{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 5,
  "global_step": 494,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004048582995951417,
      "grad_norm": 57.285691125278106,
      "learning_rate": 2e-05,
      "loss": 20.1406,
      "mean_token_accuracy": 0.5111725330352783,
      "num_tokens": 72429.0,
      "step": 1
    },
    {
      "epoch": 0.008097165991902834,
      "grad_norm": 41.99363066045992,
      "learning_rate": 1.9959514170040488e-05,
      "loss": 18.6875,
      "mean_token_accuracy": 0.5240641832351685,
      "num_tokens": 136634.0,
      "step": 2
    },
    {
      "epoch": 0.012145748987854251,
      "grad_norm": 38.5726807470561,
      "learning_rate": 1.9919028340080974e-05,
      "loss": 19.4375,
      "mean_token_accuracy": 0.5193748474121094,
      "num_tokens": 204776.0,
      "step": 3
    },
    {
      "epoch": 0.016194331983805668,
      "grad_norm": 36.354880235444575,
      "learning_rate": 1.987854251012146e-05,
      "loss": 18.5859,
      "mean_token_accuracy": 0.5202323198318481,
      "num_tokens": 271293.0,
      "step": 4
    },
    {
      "epoch": 0.020242914979757085,
      "grad_norm": 28.213868324309114,
      "learning_rate": 1.9838056680161946e-05,
      "loss": 17.4609,
      "mean_token_accuracy": 0.5324159860610962,
      "num_tokens": 332546.0,
      "step": 5
    },
    {
      "epoch": 0.020242914979757085,
      "eval_loss": 2.1596875190734863,
      "eval_mean_token_accuracy": 0.5420514196157455,
      "eval_num_tokens": 332546.0,
      "eval_runtime": 3.902,
      "eval_samples_per_second": 51.256,
      "eval_steps_per_second": 1.025,
      "step": 5
    },
    {
      "epoch": 0.024291497975708502,
      "grad_norm": 27.134077029097234,
      "learning_rate": 1.979757085020243e-05,
      "loss": 17.4922,
      "mean_token_accuracy": 0.5339857339859009,
      "num_tokens": 402301.0,
      "step": 6
    },
    {
      "epoch": 0.02834008097165992,
      "grad_norm": 18.222480703668843,
      "learning_rate": 1.9757085020242915e-05,
      "loss": 17.1172,
      "mean_token_accuracy": 0.5464459657669067,
      "num_tokens": 461692.0,
      "step": 7
    },
    {
      "epoch": 0.032388663967611336,
      "grad_norm": 15.082689481608242,
      "learning_rate": 1.9716599190283405e-05,
      "loss": 16.8516,
      "mean_token_accuracy": 0.5461285710334778,
      "num_tokens": 538227.0,
      "step": 8
    },
    {
      "epoch": 0.03643724696356275,
      "grad_norm": 12.197205828835985,
      "learning_rate": 1.9676113360323887e-05,
      "loss": 16.1562,
      "mean_token_accuracy": 0.5573914647102356,
      "num_tokens": 603109.0,
      "step": 9
    },
    {
      "epoch": 0.04048582995951417,
      "grad_norm": 13.717242431104397,
      "learning_rate": 1.9635627530364373e-05,
      "loss": 15.7266,
      "mean_token_accuracy": 0.5678688287734985,
      "num_tokens": 659598.0,
      "step": 10
    },
    {
      "epoch": 0.04048582995951417,
      "eval_loss": 2.0374999046325684,
      "eval_mean_token_accuracy": 0.5590415745973587,
      "eval_num_tokens": 659598.0,
      "eval_runtime": 0.6198,
      "eval_samples_per_second": 322.698,
      "eval_steps_per_second": 6.454,
      "step": 10
    },
    {
      "epoch": 0.044534412955465584,
      "grad_norm": 12.306486360056866,
      "learning_rate": 1.959514170040486e-05,
      "loss": 15.6484,
      "mean_token_accuracy": 0.5803827047348022,
      "num_tokens": 714481.0,
      "step": 11
    },
    {
      "epoch": 0.048582995951417005,
      "grad_norm": 11.708980436467652,
      "learning_rate": 1.9554655870445346e-05,
      "loss": 15.4062,
      "mean_token_accuracy": 0.578364908695221,
      "num_tokens": 778432.0,
      "step": 12
    },
    {
      "epoch": 0.05263157894736842,
      "grad_norm": 11.35292985675568,
      "learning_rate": 1.9514170040485832e-05,
      "loss": 16.2422,
      "mean_token_accuracy": 0.5603257417678833,
      "num_tokens": 843576.0,
      "step": 13
    },
    {
      "epoch": 0.05668016194331984,
      "grad_norm": 11.721306843681429,
      "learning_rate": 1.9473684210526318e-05,
      "loss": 15.4141,
      "mean_token_accuracy": 0.5766724944114685,
      "num_tokens": 914252.0,
      "step": 14
    },
    {
      "epoch": 0.06072874493927125,
      "grad_norm": 11.371453682604804,
      "learning_rate": 1.94331983805668e-05,
      "loss": 15.2969,
      "mean_token_accuracy": 0.5723595023155212,
      "num_tokens": 976160.0,
      "step": 15
    },
    {
      "epoch": 0.06072874493927125,
      "eval_loss": 1.9924999475479126,
      "eval_mean_token_accuracy": 0.5654653459787369,
      "eval_num_tokens": 976160.0,
      "eval_runtime": 0.6214,
      "eval_samples_per_second": 321.855,
      "eval_steps_per_second": 6.437,
      "step": 15
    },
    {
      "epoch": 0.06477732793522267,
      "grad_norm": 13.380354993808568,
      "learning_rate": 1.939271255060729e-05,
      "loss": 17.2188,
      "mean_token_accuracy": 0.5414544939994812,
      "num_tokens": 1046132.0,
      "step": 16
    },
    {
      "epoch": 0.06882591093117409,
      "grad_norm": 12.756822173097444,
      "learning_rate": 1.9352226720647776e-05,
      "loss": 15.7188,
      "mean_token_accuracy": 0.5702171325683594,
      "num_tokens": 1111408.0,
      "step": 17
    },
    {
      "epoch": 0.0728744939271255,
      "grad_norm": 10.541413178483012,
      "learning_rate": 1.931174089068826e-05,
      "loss": 16.6406,
      "mean_token_accuracy": 0.5545904040336609,
      "num_tokens": 1181008.0,
      "step": 18
    },
    {
      "epoch": 0.07692307692307693,
      "grad_norm": 9.058216089424022,
      "learning_rate": 1.9271255060728745e-05,
      "loss": 15.4922,
      "mean_token_accuracy": 0.5672646164894104,
      "num_tokens": 1242895.0,
      "step": 19
    },
    {
      "epoch": 0.08097165991902834,
      "grad_norm": 8.224696140023758,
      "learning_rate": 1.923076923076923e-05,
      "loss": 14.5,
      "mean_token_accuracy": 0.5876371264457703,
      "num_tokens": 1321339.0,
      "step": 20
    },
    {
      "epoch": 0.08097165991902834,
      "eval_loss": 1.959375023841858,
      "eval_mean_token_accuracy": 0.5695992410182953,
      "eval_num_tokens": 1321339.0,
      "eval_runtime": 0.6175,
      "eval_samples_per_second": 323.911,
      "eval_steps_per_second": 6.478,
      "step": 20
    },
    {
      "epoch": 0.08502024291497975,
      "grad_norm": 9.194280544253612,
      "learning_rate": 1.9190283400809718e-05,
      "loss": 15.7969,
      "mean_token_accuracy": 0.5668182373046875,
      "num_tokens": 1388271.0,
      "step": 21
    },
    {
      "epoch": 0.08906882591093117,
      "grad_norm": 9.173673841849936,
      "learning_rate": 1.9149797570850204e-05,
      "loss": 16.1875,
      "mean_token_accuracy": 0.5593817234039307,
      "num_tokens": 1456462.0,
      "step": 22
    },
    {
      "epoch": 0.0931174089068826,
      "grad_norm": 9.571035474364693,
      "learning_rate": 1.910931174089069e-05,
      "loss": 15.8516,
      "mean_token_accuracy": 0.5684004426002502,
      "num_tokens": 1518799.0,
      "step": 23
    },
    {
      "epoch": 0.09716599190283401,
      "grad_norm": 7.975715179302099,
      "learning_rate": 1.9068825910931176e-05,
      "loss": 14.9844,
      "mean_token_accuracy": 0.5802298188209534,
      "num_tokens": 1588301.0,
      "step": 24
    },
    {
      "epoch": 0.10121457489878542,
      "grad_norm": 8.06780763084373,
      "learning_rate": 1.9028340080971662e-05,
      "loss": 15.9766,
      "mean_token_accuracy": 0.5611966848373413,
      "num_tokens": 1665044.0,
      "step": 25
    },
    {
      "epoch": 0.10121457489878542,
      "eval_loss": 1.9378124475479126,
      "eval_mean_token_accuracy": 0.571557804942131,
      "eval_num_tokens": 1665044.0,
      "eval_runtime": 0.6132,
      "eval_samples_per_second": 326.145,
      "eval_steps_per_second": 6.523,
      "step": 25
    },
    {
      "epoch": 0.10526315789473684,
      "grad_norm": 9.003512171360999,
      "learning_rate": 1.8987854251012148e-05,
      "loss": 15.9062,
      "mean_token_accuracy": 0.5626139640808105,
