,
      "grad_norm": 1.678519606590271,
      "learning_rate": 1.7113924050632913e-06,
      "loss": 0.078433096408844,
      "num_input_tokens_seen": 3415632,
      "step": 1812,
      "train_runtime": 4944.2397,
      "train_tokens_per_second": 690.831
    },
    {
      "epoch": 1.8319433914581755,
      "grad_norm": 1.8285229206085205,
      "learning_rate": 1.7012658227848104e-06,
      "loss": 0.0811389610171318,
      "num_input_tokens_seen": 3417608,
      "step": 1813,
      "train_runtime": 4946.5695,
      "train_tokens_per_second": 690.905
    },
    {
      "epoch": 1.8329542582764722,
      "grad_norm": 1.7586532831192017,
      "learning_rate": 1.6911392405063292e-06,
      "loss": 0.10236801952123642,
      "num_input_tokens_seen": 3419590,
      "step": 1814,
      "train_runtime": 4948.8475,
      "train_tokens_per_second": 690.987
    },
    {
      "epoch": 1.8339651250947688,
      "grad_norm": 1.28721022605896,
      "learning_rate": 1.6810126582278483e-06,
      "loss": 0.055602312088012695,
      "num_input_tokens_seen": 3421706,
      "step": 1815,
      "train_runtime": 4951.1041,
      "train_tokens_per_second": 691.1
    },
    {
      "epoch": 1.8349759919130655,
      "grad_norm": 1.5468133687973022,
      "learning_rate": 1.6708860759493672e-06,
      "loss": 0.0795050710439682,
      "num_input_tokens_seen": 3423888,
      "step": 1816,
      "train_runtime": 4953.3492,
      "train_tokens_per_second": 691.227
    },
    {
      "epoch": 1.8359868587313621,
      "grad_norm": 2.1119022369384766,
      "learning_rate": 1.6607594936708863e-06,
      "loss": 0.08815045654773712,
      "num_input_tokens_seen": 3425318,
      "step": 1817,
      "train_runtime": 4955.459,
      "train_tokens_per_second": 691.221
    },
    {
      "epoch": 1.8369977255496588,
      "grad_norm": 1.8005280494689941,
      "learning_rate": 1.6506329113924051e-06,
      "loss": 0.06341415643692017,
      "num_input_tokens_seen": 3426994,
      "step": 1818,
      "train_runtime": 4957.6288,
      "train_tokens_per_second": 691.257
    },
    {
      "epoch": 1.8380085923679554,
      "grad_norm": 2.0441479682922363,
      "learning_rate": 1.6405063291139242e-06,
      "loss": 0.07308895885944366,
      "num_input_tokens_seen": 3428430,
      "step": 1819,
      "train_runtime": 4959.7122,
      "train_tokens_per_second": 691.256
    },
    {
      "epoch": 1.839019459186252,
      "grad_norm": 2.1781206130981445,
      "learning_rate": 1.630379746835443e-06,
      "loss": 0.06979721784591675,
      "num_input_tokens_seen": 3429958,
      "step": 1820,
      "train_runtime": 4961.8539,
      "train_tokens_per_second": 691.265
    },
    {
      "epoch": 1.8400303260045487,
      "grad_norm": 1.9482448101043701,
      "learning_rate": 1.6202531645569622e-06,
      "loss": 0.08638764917850494,
      "num_input_tokens_seen": 3431730,
      "step": 1821,
      "train_runtime": 4964.0503,
      "train_tokens_per_second": 691.317
    },
    {
      "epoch": 1.8410411928228456,
      "grad_norm": 2.3552582263946533,
      "learning_rate": 1.6101265822784813e-06,
      "loss": 0.08728457987308502,
      "num_input_tokens_seen": 3433280,
      "step": 1822,
      "train_runtime": 4966.1523,
      "train_tokens_per_second": 691.336
    },
    {
      "epoch": 1.8420520596411423,
      "grad_norm": 1.7918983697891235,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 0.0671626627445221,
      "num_input_tokens_seen": 3435124,
      "step": 1823,
      "train_runtime": 4968.3949,
      "train_tokens_per_second": 691.395
    },
    {
      "epoch": 1.843062926459439,
      "grad_norm": 2.1483774185180664,
      "learning_rate": 1.5898734177215192e-06,
      "loss": 0.09187205135822296,
      "num_input_tokens_seen": 3436512,
      "step": 1824,
      "train_runtime": 4970.4961,
      "train_tokens_per_second": 691.382
    },
    {
      "epoch": 1.8440737932777358,
      "grad_norm": 2.024263858795166,
      "learning_rate": 1.579746835443038e-06,
      "loss": 0.061750929802656174,
      "num_input_tokens_seen": 3437948,
      "step": 1825,
      "train_runtime": 4972.5933,
      "train_tokens_per_second": 691.379
    },
    {
      "epoch": 1.8450846600960324,
      "grad_norm": 1.584208607673645,
      "learning_rate": 1.5696202531645572e-06,
      "loss": 0.07746241241693497,
      "num_input_tokens_seen": 3439876,
      "step": 1826,
      "train_runtime": 4974.8569,
      "train_tokens_per_second": 691.452
    },
    {
      "epoch": 1.846095526914329,
      "grad_norm": 1.7835091352462769,
      "learning_rate": 1.559493670886076e-06,
      "loss": 0.08446493744850159,
      "num_input_tokens_seen": 3441730,
      "step": 1827,
      "train_runtime": 4977.0636,
      "train_tokens_per_second": 691.518
    },
    {
      "epoch": 1.8471063937326258,
      "grad_norm": 1.7341301441192627,
      "learning_rate": 1.5493670886075951e-06,
      "loss": 0.05569569021463394,
      "num_input_tokens_seen": 3443580,
      "step": 1828,
      "train_runtime": 4979.2187,
      "train_tokens_per_second": 691.59
    },
    {
      "epoch": 1.8481172605509224,
      "grad_norm": 1.8013854026794434,
      "learning_rate": 1.539240506329114e-06,
      "loss": 0.05938012897968292,
      "num_input_tokens_seen": 3445306,
      "step": 1829,
      "train_runtime": 4981.4066,
      "train_tokens_per_second": 691.633
    },
    {
      "epoch": 1.849128127369219,
      "grad_norm": 1.6440705060958862,
      "learning_rate": 1.529113924050633e-06,
      "loss": 0.08171625435352325,
      "num_input_tokens_seen": 3447488,
      "step": 1830,
      "train_runtime": 4983.7382,
      "train_tokens_per_second": 691.747
    },
    {
      "epoch": 1.8501389941875157,
      "grad_norm": 2.135153293609619,
      "learning_rate": 1.518987341772152e-06,
      "loss": 0.09053754806518555,
      "num_input_tokens_seen": 3449220,
      "step": 1831,
      "train_runtime": 4992.5622,
      "train_tokens_per_second": 690.872
    },
    {
      "epoch": 1.8511498610058124,
      "grad_norm": 1.5982813835144043,
      "learning_rate": 1.508860759493671e-06,
      "loss": 0.07078617811203003,
      "num_input_tokens_seen": 3451032,
      "step": 1832,
      "train_runtime": 4994.7765,
      "train_tokens_per_second": 690.928
    },
    {
      "epoch": 1.852160727824109,
      "grad_norm": 1.5052188634872437,
      "learning_rate": 1.4987341772151901e-06,