      "num_tokens": 1731459.0,
      "step": 26
    },
    {
      "epoch": 0.10931174089068826,
      "grad_norm": 8.045437012882436,
      "learning_rate": 1.894736842105263e-05,
      "loss": 14.9844,
      "mean_token_accuracy": 0.5767053365707397,
      "num_tokens": 1794413.0,
      "step": 27
    },
    {
      "epoch": 0.11336032388663968,
      "grad_norm": 7.791818905787103,
      "learning_rate": 1.8906882591093117e-05,
      "loss": 16.0703,
      "mean_token_accuracy": 0.5646764636039734,
      "num_tokens": 1864000.0,
      "step": 28
    },
    {
      "epoch": 0.11740890688259109,
      "grad_norm": 7.899247736502825,
      "learning_rate": 1.8866396761133607e-05,
      "loss": 14.6562,
      "mean_token_accuracy": 0.5848944187164307,
      "num_tokens": 1926306.0,
      "step": 29
    },
    {
      "epoch": 0.1214574898785425,
      "grad_norm": 9.046927442368185,
      "learning_rate": 1.882591093117409e-05,
      "loss": 15.0781,
      "mean_token_accuracy": 0.5783066749572754,
      "num_tokens": 2001619.0,
      "step": 30
    },
    {
      "epoch": 0.1214574898785425,
      "eval_loss": 1.9221874475479126,
      "eval_mean_token_accuracy": 0.5738949328660965,
      "eval_num_tokens": 2001619.0,
      "eval_runtime": 0.6128,
      "eval_samples_per_second": 326.396,
      "eval_steps_per_second": 6.528,
      "step": 30
    },
    {
      "epoch": 0.12550607287449392,
      "grad_norm": 8.961465962055085,
      "learning_rate": 1.8785425101214576e-05,
      "loss": 16.1797,
      "mean_token_accuracy": 0.5587651133537292,
      "num_tokens": 2066109.0,
      "step": 31
    },
    {
      "epoch": 0.12955465587044535,
      "grad_norm": 7.562379064010745,
      "learning_rate": 1.874493927125506e-05,
      "loss": 15.4922,
      "mean_token_accuracy": 0.5650998950004578,
      "num_tokens": 2142786.0,
      "step": 32
    },
    {
      "epoch": 0.13360323886639677,
      "grad_norm": 8.562712496801407,
      "learning_rate": 1.8704453441295548e-05,
      "loss": 14.9219,
      "mean_token_accuracy": 0.5784920454025269,
      "num_tokens": 2199722.0,
      "step": 33
    },
    {
      "epoch": 0.13765182186234817,
      "grad_norm": 9.988663916747383,
      "learning_rate": 1.8663967611336034e-05,
      "loss": 17.1641,
      "mean_token_accuracy": 0.5491723418235779,
      "num_tokens": 2275906.0,
      "step": 34
    },
    {
      "epoch": 0.1417004048582996,
      "grad_norm": 7.709526019356417,
      "learning_rate": 1.862348178137652e-05,
      "loss": 15.7656,
      "mean_token_accuracy": 0.5621457099914551,
      "num_tokens": 2358888.0,
      "step": 35
    },
    {
      "epoch": 0.1417004048582996,
      "eval_loss": 1.9071874618530273,
      "eval_mean_token_accuracy": 0.5757417529821396,
      "eval_num_tokens": 2358888.0,
      "eval_runtime": 0.6136,
      "eval_samples_per_second": 325.959,
      "eval_steps_per_second": 6.519,
      "step": 35
    },
    {
      "epoch": 0.145748987854251,
      "grad_norm": 8.03798824892872,
      "learning_rate": 1.8582995951417006e-05,
      "loss": 15.8047,
      "mean_token_accuracy": 0.5641842484474182,
      "num_tokens": 2428938.0,
      "step": 36
    },
    {
      "epoch": 0.14979757085020243,
      "grad_norm": 8.32083756121802,
      "learning_rate": 1.8542510121457492e-05,
      "loss": 15.2578,
      "mean_token_accuracy": 0.5743291974067688,
      "num_tokens": 2493963.0,
      "step": 37
    },
    {
      "epoch": 0.15384615384615385,
      "grad_norm": 8.354287332937643,
      "learning_rate": 1.850202429149798e-05,
      "loss": 15.4141,
      "mean_token_accuracy": 0.5688024759292603,
      "num_tokens": 2559904.0,
      "step": 38
    },
    {
      "epoch": 0.15789473684210525,
      "grad_norm": 8.579778574791483,
      "learning_rate": 1.8461538461538465e-05,
      "loss": 15.5078,
      "mean_token_accuracy": 0.5710554122924805,
      "num_tokens": 2614334.0,
      "step": 39
    },
    {
      "epoch": 0.16194331983805668,
      "grad_norm": 8.994514132460578,
      "learning_rate": 1.8421052631578947e-05,
      "loss": 14.1875,
      "mean_token_accuracy": 0.5983923673629761,
      "num_tokens": 2677721.0,
      "step": 40
    },
    {
      "epoch": 0.16194331983805668,
      "eval_loss": 1.894374966621399,
      "eval_mean_token_accuracy": 0.577251747250557,
      "eval_num_tokens": 2677721.0,
      "eval_runtime": 0.6381,
      "eval_samples_per_second": 313.434,
      "eval_steps_per_second": 6.269,
      "step": 40
    },
    {
      "epoch": 0.1659919028340081,
      "grad_norm": 9.563884911162102,
      "learning_rate": 1.8380566801619433e-05,
      "loss": 15.8906,
      "mean_token_accuracy": 0.5601885914802551,
      "num_tokens": 2736542.0,
      "step": 41
    },
    {
      "epoch": 0.1700404858299595,
      "grad_norm": 8.720473365423832,
      "learning_rate": 1.8340080971659923e-05,
      "loss": 14.7969,
      "mean_token_accuracy": 0.5805999636650085,
      "num_tokens": 2795142.0,
      "step": 42
    },
    {
      "epoch": 0.17408906882591094,
      "grad_norm": 7.269013839644517,
      "learning_rate": 1.8299595141700406e-05,
      "loss": 14.7031,
      "mean_token_accuracy": 0.578359067440033,
      "num_tokens": 2863022.0,
      "step": 43
    },
    {
      "epoch": 0.17813765182186234,
      "grad_norm": 7.272014286398225,
      "learning_rate": 1.8259109311740892e-05,
      "loss": 14.8516,
      "mean_token_accuracy": 0.5769524574279785,
      "num_tokens": 2931180.0,
      "step": 44
    },
    {
      "epoch": 0.18218623481781376,
      "grad_norm": 8.851418056520739,
      "learning_rate": 1.8218623481781378e-05,
      "loss": 15.3594,
      "mean_token_accuracy": 0.5745086669921875,
      "num_tokens": 2995713.0,
      "step": 45
    },
    {
      "epoch": 0.18218623481781376,
      "eval_loss": 1.885312557220459,
      "eval_mean_token_accuracy": 0.5781335979700089,
      "eval_num_tokens": 2995713.0,
      "eval_runtime": 0.6226,
      "eval_samples_per_second": 321.249,
      "eval_steps_per_second": 6.425,
      "step": 45
    },
    {
      "epoch": 0.1862348178137652,
      "grad_norm": 8.587589121567573,
      "learning_rate": 1.8178137651821864e-05,
      "loss": 14.9531,
      "mean_token_accuracy": 0.5738093852996826,
      "num_tokens": 3066642.0,
      "step": 46
    },
    {
      "epoch": 0.1902834008097166,
      "grad_norm": 7.890409770531286,
      "learning_rate": 1.813765182186235e-05,
      "loss": 15.1016,
      "mean_token_accuracy": 0.5790749192237854,
      "num_tokens": 3128603.0,
      "step": 47
    },
    {
      "epoch": 0.19433198380566802,
      "grad_norm": 7.793678892191199,
      "learning_rate": 1.8097165991902836e-05,
      "loss": 14.8281,
      "mean_token_accuracy": 0.5814535021781921,
      "num_tokens": 3187449.0,
      "step": 48
    },
    {
      "epoch": 0.19838056680161945,
      "grad_norm": 7.256161611883218,
      "learning_rate": 1.805668016194332e-05,
      "loss": 14.9062,
      "mean_token_accuracy": 0.5790348649024963,
      "num_tokens": 3262113.0,
      "step": 49
    },
    {
      "epoch": 0.20242914979757085,
      "grad_norm": 7.3121519125177254,
      "learning_rate": 1.801619433198381e-05,
      "loss": 14.6484,
      "mean_token_accuracy": 0.5866994261741638,
      "num_tokens": 3332956.0,
      "step": 50
    },
    {
      "epoch": 0.20242914979757085,
      "eval_loss": 1.8756250143051147,
      "eval_mean_token_accuracy": 0.5786750316619873,
      "eval_num_tokens": 3332956.0,
      "eval_runtime": 0.616,
      "eval_samples_per_second": 324.656,
      "eval_steps_per_second": 6.493,
      "step": 50
    },
    {
      "epoch": 0.20647773279352227,
      "grad_norm": 7.4168976628601095,
      "learning_rate": 1.7975708502024295e-05,
      "loss": 15.9141,
      "mean_token_accuracy": 0.5570866465568542,
      "num_tokens": 3402686.0,
      "step": 51
    },
    {
      "epoch": 0.21052631578947367,
      "grad_norm": 8.132782257191536,
      "learning_rate": 1.7935222672064778e-05,
      "loss": 13.6328,
      "mean_token_accuracy": 0.6026627421379089,
      "num_tokens": 3463439.0,