      "loss": 0.050244953483343124,
      "num_input_tokens_seen": 3452824,
      "step": 1833,
      "train_runtime": 4997.0024,
      "train_tokens_per_second": 690.979
    },
    {
      "epoch": 1.8531715946424059,
      "grad_norm": 1.3054684400558472,
      "learning_rate": 1.488607594936709e-06,
      "loss": 0.045286163687705994,
      "num_input_tokens_seen": 3454654,
      "step": 1834,
      "train_runtime": 4999.2124,
      "train_tokens_per_second": 691.04
    },
    {
      "epoch": 1.8541824614607025,
      "grad_norm": 1.6632694005966187,
      "learning_rate": 1.478481012658228e-06,
      "loss": 0.04870486631989479,
      "num_input_tokens_seen": 3456194,
      "step": 1835,
      "train_runtime": 5001.421,
      "train_tokens_per_second": 691.042
    },
    {
      "epoch": 1.8551933282789992,
      "grad_norm": 1.638814926147461,
      "learning_rate": 1.468354430379747e-06,
      "loss": 0.04414809122681618,
      "num_input_tokens_seen": 3458628,
      "step": 1836,
      "train_runtime": 5003.8432,
      "train_tokens_per_second": 691.194
    },
    {
      "epoch": 1.856204195097296,
      "grad_norm": 1.8952370882034302,
      "learning_rate": 1.458227848101266e-06,
      "loss": 0.08647635579109192,
      "num_input_tokens_seen": 3460506,
      "step": 1837,
      "train_runtime": 5006.0602,
      "train_tokens_per_second": 691.263
    },
    {
      "epoch": 1.8572150619155927,
      "grad_norm": 1.6776092052459717,
      "learning_rate": 1.448101265822785e-06,
      "loss": 0.06300491839647293,
      "num_input_tokens_seen": 3462502,
      "step": 1838,
      "train_runtime": 5008.3496,
      "train_tokens_per_second": 691.346
    },
    {
      "epoch": 1.8582259287338894,
      "grad_norm": 1.5861340761184692,
      "learning_rate": 1.437974683544304e-06,
      "loss": 0.05638999491930008,
      "num_input_tokens_seen": 3464010,
      "step": 1839,
      "train_runtime": 5010.5477,
      "train_tokens_per_second": 691.344
    },
    {
      "epoch": 1.859236795552186,
      "grad_norm": 1.8318597078323364,
      "learning_rate": 1.4278481012658229e-06,
      "loss": 0.05801784247159958,
      "num_input_tokens_seen": 3465982,
      "step": 1840,
      "train_runtime": 5012.9244,
      "train_tokens_per_second": 691.409
    },
    {
      "epoch": 1.8602476623704827,
      "grad_norm": 1.6301965713500977,
      "learning_rate": 1.417721518987342e-06,
      "loss": 0.08084286004304886,
      "num_input_tokens_seen": 3468036,
      "step": 1841,
      "train_runtime": 5015.2036,
      "train_tokens_per_second": 691.505
    },
    {
      "epoch": 1.8612585291887793,
      "grad_norm": 2.134197950363159,
      "learning_rate": 1.4075949367088608e-06,
      "loss": 0.07081994414329529,
      "num_input_tokens_seen": 3469660,
      "step": 1842,
      "train_runtime": 5017.3585,
      "train_tokens_per_second": 691.531
    },
    {
      "epoch": 1.862269396007076,
      "grad_norm": 2.1300292015075684,
      "learning_rate": 1.39746835443038e-06,
      "loss": 0.09027338027954102,
      "num_input_tokens_seen": 3471300,
      "step": 1843,
      "train_runtime": 5019.5769,
      "train_tokens_per_second": 691.552
    },
    {
      "epoch": 1.8632802628253726,
      "grad_norm": 2.171847343444824,
      "learning_rate": 1.387341772151899e-06,
      "loss": 0.07858128100633621,
      "num_input_tokens_seen": 3472838,
      "step": 1844,
      "train_runtime": 5021.725,
      "train_tokens_per_second": 691.563
    },
    {
      "epoch": 1.8642911296436695,
      "grad_norm": 1.9706461429595947,
      "learning_rate": 1.3772151898734179e-06,
      "loss": 0.08030667901039124,
      "num_input_tokens_seen": 3474660,
      "step": 1845,
      "train_runtime": 5023.9119,
      "train_tokens_per_second": 691.624
    },
    {
      "epoch": 1.8653019964619661,
      "grad_norm": 1.915985345840454,
      "learning_rate": 1.367088607594937e-06,
      "loss": 0.07971419394016266,
      "num_input_tokens_seen": 3476294,
      "step": 1846,
      "train_runtime": 5026.0838,
      "train_tokens_per_second": 691.651
    },
    {
      "epoch": 1.8663128632802628,
      "grad_norm": 1.6336525678634644,
      "learning_rate": 1.3569620253164558e-06,
      "loss": 0.05157170072197914,
      "num_input_tokens_seen": 3477836,
      "step": 1847,
      "train_runtime": 5028.2093,
      "train_tokens_per_second": 691.665
    },
    {
      "epoch": 1.8673237300985597,
      "grad_norm": 1.8403663635253906,
      "learning_rate": 1.346835443037975e-06,
      "loss": 0.07300272583961487,
      "num_input_tokens_seen": 3479554,
      "step": 1848,
      "train_runtime": 5030.4055,
      "train_tokens_per_second": 691.704
    },
    {
      "epoch": 1.8683345969168563,
      "grad_norm": 1.7771477699279785,
      "learning_rate": 1.3367088607594938e-06,
      "loss": 0.07321541011333466,
      "num_input_tokens_seen": 3481364,
      "step": 1849,
      "train_runtime": 5032.6183,
      "train_tokens_per_second": 691.76
    },
    {
      "epoch": 1.869345463735153,
      "grad_norm": 1.665406346321106,
      "learning_rate": 1.3265822784810129e-06,
      "loss": 0.09364591538906097,
      "num_input_tokens_seen": 3483210,
      "step": 1850,
      "train_runtime": 5034.8219,
      "train_tokens_per_second": 691.824
    },
    {
      "epoch": 1.8703563305534496,
      "grad_norm": 1.1183644533157349,
      "learning_rate": 1.3164556962025317e-06,
      "loss": 0.04247704893350601,
      "num_input_tokens_seen": 3484998,
      "step": 1851,
      "train_runtime": 5037.0281,
      "train_tokens_per_second": 691.876
    },
    {
      "epoch": 1.8713671973717463,
      "grad_norm": 1.6386619806289673,
      "learning_rate": 1.3063291139240508e-06,
      "loss": 0.04744293540716171,
      "num_input_tokens_seen": 3486614,
      "step": 1852,
      "train_runtime": 5039.1912,
      "train_tokens_per_second": 691.9
    },
    {
      "epoch": 1.872378064190043,
      "grad_norm": 1.7003967761993408,
      "learning_rate": 1.2962025316455697e-06,
      "loss": 0.06002739444375038,
      "num_input_tokens_seen": 3488406,
      "step": 1853,
      "train_runtime": 5041.3562,
      "train_tokens_per_second": 691.958
    },
    {
      "epoch": 1.8733889310083396,
      "grad_norm": 2.393953800201416,
      "learning_rate": 1.2860759493670888e-06,
      "loss": 0.09052784740924835,
      "num_input_tokens_seen": 3489924,
      "step": 1854,
      "train_runtime": 5043.5327,
      "train_tokens_per_second": 691.96
    },
    {
      "epoch": 1.8743997978266362,
      "grad_norm": 1.5386276245117188,
      "learning_rate": 1.2759493670886079e-06,
      "loss": 0.05487651377916336,
      "num_input_tokens_seen": 3492204,
      "step": 1855,
      "train_runtime": 5045.8265,
      "train_tokens_per_second": 692.098
    },
    {
      "epoch": 1.875410664644933,
      "grad_norm": 1.4505112171173096,
      "learning_rate": 1.2658227848101267e-06,
      "loss": 0.04382224380970001,
      "num_input_tokens_seen": 3493908,
      "step": 1856,
      "train_runtime": 5048.0202,
      "train_tokens_per_second": 692.134
    },
    {
      "epoch": 1.8764215314632298,
      "grad_norm": 1.3902812004089355,
      "learning_rate": 1.2556962025316458e-06,
      "loss": 0.06371606141328812,
      "num_input_tokens_seen": 3495948,
      "step": 1857,
      "train_runtime": 5050.2433,
      "train_tokens_per_second": 692.234
    },
    {
      "epoch": 1.8774323982815264,
      "grad_norm": 1.582065224647522,
      "learning_rate": 1.2455696202531647e-06,
      "loss": 0.06060520187020302,
      "num_input_tokens_seen": 3498066,
      "step": 1858,
      "train_runtime": 5052.5335,
      "train_tokens_per_second": 692.339
    },
    {
      "epoch": 1.878443265099823,
      "grad_norm": 1.4478693008422852,
      "learning_rate": 1.2354430379746836e-06,
      "loss": 0.06082356348633766,
      "num_input_tokens_seen": 3500240,
      "step": 1859,
      "train_runtime": 5054.808,
      "train_tokens_per_second": 692.458
    },
    {
      "epoch": 1.87945413191812,
      "grad_norm": 1.8662803173065186,
      "learning_rate": 1.2253164556962027e-06,
      "loss": 0.08540136367082596,
      "num_input_tokens_seen": 3502098,
      "step": 1860,
      "train_runtime": 5057.0231,
      "train_tokens_per_second": 692.522
    },
    {
      "epoch": 1.8804649987364166,
      "grad_norm": 1.6401348114013672,
      "learning_rate": 1.2151898734177215e-06,
      "loss": 0.05659695714712143,
      "num_input_tokens_seen": 3504010,
      "step": 1861,
      "train_runtime": 5065.6998,
      "train_tokens_per_second": 691.713
    },
    {
      "epoch": 1.8814758655547132,
      "grad_norm": 1.8000669479370117,
      "learning_rate": 1.2050632911392406e-06,
      "loss": 0.0997447595000267,
      "num_input_tokens_seen": 3506308,
      "step": 1862,
      "train_runtime": 5067.9939,
      "train_tokens_per_second": 691.853
    },
    {
      "epoch": 1.88248673237301,
      "grad_norm": 1.4725761413574219,
      "learning_rate": 1.1949367088607595e-06,
      "loss": 0.05516906827688217,
      "num_input_tokens_seen": 3508312,
      "step": 1863,
      "train_runtime": 5070.2737,
      "train_tokens_per_second": 691.937
    },
    {
      "epoch": 1.8834975991913065,
      "grad_norm": 1.6513181924819946,
      "learning_rate": 1.1848101265822786e-06,
      "loss": 0.06855172663927078,
      "num_input_tokens_seen": 3510196,
      "step": 1864,
      "train_runtime": 5072.4809,
      "train_tokens_per_second": 692.008
    },
    {
      "epoch": 1.8845084660096032,
      "grad_norm": 1.6698564291000366,
      "learning_rate": 1.1746835443037974e-06,
      "loss": 0.05972818285226822,
      "num_input_tokens_seen": 3511816,
      "step": 1865,
      "train_runtime": 5074.6547,
      "train_tokens_per_second": 692.031
    },
    {
      "epoch": 1.8855193328278999,
      "grad_norm": 1.4553149938583374,
      "learning_rate": 1.1645569620253165e-06,
      "loss": 0.05890914425253868,
      "num_input_tokens_seen": 3513736,
      "step": 1866,
      "train_runtime": 5076.8634,
      "train_tokens_per_second": 692.108
    },
    {
      "epoch": 1.8865301996461965,
      "grad_norm": 1.9039795398712158,
      "learning_rate": 1.1544303797468354e-06,
      "loss": 0.06471452116966248,
      "num_input_tokens_seen": 3515274,
      "step": 1867,
      "train_runtime": 5078.9868,
      "train_tokens_per_second": 692.121
    },
    {
      "epoch": 1.8875410664644932,
      "grad_norm": 1.6786473989486694,
      "learning_rate": 1.1443037974683545e-06,
      "loss": 0.07185996323823929,
      "num_input_tokens_seen": 3517364,
      "step": 1868,
      "train_runtime": 5081.271,
      "train_tokens_per_second": 692.221
    },
    {
      "epoch": 1.88855193328279,
      "grad_norm": 1.8660085201263428,
      "learning_rate": 1.1341772151898736e-06,
      "loss": 0.07593970000743866,
      "num_input_tokens_seen": 3519344,
      "step": 1869,
      "train_runtime": 5083.6076,
      "train_tokens_per_second": 692.293
    },
    {
      "epoch": 1.8895628001010867,
      "grad_norm": 1.8927536010742188,
      "learning_rate": 1.1240506329113924e-06,
      "loss": 0.0532231330871582,
      "num_input_tokens_seen": 3521038,
      "step": 1870,
      "train_runtime": 5085.7955,
      "train_tokens_per_second": 692.328
    },
    {
      "epoch": 1.8905736669193833,
      "grad_norm": 1.4850199222564697,
      "learning_rate": 1.1139240506329115e-06,
      "loss": 0.06374198943376541,
      "num_input_tokens_seen": 3522880,
      "step": 1871,
      "train_runtime": 5088.0351,
      "train_tokens_per_second": 692.385
    },
    {
      "epoch": 1.8915845337376802,
      "grad_norm": 1.70200777053833,
      "learning_rate": 1.1037974683544304e-06,
      "loss": 0.07828419655561447,
      "num_input_tokens_seen": 3525122,
      "step": 1872,
      "train_runtime": 5090.367,
      "train_tokens_per_second": 692.508
    },
    {
      "epoch": 1.8925954005559769,
      "grad_norm": 2.2650492191314697,
      "learning_rate": 1.0936708860759495e-06,
      "loss": 0.07213446497917175,
      "num_input_tokens_seen": 3526860,
      "step": 1873,
      "train_runtime": 5092.5173,
      "train_tokens_per_second": 692.557
    },
    {
      "epoch": 1.8936062673742735,
      "grad_norm": 1.9589587450027466,
      "learning_rate": 1.0835443037974684e-06,
      "loss": 0.07544541358947754,
      "num_input_tokens_seen": 3528716,
      "step": 1874,
      "train_runtime": 5094.7493,
      "train_tokens_per_second": 692.618
    },
    {
      "epoch": 1.8946171341925702,
      "grad_norm": 2.1888937950134277,
      "learning_rate": 1.0734177215189874e-06,
      "loss": 0.08355741947889328,
      "num_input_tokens_seen": 3530244,
      "step": 1875,
      "train_runtime": 5096.9047,
      "train_tokens_per_second": 692.625
    },