      "step": 52
    },
    {
      "epoch": 0.2145748987854251,
      "grad_norm": 8.722536735498021,
      "learning_rate": 1.7894736842105264e-05,
      "loss": 15.9922,
      "mean_token_accuracy": 0.5636070370674133,
      "num_tokens": 3534808.0,
      "step": 53
    },
    {
      "epoch": 0.21862348178137653,
      "grad_norm": 9.5925704794369,
      "learning_rate": 1.785425101214575e-05,
      "loss": 16.5859,
      "mean_token_accuracy": 0.5475066900253296,
      "num_tokens": 3598221.0,
      "step": 54
    },
    {
      "epoch": 0.22267206477732793,
      "grad_norm": 8.403059172595993,
      "learning_rate": 1.7813765182186236e-05,
      "loss": 15.0859,
      "mean_token_accuracy": 0.5766125321388245,
      "num_tokens": 3667399.0,
      "step": 55
    },
    {
      "epoch": 0.22267206477732793,
      "eval_loss": 1.8668750524520874,
      "eval_mean_token_accuracy": 0.5796706676483154,
      "eval_num_tokens": 3667399.0,
      "eval_runtime": 0.6158,
      "eval_samples_per_second": 324.774,
      "eval_steps_per_second": 6.495,
      "step": 55
    },
    {
      "epoch": 0.22672064777327935,
      "grad_norm": 7.854856510906883,
      "learning_rate": 1.7773279352226722e-05,
      "loss": 14.5781,
      "mean_token_accuracy": 0.5866025686264038,
      "num_tokens": 3741537.0,
      "step": 56
    },
    {
      "epoch": 0.23076923076923078,
      "grad_norm": 9.026696458622231,
      "learning_rate": 1.7732793522267208e-05,
      "loss": 15.4062,
      "mean_token_accuracy": 0.5732284784317017,
      "num_tokens": 3817002.0,
      "step": 57
    },
    {
      "epoch": 0.23481781376518218,
      "grad_norm": 8.576517147098428,
      "learning_rate": 1.7692307692307694e-05,
      "loss": 14.7969,
      "mean_token_accuracy": 0.577850878238678,
      "num_tokens": 3875145.0,
      "step": 58
    },
    {
      "epoch": 0.2388663967611336,
      "grad_norm": 9.62049525740993,
      "learning_rate": 1.765182186234818e-05,
      "loss": 14.9531,
      "mean_token_accuracy": 0.5829423069953918,
      "num_tokens": 3947880.0,
      "step": 59
    },
    {
      "epoch": 0.242914979757085,
      "grad_norm": 7.546747277273659,
      "learning_rate": 1.7611336032388667e-05,
      "loss": 13.5547,
      "mean_token_accuracy": 0.6001456379890442,
      "num_tokens": 4015226.0,
      "step": 60
    },
    {
      "epoch": 0.242914979757085,
      "eval_loss": 1.859375,
      "eval_mean_token_accuracy": 0.5801091939210892,
      "eval_num_tokens": 4015226.0,
      "eval_runtime": 0.6356,
      "eval_samples_per_second": 314.646,
      "eval_steps_per_second": 6.293,
      "step": 60
    },
    {
      "epoch": 0.24696356275303644,
      "grad_norm": 8.432094147883522,
      "learning_rate": 1.757085020242915e-05,
      "loss": 14.3438,
      "mean_token_accuracy": 0.5897209048271179,
      "num_tokens": 4076385.0,
      "step": 61
    },
    {
      "epoch": 0.25101214574898784,
      "grad_norm": 8.727387350173785,
      "learning_rate": 1.7530364372469636e-05,
      "loss": 14.0859,
      "mean_token_accuracy": 0.5868391394615173,
      "num_tokens": 4130336.0,
      "step": 62
    },
    {
      "epoch": 0.2550607287449393,
      "grad_norm": 7.853461601699973,
      "learning_rate": 1.7489878542510125e-05,
      "loss": 14.5625,
      "mean_token_accuracy": 0.5877260565757751,
      "num_tokens": 4205714.0,
      "step": 63
    },
    {
      "epoch": 0.2591093117408907,
      "grad_norm": 10.618114952893986,
      "learning_rate": 1.7449392712550608e-05,
      "loss": 14.7891,
      "mean_token_accuracy": 0.5806258916854858,
      "num_tokens": 4276016.0,
      "step": 64
    },
    {
      "epoch": 0.2631578947368421,
      "grad_norm": 8.368519887093084,
      "learning_rate": 1.7408906882591094e-05,
      "loss": 14.6094,
      "mean_token_accuracy": 0.5852103233337402,
      "num_tokens": 4341166.0,
      "step": 65
    },
    {
      "epoch": 0.2631578947368421,
      "eval_loss": 1.8528125286102295,
      "eval_mean_token_accuracy": 0.5809510350227356,
      "eval_num_tokens": 4341166.0,
      "eval_runtime": 0.6479,
      "eval_samples_per_second": 308.698,
      "eval_steps_per_second": 6.174,
      "step": 65
    },
    {
      "epoch": 0.26720647773279355,
      "grad_norm": 8.961964950224539,
      "learning_rate": 1.736842105263158e-05,
      "loss": 13.7656,
      "mean_token_accuracy": 0.5963848829269409,
      "num_tokens": 4394174.0,
      "step": 66
    },
    {
      "epoch": 0.27125506072874495,
      "grad_norm": 8.15926394629814,
      "learning_rate": 1.7327935222672066e-05,
      "loss": 14.8906,
      "mean_token_accuracy": 0.5760456323623657,
      "num_tokens": 4463144.0,
      "step": 67
    },
    {
      "epoch": 0.27530364372469635,
      "grad_norm": 7.093290877100628,
      "learning_rate": 1.7287449392712552e-05,
      "loss": 14.375,
      "mean_token_accuracy": 0.5898444056510925,
      "num_tokens": 4533140.0,
      "step": 68
    },
    {
      "epoch": 0.2793522267206478,
      "grad_norm": 8.262028322497853,
      "learning_rate": 1.724696356275304e-05,
      "loss": 15.3672,
      "mean_token_accuracy": 0.5701308250427246,
      "num_tokens": 4596279.0,
      "step": 69
    },
    {
      "epoch": 0.2834008097165992,
      "grad_norm": 7.614011800242887,
      "learning_rate": 1.720647773279352e-05,
      "loss": 13.5547,
      "mean_token_accuracy": 0.5994280576705933,
      "num_tokens": 4661033.0,
      "step": 70
    },
    {
      "epoch": 0.2834008097165992,
      "eval_loss": 1.8446874618530273,
      "eval_mean_token_accuracy": 0.5816306471824646,
      "eval_num_tokens": 4661033.0,
      "eval_runtime": 0.6219,
      "eval_samples_per_second": 321.57,
      "eval_steps_per_second": 6.431,
      "step": 70
    },
    {
      "epoch": 0.2874493927125506,
      "grad_norm": 8.501649392228297,
      "learning_rate": 1.716599190283401e-05,
      "loss": 14.5312,
      "mean_token_accuracy": 0.5782850384712219,
      "num_tokens": 4730708.0,
      "step": 71
    },
    {
      "epoch": 0.291497975708502,
      "grad_norm": 7.022173928165628,
      "learning_rate": 1.7125506072874497e-05,
      "loss": 14.8203,
      "mean_token_accuracy": 0.5837699770927429,
      "num_tokens": 4804253.0,
      "step": 72
    },
    {
      "epoch": 0.29554655870445345,
      "grad_norm": 7.489428149228862,
      "learning_rate": 1.708502024291498e-05,
      "loss": 13.9219,
      "mean_token_accuracy": 0.594904363155365,
      "num_tokens": 4873160.0,
      "step": 73
    },
    {
      "epoch": 0.29959514170040485,
      "grad_norm": 8.856719704451924,
      "learning_rate": 1.7044534412955466e-05,
      "loss": 15.5781,
      "mean_token_accuracy": 0.5685590505599976,
      "num_tokens": 4942325.0,
      "step": 74
    },
    {
      "epoch": 0.30364372469635625,
      "grad_norm": 7.197286431718064,
      "learning_rate": 1.7004048582995952e-05,
      "loss": 14.4922,
      "mean_token_accuracy": 0.5876879096031189,
      "num_tokens": 5014497.0,
      "step": 75
    },
    {
      "epoch": 0.30364372469635625,
      "eval_loss": 1.8378125429153442,
      "eval_mean_token_accuracy": 0.5824010521173477,
      "eval_num_tokens": 5014497.0,
      "eval_runtime": 0.6338,
      "eval_samples_per_second": 315.564,
      "eval_steps_per_second": 6.311,
      "step": 75
    },
    {
      "epoch": 0.3076923076923077,
      "grad_norm": 9.04302504332728,
      "learning_rate": 1.6963562753036438e-05,
      "loss": 15.4844,
      "mean_token_accuracy": 0.571445107460022,
      "num_tokens": 5075118.0,
      "step": 76
    },
    {
      "epoch": 0.3117408906882591,
      "grad_norm": 7.558608724226377,
      "learning_rate": 1.6923076923076924e-05,
      "loss": 14.5156,
      "mean_token_accuracy": 0.5831646919250488,
      "num_tokens": 5142386.0,
      "step": 77
    },
    {
      "epoch": 0.3157894736842105,
      "grad_norm": 7.8941192536661955,
      "learning_rate": 1.688259109311741e-05,
      "loss": 13.6406,
      "mean_token_accuracy": 0.5994213223457336,
      "num_tokens": 5208462.0,
      "step": 78
    },
    {