    {
      "epoch": 1.8956280010108668,
      "grad_norm": 2.0891592502593994,
      "learning_rate": 1.0632911392405063e-06,
      "loss": 0.07398049533367157,
      "num_input_tokens_seen": 3531840,
      "step": 1876,
      "train_runtime": 5099.1005,
      "train_tokens_per_second": 692.64
    },
    {
      "epoch": 1.8966388678291635,
      "grad_norm": 1.6804112195968628,
      "learning_rate": 1.0531645569620254e-06,
      "loss": 0.0694270059466362,
      "num_input_tokens_seen": 3533806,
      "step": 1877,
      "train_runtime": 5101.3349,
      "train_tokens_per_second": 692.722
    },
    {
      "epoch": 1.8976497346474601,
      "grad_norm": 2.4839885234832764,
      "learning_rate": 1.0430379746835443e-06,
      "loss": 0.08326879888772964,
      "num_input_tokens_seen": 3535284,
      "step": 1878,
      "train_runtime": 5103.4445,
      "train_tokens_per_second": 692.725
    },
    {
      "epoch": 1.8986606014657568,
      "grad_norm": 1.8496500253677368,
      "learning_rate": 1.0329113924050634e-06,
      "loss": 0.10120419412851334,
      "num_input_tokens_seen": 3537260,
      "step": 1879,
      "train_runtime": 5105.6793,
      "train_tokens_per_second": 692.809
    },
    {
      "epoch": 1.8996714682840534,
      "grad_norm": 1.635189414024353,
      "learning_rate": 1.0227848101265824e-06,
      "loss": 0.05759390443563461,
      "num_input_tokens_seen": 3539170,
      "step": 1880,
      "train_runtime": 5107.956,
      "train_tokens_per_second": 692.874
    },
    {
      "epoch": 1.9006823351023503,
      "grad_norm": 1.5812994241714478,
      "learning_rate": 1.0126582278481013e-06,
      "loss": 0.06326553225517273,
      "num_input_tokens_seen": 3540982,
      "step": 1881,
      "train_runtime": 5110.2669,
      "train_tokens_per_second": 692.915
    },
    {
      "epoch": 1.901693201920647,
      "grad_norm": 1.7322474718093872,
      "learning_rate": 1.0025316455696204e-06,
      "loss": 0.08316643536090851,
      "num_input_tokens_seen": 3543062,
      "step": 1882,
      "train_runtime": 5112.5532,
      "train_tokens_per_second": 693.012
    },
    {
      "epoch": 1.9027040687389436,
      "grad_norm": 1.5347411632537842,
      "learning_rate": 9.924050632911393e-07,
      "loss": 0.0600912943482399,
      "num_input_tokens_seen": 3545132,
      "step": 1883,
      "train_runtime": 5114.7691,
      "train_tokens_per_second": 693.117
    },
    {
      "epoch": 1.9037149355572405,
      "grad_norm": 1.78092622756958,
      "learning_rate": 9.822784810126584e-07,
      "loss": 0.06544648110866547,
      "num_input_tokens_seen": 3546714,
      "step": 1884,
      "train_runtime": 5116.9727,
      "train_tokens_per_second": 693.127
    },
    {
      "epoch": 1.9047258023755371,
      "grad_norm": 1.6052136421203613,
      "learning_rate": 9.721518987341772e-07,
      "loss": 0.06100933626294136,
      "num_input_tokens_seen": 3548604,
      "step": 1885,
      "train_runtime": 5119.1822,
      "train_tokens_per_second": 693.197
    },
    {
      "epoch": 1.9057366691938338,
      "grad_norm": 2.2423882484436035,
      "learning_rate": 9.620253164556963e-07,
      "loss": 0.09639609605073929,
      "num_input_tokens_seen": 3550542,
      "step": 1886,
      "train_runtime": 5121.3875,
      "train_tokens_per_second": 693.277
    },
    {
      "epoch": 1.9067475360121304,
      "grad_norm": 1.7005420923233032,
      "learning_rate": 9.518987341772153e-07,
      "loss": 0.06617413461208344,
      "num_input_tokens_seen": 3552096,
      "step": 1887,
      "train_runtime": 5123.5454,
      "train_tokens_per_second": 693.289
    },
    {
      "epoch": 1.907758402830427,
      "grad_norm": 1.886470079421997,
      "learning_rate": 9.417721518987343e-07,
      "loss": 0.0866304561495781,
      "num_input_tokens_seen": 3553786,
      "step": 1888,
      "train_runtime": 5125.7007,
      "train_tokens_per_second": 693.327
    },
    {
      "epoch": 1.9087692696487237,
      "grad_norm": 1.8155748844146729,
      "learning_rate": 9.316455696202532e-07,
      "loss": 0.08397430181503296,
      "num_input_tokens_seen": 3555838,
      "step": 1889,
      "train_runtime": 5127.9423,
      "train_tokens_per_second": 693.424
    },
    {
      "epoch": 1.9097801364670204,
      "grad_norm": 1.4033902883529663,
      "learning_rate": 9.215189873417722e-07,
      "loss": 0.06193836033344269,
      "num_input_tokens_seen": 3557936,
      "step": 1890,
      "train_runtime": 5130.2064,
      "train_tokens_per_second": 693.527
    },
    {
      "epoch": 1.910791003285317,
      "grad_norm": 1.5796843767166138,
      "learning_rate": 9.113924050632912e-07,
      "loss": 0.06924757361412048,
      "num_input_tokens_seen": 3559842,
      "step": 1891,
      "train_runtime": 5138.9183,
      "train_tokens_per_second": 692.722
    },
    {
      "epoch": 1.9118018701036137,
      "grad_norm": 1.8680285215377808,
      "learning_rate": 9.012658227848102e-07,
      "loss": 0.07084711641073227,
      "num_input_tokens_seen": 3561686,
      "step": 1892,
      "train_runtime": 5141.1268,
      "train_tokens_per_second": 692.783
    },
    {
      "epoch": 1.9128127369219106,
      "grad_norm": 1.7338470220565796,
      "learning_rate": 8.911392405063292e-07,
      "loss": 0.07172862440347672,
      "num_input_tokens_seen": 3563160,
      "step": 1893,
      "train_runtime": 5143.2457,
      "train_tokens_per_second": 692.784
    },
    {
      "epoch": 1.9138236037402072,
      "grad_norm": 1.7504284381866455,
      "learning_rate": 8.810126582278481e-07,
      "loss": 0.077755868434906,
      "num_input_tokens_seen": 3564766,
      "step": 1894,
      "train_runtime": 5145.3647,
      "train_tokens_per_second": 692.811
    },
    {
      "epoch": 1.9148344705585039,
      "grad_norm": 1.5047736167907715,
      "learning_rate": 8.708860759493671e-07,
      "loss": 0.05209915712475777,
      "num_input_tokens_seen": 3566548,
      "step": 1895,
      "train_runtime": 5147.5382,
      "train_tokens_per_second": 692.865
    },
    {
      "epoch": 1.9158453373768007,
      "grad_norm": 1.5364779233932495,
      "learning_rate": 8.607594936708862e-07,
      "loss": 0.06482240557670593,
      "num_input_tokens_seen": 3568356,
      "step": 1896,
      "train_runtime": 5149.8071,
      "train_tokens_per_second": 692.911
    },
    {
      "epoch": 1.9168562041950974,
      "grad_norm": 2.100857734680176,
      "learning_rate": 8.506329113924052e-07,
      "loss": 0.08881665766239166,
      "num_input_tokens_seen": 3570256,
      "step": 1897,
      "train_runtime": 5152.0468,
      "train_tokens_per_second": 692.978
    },
    {
      "epoch": 1.917867071013394,
      "grad_norm": 1.7566105127334595,
      "learning_rate": 8.405063291139242e-07,
      "loss": 0.08907251060009003,
      "num_input_tokens_seen": 3572150,
      "step": 1898,
      "train_runtime": 5154.3416,
      "train_tokens_per_second": 693.037
    },
    {
      "epoch": 1.9188779378316907,
      "grad_norm": 1.6315668821334839,
      "learning_rate": 8.303797468354431e-07,
      "loss": 0.08250783383846283,
      "num_input_tokens_seen": 3574394,
      "step": 1899,
      "train_runtime": 5156.7541,
      "train_tokens_per_second": 693.148
    },
    {
      "epoch": 1.9198888046499873,
      "grad_norm": 1.550088882446289,
      "learning_rate": 8.202531645569621e-07,
      "loss": 0.060801491141319275,
      "num_input_tokens_seen": 3576864,
      "step": 1900,
      "train_runtime": 5159.1141,
      "train_tokens_per_second": 693.31
    },
    {
      "epoch": 1.920899671468284,
      "grad_norm": 1.800826907157898,
      "learning_rate": 8.101265822784811e-07,
      "loss": 0.07329748570919037,
      "num_input_tokens_seen": 3578852,
      "step": 1901,
      "train_runtime": 5161.3988,
      "train_tokens_per_second": 693.388
    },
    {
      "epoch": 1.9219105382865806,
      "grad_norm": 2.830747365951538,
      "learning_rate": 8.000000000000001e-07,
      "loss": 0.0832766741514206,
      "num_input_tokens_seen": 3580110,
      "step": 1902,
      "train_runtime": 5163.4533,
      "train_tokens_per_second": 693.356
    },
    {
      "epoch": 1.9229214051048773,
      "grad_norm": 1.718701958656311,
      "learning_rate": 7.89873417721519e-07,
      "loss": 0.07386311143636703,
      "num_input_tokens_seen": 3582264,
      "step": 1903,
      "train_runtime": 5165.7572,
      "train_tokens_per_second": 693.463
    },
    {
      "epoch": 1.9239322719231742,
      "grad_norm": 1.9136043787002563,
      "learning_rate": 7.79746835443038e-07,
      "loss": 0.08677150309085846,
      "num_input_tokens_seen": 3583906,
      "step": 1904,
      "train_runtime": 5167.9118,
      "train_tokens_per_second": 693.492
    },
    {
      "epoch": 1.9249431387414708,
      "grad_norm": 1.894005537033081,
      "learning_rate": 7.69620253164557e-07,
      "loss": 0.09352187067270279,
      "num_input_tokens_seen": 3585726,
      "step": 1905,
      "train_runtime": 5170.1254,
      "train_tokens_per_second": 693.547
    },
    {
      "epoch": 1.9259540055597675,
      "grad_norm": 2.1736397743225098,
      "learning_rate": 7.59493670886076e-07,
      "loss": 0.06525714695453644,
      "num_input_tokens_seen": 3587382,
      "step": 1906,
      "train_runtime": 5172.3277,
      "train_tokens_per_second": 693.572
    },
    {
      "epoch": 1.9269648723780644,
      "grad_norm": 1.7275938987731934,
      "learning_rate": 7.493670886075951e-07,
      "loss": 0.06097712367773056,
      "num_input_tokens_seen": 3589402,
      "step": 1907,
      "train_runtime": 5174.6275,
      "train_tokens_per_second": 693.654
    },
    {
      "epoch": 1.927975739196361,
      "grad_norm": 1.5570915937423706,
      "learning_rate": 7.39240506329114e-07,
      "loss": 0.05133200064301491,
      "num_input_tokens_seen": 3590808,
      "step": 1908,
      "train_runtime": 5176.7731,
      "train_tokens_per_second": 693.638
    },
    {
      "epoch": 1.9289866060146577,
      "grad_norm": 1.765472173690796,
      "learning_rate": 7.29113924050633e-07,
      "loss": 0.07087323814630508,
      "num_input_tokens_seen": 3593456,
      "step": 1909,
      "train_runtime": 5179.1694,
      "train_tokens_per_second": 693.829
    },
    {
      "epoch": 1.9299974728329543,
      "grad_norm": 1.6926285028457642,
      "learning_rate": 7.18987341772152e-07,
      "loss": 0.07390089333057404,
      "num_input_tokens_seen": 3595598,
      "step": 1910,
      "train_runtime": 5181.4103,
      "train_tokens_per_second": 693.942
    },
    {
      "epoch": 1.931008339651251,
      "grad_norm": 1.7235076427459717,
      "learning_rate": 7.08860759493671e-07,
      "loss": 0.06663314253091812,
      "num_input_tokens_seen": 3597450,
      "step": 1911,
      "train_runtime": 5183.6001,
      "train_tokens_per_second": 694.006
    },
    {
      "epoch": 1.9320192064695476,
      "grad_norm": 1.5091921091079712,
      "learning_rate": 6.9873417721519e-07,
      "loss": 0.06591759622097015,
      "num_input_tokens_seen": 3599768,
      "step": 1912,
      "train_runtime": 5185.8958,
      "train_tokens_per_second": 694.146
    },
    {
      "epoch": 1.9330300732878443,
      "grad_norm": 1.685353398323059,
      "learning_rate": 6.886075949367089e-07,
      "loss": 0.06985824555158615,
      "num_input_tokens_seen": 3601768,
      "step": 1913,
      "train_runtime": 5188.1358,
      "train_tokens_per_second": 694.232
    },
    {
      "epoch": 1.934040940106141,
      "grad_norm": 1.5597984790802002,
      "learning_rate": 6.784810126582279e-07,
      "loss": 0.0746968686580658,
      "num_input_tokens_seen": 3604040,
      "step": 1914,
      "train_runtime": 5190.4279,
      "train_tokens_per_second": 694.363
    },
    {
      "epoch": 1.9350518069244376,
      "grad_norm": 1.6250715255737305,
      "learning_rate": 6.683544303797469e-07,
      "loss": 0.058975815773010254,
      "num_input_tokens_seen": 3605882,
      "step": 1915,
      "train_runtime": 5192.6106,
      "train_tokens_per_second": 694.426
    },
    {
      "epoch": 1.9360626737427344,
      "grad_norm": 1.5663336515426636,
      "learning_rate": 6.582278481012659e-07,
      "loss": 0.06764580309391022,
      "num_input_tokens_seen": 3607992,
      "step": 1916,
      "train_runtime": 5194.8791,
      "train_tokens_per_second": 694.529
    },
    {
      "epoch": 1.937073540561031,
      "grad_norm": 2.1358561515808105,
      "learning_rate": 6.481012658227849e-07,
      "loss": 0.06255315244197845,
      "num_input_tokens_seen": 3609462,
      "step": 1917,
      "train_runtime": 5196.9891,
      "train_tokens_per_second": 694.529
    },
    {
      "epoch": 1.9380844073793277,
      "grad_norm": 1.297664999961853,
      "learning_rate": 6.379746835443039e-07,
      "loss": 0.05662040412425995,
      "num_input_tokens_seen": 3611516,