      "epoch": 0.31983805668016196,
      "grad_norm": 7.560860397902134,
      "learning_rate": 1.6842105263157896e-05,
      "loss": 15.3047,
      "mean_token_accuracy": 0.5730275511741638,
      "num_tokens": 5285004.0,
      "step": 79
    },
    {
      "epoch": 0.32388663967611336,
      "grad_norm": 7.257123559107278,
      "learning_rate": 1.6801619433198383e-05,
      "loss": 14.5312,
      "mean_token_accuracy": 0.586321234703064,
      "num_tokens": 5349606.0,
      "step": 80
    },
    {
      "epoch": 0.32388663967611336,
      "eval_loss": 1.8290624618530273,
      "eval_mean_token_accuracy": 0.583142414689064,
      "eval_num_tokens": 5349606.0,
      "eval_runtime": 0.6206,
      "eval_samples_per_second": 322.278,
      "eval_steps_per_second": 6.446,
      "step": 80
    },
    {
      "epoch": 0.32793522267206476,
      "grad_norm": 8.385990372203198,
      "learning_rate": 1.676113360323887e-05,
      "loss": 15.8828,
      "mean_token_accuracy": 0.566261887550354,
      "num_tokens": 5413817.0,
      "step": 81
    },
    {
      "epoch": 0.3319838056680162,
      "grad_norm": 8.40823722369304,
      "learning_rate": 1.672064777327935e-05,
      "loss": 15.5781,
      "mean_token_accuracy": 0.5658481121063232,
      "num_tokens": 5487505.0,
      "step": 82
    },
    {
      "epoch": 0.3360323886639676,
      "grad_norm": 8.204736370997345,
      "learning_rate": 1.6680161943319838e-05,
      "loss": 13.2969,
      "mean_token_accuracy": 0.6081260442733765,
      "num_tokens": 5549321.0,
      "step": 83
    },
    {
      "epoch": 0.340080971659919,
      "grad_norm": 7.713082914361621,
      "learning_rate": 1.6639676113360327e-05,
      "loss": 14.375,
      "mean_token_accuracy": 0.5855867266654968,
      "num_tokens": 5612660.0,
      "step": 84
    },
    {
      "epoch": 0.3441295546558704,
      "grad_norm": 9.49322308557619,
      "learning_rate": 1.659919028340081e-05,
      "loss": 15.6953,
      "mean_token_accuracy": 0.5663869380950928,
      "num_tokens": 5685773.0,
      "step": 85
    },
    {
      "epoch": 0.3441295546558704,
      "eval_loss": 1.8246874809265137,
      "eval_mean_token_accuracy": 0.5833836942911148,
      "eval_num_tokens": 5685773.0,
      "eval_runtime": 0.6299,
      "eval_samples_per_second": 317.535,
      "eval_steps_per_second": 6.351,
      "step": 85
    },
    {
      "epoch": 0.3481781376518219,
      "grad_norm": 8.979306474955873,
      "learning_rate": 1.6558704453441296e-05,
      "loss": 14.8125,
      "mean_token_accuracy": 0.5798248648643494,
      "num_tokens": 5746238.0,
      "step": 86
    },
    {
      "epoch": 0.3522267206477733,
      "grad_norm": 7.582176162306731,
      "learning_rate": 1.6518218623481782e-05,
      "loss": 15.5625,
      "mean_token_accuracy": 0.5640476942062378,
      "num_tokens": 5814993.0,
      "step": 87
    },
    {
      "epoch": 0.3562753036437247,
      "grad_norm": 8.163709779919543,
      "learning_rate": 1.6477732793522268e-05,
      "loss": 15.125,
      "mean_token_accuracy": 0.5744800567626953,
      "num_tokens": 5884344.0,
      "step": 88
    },
    {
      "epoch": 0.3603238866396761,
      "grad_norm": 7.597205625950287,
      "learning_rate": 1.6437246963562754e-05,
      "loss": 15.1172,
      "mean_token_accuracy": 0.5687373876571655,
      "num_tokens": 5954319.0,
      "step": 89
    },
    {
      "epoch": 0.3643724696356275,
      "grad_norm": 7.2482847160987625,
      "learning_rate": 1.639676113360324e-05,
      "loss": 13.2422,
      "mean_token_accuracy": 0.6023260354995728,
      "num_tokens": 6019645.0,
      "step": 90
    },
    {
      "epoch": 0.3643724696356275,
      "eval_loss": 1.8174999952316284,
      "eval_mean_token_accuracy": 0.5845800191164017,
      "eval_num_tokens": 6019645.0,
      "eval_runtime": 0.6235,
      "eval_samples_per_second": 320.756,
      "eval_steps_per_second": 6.415,
      "step": 90
    },
    {
      "epoch": 0.3684210526315789,
      "grad_norm": 8.841871793018052,
      "learning_rate": 1.6356275303643723e-05,
      "loss": 13.7266,
      "mean_token_accuracy": 0.6017288565635681,
      "num_tokens": 6075352.0,
      "step": 91
    },
    {
      "epoch": 0.3724696356275304,
      "grad_norm": 8.434804871321719,
      "learning_rate": 1.6315789473684213e-05,
      "loss": 14.0547,
      "mean_token_accuracy": 0.5935065746307373,
      "num_tokens": 6145764.0,
      "step": 92
    },
    {
      "epoch": 0.3765182186234818,
      "grad_norm": 7.730347182665673,
      "learning_rate": 1.62753036437247e-05,
      "loss": 14.6641,
      "mean_token_accuracy": 0.5870702862739563,
      "num_tokens": 6210408.0,
      "step": 93
    },
    {
      "epoch": 0.3805668016194332,
      "grad_norm": 7.45273814710918,
      "learning_rate": 1.6234817813765185e-05,
      "loss": 14.2422,
      "mean_token_accuracy": 0.5905075669288635,
      "num_tokens": 6279074.0,
      "step": 94
    },
    {
      "epoch": 0.38461538461538464,
      "grad_norm": 11.1983859182283,
      "learning_rate": 1.6194331983805668e-05,
      "loss": 15.3906,
      "mean_token_accuracy": 0.5667217373847961,
      "num_tokens": 6348733.0,
      "step": 95
    },
    {
      "epoch": 0.38461538461538464,
      "eval_loss": 1.8106249570846558,
      "eval_mean_token_accuracy": 0.5847904682159424,
      "eval_num_tokens": 6348733.0,
      "eval_runtime": 0.6596,
      "eval_samples_per_second": 303.232,
      "eval_steps_per_second": 6.065,
      "step": 95
    },
    {
      "epoch": 0.38866396761133604,
      "grad_norm": 6.824860263185756,
      "learning_rate": 1.6153846153846154e-05,
      "loss": 13.8672,
      "mean_token_accuracy": 0.5981316566467285,
      "num_tokens": 6418698.0,
      "step": 96
    },
    {
      "epoch": 0.39271255060728744,
      "grad_norm": 7.431961140378814,
      "learning_rate": 1.6113360323886644e-05,
      "loss": 15.1719,
      "mean_token_accuracy": 0.5711833834648132,
      "num_tokens": 6488097.0,
      "step": 97
    },
    {
      "epoch": 0.3967611336032389,
      "grad_norm": 8.677956082707139,
      "learning_rate": 1.6072874493927126e-05,
      "loss": 14.4922,
      "mean_token_accuracy": 0.585637092590332,
      "num_tokens": 6549764.0,
      "step": 98
    },
    {
      "epoch": 0.4008097165991903,
      "grad_norm": 8.646021653592204,
      "learning_rate": 1.6032388663967612e-05,
      "loss": 14.75,
      "mean_token_accuracy": 0.5855600833892822,
      "num_tokens": 6609482.0,
      "step": 99
    },
    {
      "epoch": 0.4048582995951417,
      "grad_norm": 7.907811817137269,
      "learning_rate": 1.59919028340081e-05,
      "loss": 15.0469,
      "mean_token_accuracy": 0.5750304460525513,
      "num_tokens": 6679311.0,
      "step": 100
    },
    {
      "epoch": 0.4048582995951417,
      "eval_loss": 1.807187557220459,
      "eval_mean_token_accuracy": 0.585128903388977,
      "eval_num_tokens": 6679311.0,
      "eval_runtime": 0.6262,
      "eval_samples_per_second": 319.397,
      "eval_steps_per_second": 6.388,
      "step": 100
    },
    {
      "epoch": 0.4089068825910931,
      "grad_norm": 7.835498999161377,
      "learning_rate": 1.5951417004048585e-05,
      "loss": 15.0156,
      "mean_token_accuracy": 0.5781442523002625,
      "num_tokens": 6736916.0,
      "step": 101
    },
    {
      "epoch": 0.41295546558704455,
      "grad_norm": 7.972192619348582,
      "learning_rate": 1.591093117408907e-05,
      "loss": 14.3438,
      "mean_token_accuracy": 0.590062141418457,
      "num_tokens": 6802185.0,
      "step": 102
    },
    {
      "epoch": 0.41700404858299595,
      "grad_norm": 7.983491081447614,
      "learning_rate": 1.5870445344129557e-05,
      "loss": 14.75,
      "mean_token_accuracy": 0.579732358455658,
      "num_tokens": 6863598.0,
      "step": 103
    },
    {
      "epoch": 0.42105263157894735,
      "grad_norm": 10.01408214824878,
      "learning_rate": 1.582995951417004e-05,
      "loss": 14.9219,
      "mean_token_accuracy": 0.5742595791816711,
      "num_tokens": 6927748.0,
      "step": 104
    },
    {
      "epoch": 0.4251012145748988,
      "grad_norm": 7.779179257118591,