      "step": 1918,
      "train_runtime": 5199.2181,
      "train_tokens_per_second": 694.627
    },
    {
      "epoch": 1.9390952741976246,
      "grad_norm": 1.4330904483795166,
      "learning_rate": 6.278481012658229e-07,
      "loss": 0.05363224446773529,
      "num_input_tokens_seen": 3613642,
      "step": 1919,
      "train_runtime": 5201.4994,
      "train_tokens_per_second": 694.731
    },
    {
      "epoch": 1.9401061410159213,
      "grad_norm": 1.6165391206741333,
      "learning_rate": 6.177215189873418e-07,
      "loss": 0.08508194237947464,
      "num_input_tokens_seen": 3615978,
      "step": 1920,
      "train_runtime": 5203.8046,
      "train_tokens_per_second": 694.872
    },
    {
      "epoch": 1.941117007834218,
      "grad_norm": 2.0775704383850098,
      "learning_rate": 6.075949367088608e-07,
      "loss": 0.07671953737735748,
      "num_input_tokens_seen": 3617712,
      "step": 1921,
      "train_runtime": 5212.5709,
      "train_tokens_per_second": 694.036
    },
    {
      "epoch": 1.9421278746525146,
      "grad_norm": 1.7379590272903442,
      "learning_rate": 5.974683544303797e-07,
      "loss": 0.07785919308662415,
      "num_input_tokens_seen": 3619592,
      "step": 1922,
      "train_runtime": 5214.7871,
      "train_tokens_per_second": 694.102
    },
    {
      "epoch": 1.9431387414708112,
      "grad_norm": 1.744384765625,
      "learning_rate": 5.873417721518987e-07,
      "loss": 0.0654076486825943,
      "num_input_tokens_seen": 3621670,
      "step": 1923,
      "train_runtime": 5217.0521,
      "train_tokens_per_second": 694.199
    },
    {
      "epoch": 1.9441496082891079,
      "grad_norm": 2.063349962234497,
      "learning_rate": 5.772151898734177e-07,
      "loss": 0.08102155476808548,
      "num_input_tokens_seen": 3623864,
      "step": 1924,
      "train_runtime": 5219.3159,
      "train_tokens_per_second": 694.318
    },
    {
      "epoch": 1.9451604751074045,
      "grad_norm": 1.8299205303192139,
      "learning_rate": 5.670886075949368e-07,
      "loss": 0.08363678306341171,
      "num_input_tokens_seen": 3625582,
      "step": 1925,
      "train_runtime": 5221.5195,
      "train_tokens_per_second": 694.354
    },
    {
      "epoch": 1.9461713419257012,
      "grad_norm": 1.599539041519165,
      "learning_rate": 5.569620253164558e-07,
      "loss": 0.05633139610290527,
      "num_input_tokens_seen": 3627172,
      "step": 1926,
      "train_runtime": 5223.689,
      "train_tokens_per_second": 694.37
    },
    {
      "epoch": 1.9471822087439978,
      "grad_norm": 1.3668630123138428,
      "learning_rate": 5.468354430379747e-07,
      "loss": 0.04236835986375809,
      "num_input_tokens_seen": 3628574,
      "step": 1927,
      "train_runtime": 5225.7862,
      "train_tokens_per_second": 694.359
    },
    {
      "epoch": 1.9481930755622947,
      "grad_norm": 1.6555768251419067,
      "learning_rate": 5.367088607594937e-07,
      "loss": 0.057802464812994,
      "num_input_tokens_seen": 3630570,
      "step": 1928,
      "train_runtime": 5228.0254,
      "train_tokens_per_second": 694.444
    },
    {
      "epoch": 1.9492039423805914,
      "grad_norm": 1.4538333415985107,
      "learning_rate": 5.265822784810127e-07,
      "loss": 0.044277213513851166,
      "num_input_tokens_seen": 3632622,
      "step": 1929,
      "train_runtime": 5230.2967,
      "train_tokens_per_second": 694.535
    },
    {
      "epoch": 1.950214809198888,
      "grad_norm": 1.9167815446853638,
      "learning_rate": 5.164556962025317e-07,
      "loss": 0.05408067628741264,
      "num_input_tokens_seen": 3634376,
      "step": 1930,
      "train_runtime": 5232.4821,
      "train_tokens_per_second": 694.58
    },
    {
      "epoch": 1.9512256760171849,
      "grad_norm": 1.5377455949783325,
      "learning_rate": 5.063291139240507e-07,
      "loss": 0.0610002763569355,
      "num_input_tokens_seen": 3636136,
      "step": 1931,
      "train_runtime": 5234.6709,
      "train_tokens_per_second": 694.626
    },
    {
      "epoch": 1.9522365428354815,
      "grad_norm": 1.8254579305648804,
      "learning_rate": 4.962025316455696e-07,
      "loss": 0.05765355005860329,
      "num_input_tokens_seen": 3637648,
      "step": 1932,
      "train_runtime": 5236.7901,
      "train_tokens_per_second": 694.633
    },
    {
      "epoch": 1.9532474096537782,
      "grad_norm": 1.6161350011825562,
      "learning_rate": 4.860759493670886e-07,
      "loss": 0.05342883989214897,
      "num_input_tokens_seen": 3639374,
      "step": 1933,
      "train_runtime": 5239.0551,
      "train_tokens_per_second": 694.662
    },
    {
      "epoch": 1.9542582764720748,
      "grad_norm": 1.8346822261810303,
      "learning_rate": 4.7594936708860764e-07,
      "loss": 0.05740257352590561,
      "num_input_tokens_seen": 3640998,
      "step": 1934,
      "train_runtime": 5241.2028,
      "train_tokens_per_second": 694.687
    },
    {
      "epoch": 1.9552691432903715,
      "grad_norm": 1.5896601676940918,
      "learning_rate": 4.658227848101266e-07,
      "loss": 0.05597972497344017,
      "num_input_tokens_seen": 3643372,
      "step": 1935,
      "train_runtime": 5243.4973,
      "train_tokens_per_second": 694.836
    },
    {
      "epoch": 1.9562800101086681,
      "grad_norm": 1.9486325979232788,
      "learning_rate": 4.556962025316456e-07,
      "loss": 0.07691022008657455,
      "num_input_tokens_seen": 3645376,
      "step": 1936,
      "train_runtime": 5245.7323,
      "train_tokens_per_second": 694.922
    },
    {
      "epoch": 1.9572908769269648,
      "grad_norm": 1.4963641166687012,
      "learning_rate": 4.455696202531646e-07,
      "loss": 0.07239636778831482,
      "num_input_tokens_seen": 3647236,
      "step": 1937,
      "train_runtime": 5247.9063,
      "train_tokens_per_second": 694.989
    },
    {
      "epoch": 1.9583017437452614,
      "grad_norm": 3.195930004119873,
      "learning_rate": 4.3544303797468356e-07,
      "loss": 0.08541422337293625,
      "num_input_tokens_seen": 3649178,
      "step": 1938,
      "train_runtime": 5250.1139,
      "train_tokens_per_second": 695.066
    },
    {
      "epoch": 1.959312610563558,
      "grad_norm": 1.5451648235321045,
      "learning_rate": 4.253164556962026e-07,
      "loss": 0.07356944680213928,
      "num_input_tokens_seen": 3651538,
      "step": 1939,
      "train_runtime": 5252.4399,
      "train_tokens_per_second": 695.208