      "learning_rate": 1.578947368421053e-05,
      "loss": 13.8125,
      "mean_token_accuracy": 0.5934587121009827,
      "num_tokens": 6993579.0,
      "step": 105
    },
    {
      "epoch": 0.4251012145748988,
      "eval_loss": 1.8018749952316284,
      "eval_mean_token_accuracy": 0.5855304300785065,
      "eval_num_tokens": 6993579.0,
      "eval_runtime": 0.624,
      "eval_samples_per_second": 320.53,
      "eval_steps_per_second": 6.411,
      "step": 105
    },
    {
      "epoch": 0.4291497975708502,
      "grad_norm": 7.766216017833367,
      "learning_rate": 1.5748987854251015e-05,
      "loss": 14.9453,
      "mean_token_accuracy": 0.5719259977340698,
      "num_tokens": 7064007.0,
      "step": 106
    },
    {
      "epoch": 0.4331983805668016,
      "grad_norm": 8.87590967566746,
      "learning_rate": 1.5708502024291498e-05,
      "loss": 14.1328,
      "mean_token_accuracy": 0.5912887454032898,
      "num_tokens": 7126932.0,
      "step": 107
    },
    {
      "epoch": 0.43724696356275305,
      "grad_norm": 7.2824506866551015,
      "learning_rate": 1.5668016194331984e-05,
      "loss": 12.9688,
      "mean_token_accuracy": 0.6094680428504944,
      "num_tokens": 7188297.0,
      "step": 108
    },
    {
      "epoch": 0.44129554655870445,
      "grad_norm": 8.281692416649847,
      "learning_rate": 1.562753036437247e-05,
      "loss": 14.7578,
      "mean_token_accuracy": 0.5806804895401001,
      "num_tokens": 7252286.0,
      "step": 109
    },
    {
      "epoch": 0.44534412955465585,
      "grad_norm": 7.710589034957108,
      "learning_rate": 1.5587044534412957e-05,
      "loss": 13.9375,
      "mean_token_accuracy": 0.5891672968864441,
      "num_tokens": 7328970.0,
      "step": 110
    },
    {
      "epoch": 0.44534412955465585,
      "eval_loss": 1.7962499856948853,
      "eval_mean_token_accuracy": 0.5861413329839706,
      "eval_num_tokens": 7328970.0,
      "eval_runtime": 0.6541,
      "eval_samples_per_second": 305.776,
      "eval_steps_per_second": 6.116,
      "step": 110
    },
    {
      "epoch": 0.4493927125506073,
      "grad_norm": 9.932234036882006,
      "learning_rate": 1.5546558704453443e-05,
      "loss": 14.9844,
      "mean_token_accuracy": 0.583306074142456,
      "num_tokens": 7396220.0,
      "step": 111
    },
    {
      "epoch": 0.4534412955465587,
      "grad_norm": 7.818592875809941,
      "learning_rate": 1.550607287449393e-05,
      "loss": 14.6562,
      "mean_token_accuracy": 0.5830007195472717,
      "num_tokens": 7454663.0,
      "step": 112
    },
    {
      "epoch": 0.4574898785425101,
      "grad_norm": 8.077045865068051,
      "learning_rate": 1.5465587044534415e-05,
      "loss": 15.1641,
      "mean_token_accuracy": 0.5737788677215576,
      "num_tokens": 7515205.0,
      "step": 113
    },
    {
      "epoch": 0.46153846153846156,
      "grad_norm": 8.172258229119766,
      "learning_rate": 1.54251012145749e-05,
      "loss": 13.7969,
      "mean_token_accuracy": 0.5884114503860474,
      "num_tokens": 7571704.0,
      "step": 114
    },
    {
      "epoch": 0.46558704453441296,
      "grad_norm": 7.57672126802984,
      "learning_rate": 1.5384615384615387e-05,
      "loss": 13.6719,
      "mean_token_accuracy": 0.599338173866272,
      "num_tokens": 7634020.0,
      "step": 115
    },
    {
      "epoch": 0.46558704453441296,
      "eval_loss": 1.7934374809265137,
      "eval_mean_token_accuracy": 0.5860239416360855,
      "eval_num_tokens": 7634020.0,
      "eval_runtime": 0.6265,
      "eval_samples_per_second": 319.23,
      "eval_steps_per_second": 6.385,
      "step": 115
    },
    {
      "epoch": 0.46963562753036436,
      "grad_norm": 7.229443663353096,
      "learning_rate": 1.534412955465587e-05,
      "loss": 13.4922,
      "mean_token_accuracy": 0.5976777672767639,
      "num_tokens": 7702897.0,
      "step": 116
    },
    {
      "epoch": 0.47368421052631576,
      "grad_norm": 8.594596623758976,
      "learning_rate": 1.5303643724696356e-05,
      "loss": 13.7969,
      "mean_token_accuracy": 0.595205545425415,
      "num_tokens": 7761445.0,
      "step": 117
    },
    {
      "epoch": 0.4777327935222672,
      "grad_norm": 7.44049680844727,
      "learning_rate": 1.5263157894736846e-05,
      "loss": 14.4688,
      "mean_token_accuracy": 0.5831844210624695,
      "num_tokens": 7829220.0,
      "step": 118
    },
    {
      "epoch": 0.4817813765182186,
      "grad_norm": 8.010264997343702,
      "learning_rate": 1.5222672064777328e-05,
      "loss": 14.3516,
      "mean_token_accuracy": 0.5925936698913574,
      "num_tokens": 7897361.0,
      "step": 119
    },
    {
      "epoch": 0.48582995951417,
      "grad_norm": 9.11202877002977,
      "learning_rate": 1.5182186234817814e-05,
      "loss": 15.5078,
      "mean_token_accuracy": 0.5689979195594788,
      "num_tokens": 7958666.0,
      "step": 120
    },
    {
      "epoch": 0.48582995951417,
      "eval_loss": 1.786562442779541,
      "eval_mean_token_accuracy": 0.587049126625061,
      "eval_num_tokens": 7958666.0,
      "eval_runtime": 0.6451,
      "eval_samples_per_second": 310.009,
      "eval_steps_per_second": 6.2,
      "step": 120
    },
    {
      "epoch": 0.4898785425101215,
      "grad_norm": 8.733802509491014,
      "learning_rate": 1.5141700404858302e-05,
      "loss": 13.9922,
      "mean_token_accuracy": 0.591108500957489,
      "num_tokens": 8031901.0,
      "step": 121
    },
    {
      "epoch": 0.4939271255060729,
      "grad_norm": 7.43499030469072,
      "learning_rate": 1.5101214574898787e-05,
      "loss": 13.6719,
      "mean_token_accuracy": 0.6017109155654907,
      "num_tokens": 8105143.0,
      "step": 122
    },
    {
      "epoch": 0.4979757085020243,
      "grad_norm": 7.117038860943061,
      "learning_rate": 1.5060728744939273e-05,
      "loss": 14.5234,
      "mean_token_accuracy": 0.580233633518219,
      "num_tokens": 8177876.0,
      "step": 123
    },
    {
      "epoch": 0.5020242914979757,
      "grad_norm": 7.746503576535467,
      "learning_rate": 1.5020242914979759e-05,
      "loss": 14.6797,
      "mean_token_accuracy": 0.5798138380050659,
      "num_tokens": 8246274.0,
      "step": 124
    },
    {
      "epoch": 0.5060728744939271,
      "grad_norm": 7.532018226813427,
      "learning_rate": 1.4979757085020243e-05,
      "loss": 14.1562,
      "mean_token_accuracy": 0.582686722278595,
      "num_tokens": 8315418.0,
      "step": 125
    },
    {
      "epoch": 0.5060728744939271,
      "eval_loss": 1.7821874618530273,
      "eval_mean_token_accuracy": 0.5875359177589417,
      "eval_num_tokens": 8315418.0,
      "eval_runtime": 0.6323,
      "eval_samples_per_second": 316.295,
      "eval_steps_per_second": 6.326,
      "step": 125
    },
    {
      "epoch": 0.5101214574898786,
      "grad_norm": 8.864364235981503,
      "learning_rate": 1.493927125506073e-05,
      "loss": 14.4688,
      "mean_token_accuracy": 0.5993614792823792,
      "num_tokens": 8372174.0,
      "step": 126
    },
    {
      "epoch": 0.5141700404858299,
      "grad_norm": 8.95321820232742,
      "learning_rate": 1.4898785425101216e-05,
      "loss": 14.3438,
      "mean_token_accuracy": 0.5901056528091431,
      "num_tokens": 8441146.0,
      "step": 127
    },
    {
      "epoch": 0.5182186234817814,
      "grad_norm": 7.263835476936067,
      "learning_rate": 1.48582995951417e-05,
      "loss": 13.9688,
      "mean_token_accuracy": 0.594972550868988,
      "num_tokens": 8511862.0,
      "step": 128
    },
    {
      "epoch": 0.5222672064777328,
      "grad_norm": 9.946116238413886,
      "learning_rate": 1.4817813765182188e-05,
      "loss": 14.6953,
      "mean_token_accuracy": 0.5761271119117737,
      "num_tokens": 8579819.0,
      "step": 129
    },
    {
      "epoch": 0.5263157894736842,
      "grad_norm": 7.617051573886364,
      "learning_rate": 1.4777327935222674e-05,
      "loss": 12.9141,
      "mean_token_accuracy": 0.6136053800582886,
      "num_tokens": 8648885.0,
      "step": 130
    },
    {
      "epoch": 0.5263157894736842,
      "eval_loss": 1.7781250476837158,