    },
    {
      "epoch": 1.960323477381855,
      "grad_norm": 1.2501071691513062,
      "learning_rate": 4.1518987341772157e-07,
      "loss": 0.04627997800707817,
      "num_input_tokens_seen": 3653874,
      "step": 1940,
      "train_runtime": 5254.7628,
      "train_tokens_per_second": 695.345
    },
    {
      "epoch": 1.9613343442001516,
      "grad_norm": 2.022702217102051,
      "learning_rate": 4.0506329113924055e-07,
      "loss": 0.06978513300418854,
      "num_input_tokens_seen": 3655382,
      "step": 1941,
      "train_runtime": 5256.8885,
      "train_tokens_per_second": 695.351
    },
    {
      "epoch": 1.9623452110184483,
      "grad_norm": 1.7374645471572876,
      "learning_rate": 3.949367088607595e-07,
      "loss": 0.05413699895143509,
      "num_input_tokens_seen": 3656910,
      "step": 1942,
      "train_runtime": 5259.0068,
      "train_tokens_per_second": 695.361
    },
    {
      "epoch": 1.9633560778367452,
      "grad_norm": 1.7338082790374756,
      "learning_rate": 3.848101265822785e-07,
      "loss": 0.09035466611385345,
      "num_input_tokens_seen": 3658812,
      "step": 1943,
      "train_runtime": 5261.2642,
      "train_tokens_per_second": 695.424
    },
    {
      "epoch": 1.9643669446550418,
      "grad_norm": 1.4846956729888916,
      "learning_rate": 3.7468354430379753e-07,
      "loss": 0.054786644876003265,
      "num_input_tokens_seen": 3660618,
      "step": 1944,
      "train_runtime": 5263.5313,
      "train_tokens_per_second": 695.468
    },
    {
      "epoch": 1.9653778114733385,
      "grad_norm": 1.4887983798980713,
      "learning_rate": 3.645569620253165e-07,
      "loss": 0.04669232666492462,
      "num_input_tokens_seen": 3662816,
      "step": 1945,
      "train_runtime": 5265.7803,
      "train_tokens_per_second": 695.588
    },
    {
      "epoch": 1.966388678291635,
      "grad_norm": 1.803263783454895,
      "learning_rate": 3.544303797468355e-07,
      "loss": 0.062312956899404526,
      "num_input_tokens_seen": 3664510,
      "step": 1946,
      "train_runtime": 5267.9362,
      "train_tokens_per_second": 695.625
    },
    {
      "epoch": 1.9673995451099318,
      "grad_norm": 1.7119450569152832,
      "learning_rate": 3.4430379746835447e-07,
      "loss": 0.06583711504936218,
      "num_input_tokens_seen": 3666856,
      "step": 1947,
      "train_runtime": 5270.2745,
      "train_tokens_per_second": 695.762
    },
    {
      "epoch": 1.9684104119282284,
      "grad_norm": 1.9299792051315308,
      "learning_rate": 3.3417721518987345e-07,
      "loss": 0.06862461566925049,
      "num_input_tokens_seen": 3668630,
      "step": 1948,
      "train_runtime": 5272.4543,
      "train_tokens_per_second": 695.811
    },
    {
      "epoch": 1.969421278746525,
      "grad_norm": 1.6741821765899658,
      "learning_rate": 3.240506329113924e-07,
      "loss": 0.08114917576313019,
      "num_input_tokens_seen": 3670330,
      "step": 1949,
      "train_runtime": 5274.6507,
      "train_tokens_per_second": 695.843
    },
    {
      "epoch": 1.9704321455648217,
      "grad_norm": 1.4921934604644775,
      "learning_rate": 3.1392405063291146e-07,
      "loss": 0.0565418004989624,
      "num_input_tokens_seen": 3672008,
      "step": 1950,
      "train_runtime": 5276.7825,
      "train_tokens_per_second": 695.88
    },
    {
      "epoch": 1.9714430123831184,
      "grad_norm": 1.5650644302368164,
      "learning_rate": 3.037974683544304e-07,
      "loss": 0.07059604674577713,
      "num_input_tokens_seen": 3674074,
      "step": 1951,
      "train_runtime": 5285.7158,
      "train_tokens_per_second": 695.095
    },
    {
      "epoch": 1.9724538792014152,
      "grad_norm": 2.0831124782562256,
      "learning_rate": 2.9367088607594936e-07,
      "loss": 0.08382274210453033,
      "num_input_tokens_seen": 3676084,
      "step": 1952,
      "train_runtime": 5288.0106,
      "train_tokens_per_second": 695.173
    },
    {
      "epoch": 1.973464746019712,
      "grad_norm": 1.6316579580307007,
      "learning_rate": 2.835443037974684e-07,
      "loss": 0.06200099736452103,
      "num_input_tokens_seen": 3678154,
      "step": 1953,
      "train_runtime": 5290.2474,
      "train_tokens_per_second": 695.271
    },
    {
      "epoch": 1.9744756128380085,
      "grad_norm": 1.8981103897094727,
      "learning_rate": 2.7341772151898737e-07,
      "loss": 0.09692534059286118,
      "num_input_tokens_seen": 3680106,
      "step": 1954,
      "train_runtime": 5292.4631,
      "train_tokens_per_second": 695.348
    },
    {
      "epoch": 1.9754864796563054,
      "grad_norm": 1.4167271852493286,
      "learning_rate": 2.6329113924050635e-07,
      "loss": 0.03817079961299896,
      "num_input_tokens_seen": 3682152,
      "step": 1955,
      "train_runtime": 5294.7623,
      "train_tokens_per_second": 695.433
    },
    {
      "epoch": 1.976497346474602,
      "grad_norm": 1.6415997743606567,
      "learning_rate": 2.5316455696202533e-07,
      "loss": 0.05982155352830887,
      "num_input_tokens_seen": 3683756,
      "step": 1956,
      "train_runtime": 5296.9371,
      "train_tokens_per_second": 695.45
    },
    {
      "epoch": 1.9775082132928987,
      "grad_norm": 2.153597831726074,
      "learning_rate": 2.430379746835443e-07,
      "loss": 0.08247441798448563,
      "num_input_tokens_seen": 3685400,
      "step": 1957,
      "train_runtime": 5299.0824,
      "train_tokens_per_second": 695.479
    },
    {
      "epoch": 1.9785190801111954,
      "grad_norm": 2.303065776824951,
      "learning_rate": 2.329113924050633e-07,
      "loss": 0.09317882359027863,
      "num_input_tokens_seen": 3687110,
      "step": 1958,
      "train_runtime": 5301.2213,
      "train_tokens_per_second": 695.521
    },
    {
      "epoch": 1.979529946929492,
      "grad_norm": 1.790480136871338,
      "learning_rate": 2.227848101265823e-07,
      "loss": 0.05739545449614525,
      "num_input_tokens_seen": 3688784,
      "step": 1959,
      "train_runtime": 5303.3326,
      "train_tokens_per_second": 695.56
    },
    {
      "epoch": 1.9805408137477887,
      "grad_norm": 1.775923490524292,
      "learning_rate": 2.126582278481013e-07,
      "loss": 0.09501507878303528,
      "num_input_tokens_seen": 3690750,
      "step": 1960,
      "train_runtime": 5305.5403,
      "train_tokens_per_second": 695.641
    },
    {
      "epoch": 1.9815516805660853,
      "grad_norm": 1.9084643125534058,