      "eval_mean_token_accuracy": 0.5874989777803421,
      "eval_num_tokens": 8648885.0,
      "eval_runtime": 0.6456,
      "eval_samples_per_second": 309.786,
      "eval_steps_per_second": 6.196,
      "step": 130
    },
    {
      "epoch": 0.5303643724696356,
      "grad_norm": 7.995063941880695,
      "learning_rate": 1.4736842105263159e-05,
      "loss": 14.2891,
      "mean_token_accuracy": 0.5915219187736511,
      "num_tokens": 8716889.0,
      "step": 131
    },
    {
      "epoch": 0.5344129554655871,
      "grad_norm": 7.9292860672813665,
      "learning_rate": 1.4696356275303645e-05,
      "loss": 13.5078,
      "mean_token_accuracy": 0.6026769280433655,
      "num_tokens": 8777395.0,
      "step": 132
    },
    {
      "epoch": 0.5384615384615384,
      "grad_norm": 7.586666671628873,
      "learning_rate": 1.465587044534413e-05,
      "loss": 14.8672,
      "mean_token_accuracy": 0.5739372372627258,
      "num_tokens": 8848242.0,
      "step": 133
    },
    {
      "epoch": 0.5425101214574899,
      "grad_norm": 8.426747584830805,
      "learning_rate": 1.4615384615384615e-05,
      "loss": 14.6484,
      "mean_token_accuracy": 0.5826023817062378,
      "num_tokens": 8910211.0,
      "step": 134
    },
    {
      "epoch": 0.5465587044534413,
      "grad_norm": 8.34819038901423,
      "learning_rate": 1.4574898785425103e-05,
      "loss": 15.3672,
      "mean_token_accuracy": 0.5670615434646606,
      "num_tokens": 8976714.0,
      "step": 135
    },
    {
      "epoch": 0.5465587044534413,
      "eval_loss": 1.770937442779541,
      "eval_mean_token_accuracy": 0.5890246480703354,
      "eval_num_tokens": 8976714.0,
      "eval_runtime": 0.627,
      "eval_samples_per_second": 318.954,
      "eval_steps_per_second": 6.379,
      "step": 135
    },
    {
      "epoch": 0.5506072874493927,
      "grad_norm": 7.635072916228879,
      "learning_rate": 1.453441295546559e-05,
      "loss": 13.5703,
      "mean_token_accuracy": 0.5985166430473328,
      "num_tokens": 9032733.0,
      "step": 136
    },
    {
      "epoch": 0.5546558704453441,
      "grad_norm": 7.13210528504054,
      "learning_rate": 1.4493927125506074e-05,
      "loss": 14.0469,
      "mean_token_accuracy": 0.5941102504730225,
      "num_tokens": 9106960.0,
      "step": 137
    },
    {
      "epoch": 0.5587044534412956,
      "grad_norm": 7.472585203507861,
      "learning_rate": 1.445344129554656e-05,
      "loss": 13.4141,
      "mean_token_accuracy": 0.6021690964698792,
      "num_tokens": 9171843.0,
      "step": 138
    },
    {
      "epoch": 0.562753036437247,
      "grad_norm": 8.097767778322229,
      "learning_rate": 1.4412955465587046e-05,
      "loss": 14.8047,
      "mean_token_accuracy": 0.578671932220459,
      "num_tokens": 9238411.0,
      "step": 139
    },
    {
      "epoch": 0.5668016194331984,
      "grad_norm": 8.339563519464773,
      "learning_rate": 1.437246963562753e-05,
      "loss": 14.6484,
      "mean_token_accuracy": 0.5821104645729065,
      "num_tokens": 9300623.0,
      "step": 140
    },
    {
      "epoch": 0.5668016194331984,
      "eval_loss": 1.7675000429153442,
      "eval_mean_token_accuracy": 0.5898215025663376,
      "eval_num_tokens": 9300623.0,
      "eval_runtime": 0.6311,
      "eval_samples_per_second": 316.883,
      "eval_steps_per_second": 6.338,
      "step": 140
    },
    {
      "epoch": 0.5708502024291497,
      "grad_norm": 7.211581007077111,
      "learning_rate": 1.4331983805668017e-05,
      "loss": 14.1094,
      "mean_token_accuracy": 0.5875740051269531,
      "num_tokens": 9364199.0,
      "step": 141
    },
    {
      "epoch": 0.5748987854251012,
      "grad_norm": 8.547180559454508,
      "learning_rate": 1.4291497975708504e-05,
      "loss": 13.3594,
      "mean_token_accuracy": 0.6009641885757446,
      "num_tokens": 9427112.0,
      "step": 142
    },
    {
      "epoch": 0.5789473684210527,
      "grad_norm": 8.314592250089833,
      "learning_rate": 1.4251012145748989e-05,
      "loss": 13.6719,
      "mean_token_accuracy": 0.5959135890007019,
      "num_tokens": 9493444.0,
      "step": 143
    },
    {
      "epoch": 0.582995951417004,
      "grad_norm": 8.353437279154521,
      "learning_rate": 1.4210526315789475e-05,
      "loss": 14.4453,
      "mean_token_accuracy": 0.5817537307739258,
      "num_tokens": 9563780.0,
      "step": 144
    },
    {
      "epoch": 0.5870445344129555,
      "grad_norm": 7.105730625315736,
      "learning_rate": 1.4170040485829961e-05,
      "loss": 13.8828,
      "mean_token_accuracy": 0.5948812961578369,
      "num_tokens": 9634643.0,
      "step": 145
    },
    {
      "epoch": 0.5870445344129555,
      "eval_loss": 1.764062523841858,
      "eval_mean_token_accuracy": 0.589348778128624,
      "eval_num_tokens": 9634643.0,
      "eval_runtime": 0.6326,
      "eval_samples_per_second": 316.156,
      "eval_steps_per_second": 6.323,
      "step": 145
    },
    {
      "epoch": 0.5910931174089069,
      "grad_norm": 8.85529190632474,
      "learning_rate": 1.4129554655870446e-05,
      "loss": 14.9609,
      "mean_token_accuracy": 0.5746488571166992,
      "num_tokens": 9705552.0,
      "step": 146
    },
    {
      "epoch": 0.5951417004048583,
      "grad_norm": 7.404634749249628,
      "learning_rate": 1.4089068825910932e-05,
      "loss": 14.1562,
      "mean_token_accuracy": 0.5874021053314209,
      "num_tokens": 9784121.0,
      "step": 147
    },
    {
      "epoch": 0.5991902834008097,
      "grad_norm": 8.414339407163222,
      "learning_rate": 1.4048582995951418e-05,
      "loss": 14.2266,
      "mean_token_accuracy": 0.5838397741317749,
      "num_tokens": 9855434.0,
      "step": 148
    },
    {
      "epoch": 0.6032388663967612,
      "grad_norm": 8.385715636253362,
      "learning_rate": 1.4008097165991902e-05,
      "loss": 14.9531,
      "mean_token_accuracy": 0.580719530582428,
      "num_tokens": 9932270.0,
      "step": 149
    },
    {
      "epoch": 0.6072874493927125,
      "grad_norm": 8.437220430570475,
      "learning_rate": 1.396761133603239e-05,
      "loss": 13.5312,
      "mean_token_accuracy": 0.6000396013259888,
      "num_tokens": 9987882.0,
      "step": 150
    },
    {
      "epoch": 0.6072874493927125,
      "eval_loss": 1.7596875429153442,
      "eval_mean_token_accuracy": 0.592434361577034,
      "eval_num_tokens": 9987882.0,
      "eval_runtime": 0.6411,
      "eval_samples_per_second": 311.96,
      "eval_steps_per_second": 6.239,
      "step": 150
    },
    {
      "epoch": 0.611336032388664,
      "grad_norm": 7.3392719089419,
      "learning_rate": 1.3927125506072876e-05,
      "loss": 14.1953,
      "mean_token_accuracy": 0.5917255878448486,
      "num_tokens": 10053860.0,
      "step": 151
    },
    {
      "epoch": 0.6153846153846154,
      "grad_norm": 7.106128585719554,
      "learning_rate": 1.388663967611336e-05,
      "loss": 13.9062,
      "mean_token_accuracy": 0.5945723652839661,
      "num_tokens": 10118446.0,
      "step": 152
    },
    {
      "epoch": 0.6194331983805668,
      "grad_norm": 8.771091973478418,
      "learning_rate": 1.3846153846153847e-05,
      "loss": 13.6953,
      "mean_token_accuracy": 0.5965032577514648,
      "num_tokens": 10185658.0,
      "step": 153
    },
    {
      "epoch": 0.6234817813765182,
      "grad_norm": 7.463408517295952,
      "learning_rate": 1.3805668016194333e-05,
      "loss": 14.5234,
      "mean_token_accuracy": 0.5856794118881226,
      "num_tokens": 10256194.0,
      "step": 154
    },
    {
      "epoch": 0.6275303643724697,
      "grad_norm": 7.766925275194455,
      "learning_rate": 1.3765182186234817e-05,
      "loss": 14.4844,
      "mean_token_accuracy": 0.5821614861488342,
      "num_tokens": 10322676.0,
      "step": 155
    },
    {
      "epoch": 0.6275303643724697,
      "eval_loss": 1.7568750381469727,
      "eval_mean_token_accuracy": 0.593303382396698,
      "eval_num_tokens": 10322676.0,
      "eval_runtime": 0.6285,
      "eval_samples_per_second": 318.24,
      "eval_steps_per_second": 6.365,
      "step": 155
    },
    {
      "epoch": 0.631578947368421,
      "grad_norm": 11.730558732427344,
      "learning_rate": 1.3724696356275305e-05,
      "loss": 14.1797,
      "mean_token_accuracy": 0.590240478515625,
      "num_tokens": 10390182.0,
      "step": 156
    },
    {
      "epoch": 0.6356275303643725,
      "grad_norm": 6.962088111992022,
      "learning_rate": 1.3684210526315791e-05,
      "loss": 14.6328,
      "mean_token_accuracy": 0.5875160694122314,
      "num_tokens": 10482669.0,
      "step": 157
    },
    {
      "epoch": 0.6396761133603239,
      "grad_norm": 7.342884429279778,
      "learning_rate": 1.3643724696356277e-05,
      "loss": 13.4531,
      "mean_token_accuracy": 0.6036462783813477,
      "num_tokens": 10543343.0,
      "step": 158
    },
    {
      "epoch": 0.6437246963562753,
      "grad_norm": 9.243698402724284,
      "learning_rate": 1.3603238866396762e-05,
      "loss": 14.3438,
      "mean_token_accuracy": 0.5891345739364624,
      "num_tokens": 10605678.0,
      "step": 159
    },
    {
      "epoch": 0.6477732793522267,
      "grad_norm": 6.586652944449273,
      "learning_rate": 1.3562753036437248e-05,
      "loss": 13.6094,
      "mean_token_accuracy": 0.6069329977035522,
      "num_tokens": 10678928.0,
      "step": 160
    },
    {
      "epoch": 0.6477732793522267,
      "eval_loss": 1.751562476158142,
      "eval_mean_token_accuracy": 0.5941518694162369,
      "eval_num_tokens": 10678928.0,
      "eval_runtime": 0.6304,
      "eval_samples_per_second": 317.246,
      "eval_steps_per_second": 6.345,
      "step": 160
    },
    {
      "epoch": 0.6518218623481782,
      "grad_norm": 7.292010909500028,
      "learning_rate": 1.3522267206477734e-05,
      "loss": 15.5391,
      "mean_token_accuracy": 0.5630365014076233,
      "num_tokens": 10752497.0,
      "step": 161
    },
    {
      "epoch": 0.6558704453441295,
      "grad_norm": 8.517901596352386,
      "learning_rate": 1.3481781376518219e-05,
      "loss": 13.8203,
      "mean_token_accuracy": 0.606192409992218,
      "num_tokens": 10808888.0,
      "step": 162
    },
    {
      "epoch": 0.659919028340081,
      "grad_norm": 7.047539986657361,
      "learning_rate": 1.3441295546558706e-05,
      "loss": 12.6094,
      "mean_token_accuracy": 0.6234721541404724,
      "num_tokens": 10867941.0,
      "step": 163
    },
    {
      "epoch": 0.6639676113360324,
      "grad_norm": 8.193594708311382,
      "learning_rate": 1.3400809716599193e-05,
      "loss": 15.1562,
      "mean_token_accuracy": 0.5752149224281311,
      "num_tokens": 10945011.0,
      "step": 164
    },
    {
      "epoch": 0.6680161943319838,
      "grad_norm": 7.0437604094189386,
      "learning_rate": 1.3360323886639677e-05,
      "loss": 13.1094,
      "mean_token_accuracy": 0.6154734492301941,
      "num_tokens": 11020417.0,
      "step": 165
    },
    {
      "epoch": 0.6680161943319838,
      "eval_loss": 1.7496875524520874,
      "eval_mean_token_accuracy": 0.5943519324064255,
      "eval_num_tokens": 11020417.0,
      "eval_runtime": 0.6499,
      "eval_samples_per_second": 307.758,
      "eval_steps_per_second": 6.155,
      "step": 165
    },
    {
      "epoch": 0.6720647773279352,
      "grad_norm": 7.138901056236762,
      "learning_rate": 1.3319838056680163e-05,
      "loss": 13.8594,
      "mean_token_accuracy": 0.5956168174743652,
      "num_tokens": 11089883.0,
      "step": 166
    },
    {
      "epoch": 0.6761133603238867,
      "grad_norm": 8.201709305900888,
      "learning_rate": 1.327935222672065e-05,
      "loss": 13.7344,
      "mean_token_accuracy": 0.5976811051368713,
      "num_tokens": 11147819.0,
      "step": 167
    },
    {
      "epoch": 0.680161943319838,
      "grad_norm": 7.8526020934287954,
      "learning_rate": 1.3238866396761134e-05,
      "loss": 13.9766,
      "mean_token_accuracy": 0.5927841067314148,
      "num_tokens": 11219059.0,
      "step": 168
    },
    {
      "epoch": 0.6842105263157895,
      "grad_norm": 7.318409427810135,
      "learning_rate": 1.3198380566801622e-05,
      "loss": 14.3281,
      "mean_token_accuracy": 0.5857775211334229,
      "num_tokens": 11278072.0,
      "step": 169
    },
    {
      "epoch": 0.6882591093117408,
      "grad_norm": 7.389155589413203,
      "learning_rate": 1.3157894736842108e-05,
      "loss": 14.4297,
      "mean_token_accuracy": 0.587368905544281,
      "num_tokens": 11351732.0,
      "step": 170
    },
    {
      "epoch": 0.6882591093117408,
      "eval_loss": 1.746250033378601,
      "eval_mean_token_accuracy": 0.5935623943805695,
      "eval_num_tokens": 11351732.0,
      "eval_runtime": 0.6314,
      "eval_samples_per_second": 316.758,
      "eval_steps_per_second": 6.335,
      "step": 170
    },
    {
      "epoch": 0.6923076923076923,
      "grad_norm": 8.124816197332432,
      "learning_rate": 1.3117408906882592e-05,
      "loss": 14.1172,
      "mean_token_accuracy": 0.5922242999076843,
      "num_tokens": 11407714.0,
      "step": 171
    },
    {
      "epoch": 0.6963562753036437,
      "grad_norm": 7.965720782927913,
      "learning_rate": 1.3076923076923078e-05,
      "loss": 13.4609,
      "mean_token_accuracy": 0.6015889048576355,
      "num_tokens": 11484184.0,
      "step": 172
    },
    {
      "epoch": 0.7004048582995951,
      "grad_norm": 8.915221369950393,
      "learning_rate": 1.3036437246963564e-05,
      "loss": 14.1875,
      "mean_token_accuracy": 0.5905665755271912,
      "num_tokens": 11544099.0,
      "step": 173
    },
    {
      "epoch": 0.7044534412955465,
      "grad_norm": 8.790247334783299,
      "learning_rate": 1.2995951417004049e-05,
      "loss": 13.9219,
      "mean_token_accuracy": 0.5985158681869507,
      "num_tokens": 11606691.0,
      "step": 174
    },
    {
      "epoch": 0.708502024291498,
      "grad_norm": 9.868245626915737,
      "learning_rate": 1.2955465587044535e-05,
      "loss": 13.6172,
      "mean_token_accuracy": 0.5943373441696167,
      "num_tokens": 11672625.0,
      "step": 175
    },
    {
      "epoch": 0.708502024291498,
      "eval_loss": 1.7431249618530273,
      "eval_mean_token_accuracy": 0.5946680456399918,
      "eval_num_tokens": 11672625.0,
      "eval_runtime": 0.6432,
      "eval_samples_per_second": 310.949,
      "eval_steps_per_second": 6.219,
      "step": 175
    },
    {
      "epoch": 0.7125506072874493,
      "grad_norm": 8.884030860368474,
      "learning_rate": 1.2914979757085023e-05,
      "loss": 14.0,
      "mean_token_accuracy": 0.5949345827102661,
      "num_tokens": 11732625.0,
      "step": 176
    },
    {
      "epoch": 0.7165991902834008,
      "grad_norm": 9.007310004782914,
      "learning_rate": 1.2874493927125507e-05,
      "loss": 14.0156,
      "mean_token_accuracy": 0.6007354259490967,
      "num_tokens": 11801766.0,
      "step": 177
    },
    {
      "epoch": 0.7206477732793523,
      "grad_norm": 8.33208496329059,
      "learning_rate": 1.2834008097165993e-05,
      "loss": 14.8281,
      "mean_token_accuracy": 0.5790393352508545,
      "num_tokens": 11863641.0,
      "step": 178
    },
    {
      "epoch": 0.7246963562753036,
      "grad_norm": 7.426550244631567,
      "learning_rate": 1.279352226720648e-05,
      "loss": 14.5625,
      "mean_token_accuracy": 0.5841023325920105,
      "num_tokens": 11933275.0,
      "step": 179
    },
    {
      "epoch": 0.728744939271255,
      "grad_norm": 6.8034373170968525,
      "learning_rate": 1.2753036437246964e-05,
      "loss": 14.6562,
      "mean_token_accuracy": 0.5787296295166016,
      "num_tokens": 12017691.0,
      "step": 180
    },
    {
      "epoch": 0.728744939271255,
      "eval_loss": 1.7400000095367432,
      "eval_mean_token_accuracy": 0.5955071151256561,
      "eval_num_tokens": 12017691.0,
      "eval_runtime": 0.6493,
      "eval_samples_per_second": 308.043,
      "eval_steps_per_second": 6.161,
      "step": 180
    },
    {
      "epoch": 0.7327935222672065,
      "grad_norm": 7.057663809648343,
      "learning_rate": 1.271255060728745e-05,
      "loss": 13.6875,
      "mean_token_accuracy": 0.5996446013450623,
      "num_tokens": 12079093.0,
      "step": 181
    },
    {
      "epoch": 0.7368421052631579,
      "grad_norm": 7.371289124958891,