      "learning_rate": 2.0253164556962027e-07,
      "loss": 0.07032062858343124,
      "num_input_tokens_seen": 3692700,
      "step": 1961,
      "train_runtime": 5307.7393,
      "train_tokens_per_second": 695.72
    },
    {
      "epoch": 1.982562547384382,
      "grad_norm": 1.9343383312225342,
      "learning_rate": 1.9240506329113925e-07,
      "loss": 0.06499311327934265,
      "num_input_tokens_seen": 3694398,
      "step": 1962,
      "train_runtime": 5309.9181,
      "train_tokens_per_second": 695.754
    },
    {
      "epoch": 1.9835734142026786,
      "grad_norm": 2.3557677268981934,
      "learning_rate": 1.8227848101265826e-07,
      "loss": 0.09849251061677933,
      "num_input_tokens_seen": 3695908,
      "step": 1963,
      "train_runtime": 5312.0324,
      "train_tokens_per_second": 695.762
    },
    {
      "epoch": 1.9845842810209755,
      "grad_norm": 1.6780991554260254,
      "learning_rate": 1.7215189873417723e-07,
      "loss": 0.07726100087165833,
      "num_input_tokens_seen": 3698094,
      "step": 1964,
      "train_runtime": 5314.302,
      "train_tokens_per_second": 695.876
    },
    {
      "epoch": 1.9855951478392722,
      "grad_norm": 1.810806155204773,
      "learning_rate": 1.620253164556962e-07,
      "loss": 0.08548904210329056,
      "num_input_tokens_seen": 3700184,
      "step": 1965,
      "train_runtime": 5316.6599,
      "train_tokens_per_second": 695.96
    },
    {
      "epoch": 1.986606014657569,
      "grad_norm": 1.9527696371078491,
      "learning_rate": 1.518987341772152e-07,
      "loss": 0.06699540466070175,
      "num_input_tokens_seen": 3701876,
      "step": 1966,
      "train_runtime": 5318.882,
      "train_tokens_per_second": 695.988
    },
    {
      "epoch": 1.9876168814758657,
      "grad_norm": 1.7017799615859985,
      "learning_rate": 1.417721518987342e-07,
      "loss": 0.06659930944442749,
      "num_input_tokens_seen": 3703512,
      "step": 1967,
      "train_runtime": 5321.0459,
      "train_tokens_per_second": 696.012
    },
    {
      "epoch": 1.9886277482941623,
      "grad_norm": 1.539168119430542,
      "learning_rate": 1.3164556962025317e-07,
      "loss": 0.0572090670466423,
      "num_input_tokens_seen": 3705336,
      "step": 1968,
      "train_runtime": 5323.2245,
      "train_tokens_per_second": 696.07
    },
    {
      "epoch": 1.989638615112459,
      "grad_norm": 2.168595790863037,
      "learning_rate": 1.2151898734177215e-07,
      "loss": 0.07186886668205261,
      "num_input_tokens_seen": 3707076,
      "step": 1969,
      "train_runtime": 5325.4064,
      "train_tokens_per_second": 696.111
    },
    {
      "epoch": 1.9906494819307556,
      "grad_norm": 1.9676153659820557,
      "learning_rate": 1.1139240506329114e-07,
      "loss": 0.08913262188434601,
      "num_input_tokens_seen": 3708656,
      "step": 1970,
      "train_runtime": 5327.5351,
      "train_tokens_per_second": 696.13
    },
    {
      "epoch": 1.9916603487490523,
      "grad_norm": 1.9166979789733887,
      "learning_rate": 1.0126582278481014e-07,
      "loss": 0.07388445734977722,
      "num_input_tokens_seen": 3710238,
      "step": 1971,
      "train_runtime": 5329.6526,
      "train_tokens_per_second": 696.15
    },
    {
      "epoch": 1.992671215567349,
      "grad_norm": 1.543239951133728,
      "learning_rate": 9.113924050632913e-08,
      "loss": 0.06998427212238312,
      "num_input_tokens_seen": 3712330,
      "step": 1972,
      "train_runtime": 5331.8629,
      "train_tokens_per_second": 696.254
    },
    {
      "epoch": 1.9936820823856456,
      "grad_norm": 2.1475512981414795,
      "learning_rate": 8.10126582278481e-08,
      "loss": 0.07976119220256805,
      "num_input_tokens_seen": 3713826,
      "step": 1973,
      "train_runtime": 5333.9966,
      "train_tokens_per_second": 696.256
    },
    {
      "epoch": 1.9946929492039422,
      "grad_norm": 1.6464298963546753,
      "learning_rate": 7.08860759493671e-08,
      "loss": 0.06630692631006241,
      "num_input_tokens_seen": 3716112,
      "step": 1974,
      "train_runtime": 5336.2694,
      "train_tokens_per_second": 696.388
    },
    {
      "epoch": 1.9957038160222391,
      "grad_norm": 1.7383606433868408,
      "learning_rate": 6.075949367088608e-08,
      "loss": 0.07431012392044067,
      "num_input_tokens_seen": 3717602,
      "step": 1975,
      "train_runtime": 5338.396,
      "train_tokens_per_second": 696.389
    },
    {
      "epoch": 1.9967146828405358,
      "grad_norm": 1.1304759979248047,
      "learning_rate": 5.063291139240507e-08,
      "loss": 0.03536690026521683,
      "num_input_tokens_seen": 3719202,
      "step": 1976,
      "train_runtime": 5340.5275,
      "train_tokens_per_second": 696.411
    },
    {
      "epoch": 1.9977255496588324,
      "grad_norm": 1.9554280042648315,
      "learning_rate": 4.050632911392405e-08,
      "loss": 0.07763595134019852,
      "num_input_tokens_seen": 3720932,
      "step": 1977,
      "train_runtime": 5342.7,
      "train_tokens_per_second": 696.452
    },
    {
      "epoch": 1.9987364164771293,
      "grad_norm": 1.8227564096450806,
      "learning_rate": 3.037974683544304e-08,
      "loss": 0.06695139408111572,
      "num_input_tokens_seen": 3722462,
      "step": 1978,
      "train_runtime": 5344.7916,
      "train_tokens_per_second": 696.465
    },
    {
      "epoch": 1.999747283295426,
      "grad_norm": 1.4120136499404907,
      "learning_rate": 2.0253164556962027e-08,
      "loss": 0.07311323285102844,
      "num_input_tokens_seen": 3724864,
      "step": 1979,
      "train_runtime": 5347.1284,
      "train_tokens_per_second": 696.61
    },
    {
      "epoch": 2.0,
      "grad_norm": 18.52170181274414,
      "learning_rate": 1.0126582278481013e-08,
      "loss": 0.13481095433235168,
      "num_input_tokens_seen": 3724995,
      "step": 1980,
      "train_runtime": 5347.6879,
      "train_tokens_per_second": 696.562
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.39007651805877686,
      "eval_runtime": 52.4712,
      "eval_samples_per_second": 16.752,
      "eval_steps_per_second": 8.386,
      "num_input_tokens_seen": 3724995,
      "step": 1980
    }
  ],
  "logging_steps": 1,
  "max_steps": 1980,
  "num_input_tokens_seen": 3724995,
  "num_train_epochs": 2,
  "save_steps": 30,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.334151885669696e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}