      "learning_rate": 1.2672064777327936e-05,
      "loss": 14.5234,
      "mean_token_accuracy": 0.5833817720413208,
      "num_tokens": 12160008.0,
      "step": 182
    },
    {
      "epoch": 0.7408906882591093,
      "grad_norm": 8.707687219264766,
      "learning_rate": 1.263157894736842e-05,
      "loss": 14.2266,
      "mean_token_accuracy": 0.5921141505241394,
      "num_tokens": 12229741.0,
      "step": 183
    },
    {
      "epoch": 0.7449392712550608,
      "grad_norm": 7.516841864927085,
      "learning_rate": 1.2591093117408908e-05,
      "loss": 14.7891,
      "mean_token_accuracy": 0.5818018913269043,
      "num_tokens": 12301362.0,
      "step": 184
    },
    {
      "epoch": 0.7489878542510121,
      "grad_norm": 7.6633569797452346,
      "learning_rate": 1.2550607287449395e-05,
      "loss": 13.6406,
      "mean_token_accuracy": 0.6013834476470947,
      "num_tokens": 12370239.0,
      "step": 185
    },
    {
      "epoch": 0.7489878542510121,
      "eval_loss": 1.7356250286102295,
      "eval_mean_token_accuracy": 0.5960101634263992,
      "eval_num_tokens": 12370239.0,
      "eval_runtime": 0.6504,
      "eval_samples_per_second": 307.492,
      "eval_steps_per_second": 6.15,
      "step": 185
    },
    {
      "epoch": 0.7530364372469636,
      "grad_norm": 6.968835438956393,
      "learning_rate": 1.2510121457489879e-05,
      "loss": 13.5625,
      "mean_token_accuracy": 0.6022695302963257,
      "num_tokens": 12441507.0,
      "step": 186
    },
    {
      "epoch": 0.757085020242915,
      "grad_norm": 7.195791337914372,
      "learning_rate": 1.2469635627530365e-05,
      "loss": 13.4609,
      "mean_token_accuracy": 0.6029326915740967,
      "num_tokens": 12502266.0,
      "step": 187
    },
    {
      "epoch": 0.7611336032388664,
      "grad_norm": 7.668677247147722,
      "learning_rate": 1.2429149797570851e-05,
      "loss": 13.8047,
      "mean_token_accuracy": 0.5992096662521362,
      "num_tokens": 12566099.0,
      "step": 188
    },
    {
      "epoch": 0.7651821862348178,
      "grad_norm": 7.04566661191058,
      "learning_rate": 1.2388663967611336e-05,
      "loss": 13.6875,
      "mean_token_accuracy": 0.6022654175758362,
      "num_tokens": 12634407.0,
      "step": 189
    },
    {
      "epoch": 0.7692307692307693,
      "grad_norm": 7.697888634722549,
      "learning_rate": 1.2348178137651824e-05,
      "loss": 13.5703,
      "mean_token_accuracy": 0.6025985479354858,
      "num_tokens": 12695120.0,
      "step": 190
    },
    {
      "epoch": 0.7692307692307693,
      "eval_loss": 1.734375,
      "eval_mean_token_accuracy": 0.5960522890090942,
      "eval_num_tokens": 12695120.0,
      "eval_runtime": 0.6494,
      "eval_samples_per_second": 307.96,
      "eval_steps_per_second": 6.159,
      "step": 190
    },
    {
      "epoch": 0.7732793522267206,
      "grad_norm": 6.785752803586286,
      "learning_rate": 1.230769230769231e-05,
      "loss": 13.2031,
      "mean_token_accuracy": 0.6083678603172302,
      "num_tokens": 12765477.0,
      "step": 191
    },
    {
      "epoch": 0.7773279352226721,
      "grad_norm": 8.429385939587196,
      "learning_rate": 1.2267206477732794e-05,
      "loss": 13.6719,
      "mean_token_accuracy": 0.5998564958572388,
      "num_tokens": 12835216.0,
      "step": 192
    },
    {
      "epoch": 0.7813765182186235,
      "grad_norm": 7.85611095251478,
      "learning_rate": 1.222672064777328e-05,
      "loss": 13.1328,
      "mean_token_accuracy": 0.6025688648223877,
      "num_tokens": 12902937.0,
      "step": 193
    },
    {
      "epoch": 0.7854251012145749,
      "grad_norm": 7.6755487929843556,
      "learning_rate": 1.2186234817813766e-05,
      "loss": 13.0703,
      "mean_token_accuracy": 0.6151934266090393,
      "num_tokens": 12958920.0,
      "step": 194
    },
    {
      "epoch": 0.7894736842105263,
      "grad_norm": 9.563095955991269,
      "learning_rate": 1.2145748987854251e-05,
      "loss": 13.75,
      "mean_token_accuracy": 0.5952048301696777,
      "num_tokens": 13019128.0,
      "step": 195
    },
    {
      "epoch": 0.7894736842105263,
      "eval_loss": 1.730625033378601,
      "eval_mean_token_accuracy": 0.5964938104152679,
      "eval_num_tokens": 13019128.0,
      "eval_runtime": 0.6302,
      "eval_samples_per_second": 317.371,
      "eval_steps_per_second": 6.347,
      "step": 195
    },
    {
      "epoch": 0.7935222672064778,
      "grad_norm": 7.513918136227913,
      "learning_rate": 1.2105263157894737e-05,
      "loss": 14.5703,
      "mean_token_accuracy": 0.5844899415969849,
      "num_tokens": 13085188.0,
      "step": 196
    },
    {
      "epoch": 0.7975708502024291,
      "grad_norm": 7.287761719869118,
      "learning_rate": 1.2064777327935225e-05,
      "loss": 13.6484,
      "mean_token_accuracy": 0.5984832048416138,
      "num_tokens": 13155269.0,
      "step": 197
    },
    {
      "epoch": 0.8016194331983806,
      "grad_norm": 7.383829587566862,
      "learning_rate": 1.202429149797571e-05,
      "loss": 14.5156,
      "mean_token_accuracy": 0.5839080214500427,
      "num_tokens": 13229283.0,
      "step": 198
    },
    {
      "epoch": 0.805668016194332,
      "grad_norm": 12.228642188104056,
      "learning_rate": 1.1983805668016195e-05,
      "loss": 14.0234,
      "mean_token_accuracy": 0.5865586400032043,
      "num_tokens": 13298358.0,
      "step": 199
    },
    {
      "epoch": 0.8097165991902834,
      "grad_norm": 8.01540634329265,
      "learning_rate": 1.1943319838056682e-05,
      "loss": 13.4844,
      "mean_token_accuracy": 0.603053867816925,
      "num_tokens": 13366925.0,
      "step": 200
    },
    {
      "epoch": 0.8097165991902834,
      "eval_loss": 1.7278125286102295,
      "eval_mean_token_accuracy": 0.5974646657705307,
      "eval_num_tokens": 13366925.0,
      "eval_runtime": 0.6458,
      "eval_samples_per_second": 309.671,
      "eval_steps_per_second": 6.193,
      "step": 200
    },
    {
      "epoch": 0.8137651821862348,
      "grad_norm": 8.670916760818496,
      "learning_rate": 1.1902834008097166e-05,
      "loss": 13.5781,
      "mean_token_accuracy": 0.5980917811393738,
      "num_tokens": 13423691.0,
      "step": 201
    },
    {
      "epoch": 0.8178137651821862,
      "grad_norm": 8.271091772585052,
      "learning_rate": 1.1862348178137652e-05,
      "loss": 13.5938,
      "mean_token_accuracy": 0.6054921746253967,
      "num_tokens": 13484423.0,
      "step": 202
    },
    {
      "epoch": 0.8218623481781376,
      "grad_norm": 7.300340294135019,
      "learning_rate": 1.182186234817814e-05,
      "loss": 13.1016,
      "mean_token_accuracy": 0.60859215259552,
      "num_tokens": 13559695.0,
      "step": 203
    },
    {
      "epoch": 0.8259109311740891,
      "grad_norm": 9.1407332883229,
      "learning_rate": 1.1781376518218623e-05,
      "loss": 14.1406,
      "mean_token_accuracy": 0.5899603366851807,
      "num_tokens": 13645280.0,
      "step": 204
    },
    {
      "epoch": 0.8299595141700404,
      "grad_norm": 9.22684063564975,
      "learning_rate": 1.174089068825911e-05,
      "loss": 13.6562,
      "mean_token_accuracy": 0.5998020768165588,
      "num_tokens": 13712046.0,
      "step": 205
    },
    {
      "epoch": 0.8299595141700404,
      "eval_loss": 1.7253124713897705,
      "eval_mean_token_accuracy": 0.5980972796678543,
      "eval_num_tokens": 13712046.0,
      "eval_runtime": 0.6282,
      "eval_samples_per_second": 318.347,
      "eval_steps_per_second": 6.367,
      "step": 205
    },
    {
      "epoch": 0.8340080971659919,
      "grad_norm": 7.12821183237729,
      "learning_rate": 1.1700404858299597e-05,
      "loss": 14.1094,
      "mean_token_accuracy": 0.591342568397522,
      "num_tokens": 13776702.0,
      "step": 206
    },
    {
      "epoch": 0.8380566801619433,
      "grad_norm": 8.729950014794346,
      "learning_rate": 1.1659919028340081e-05,
      "loss": 14.2266,
      "mean_token_accuracy": 0.5930887460708618,
      "num_tokens": 13836408.0,
      "step": 207
    },
    {
      "epoch": 0.8421052631578947,
      "grad_norm": 7.301044795096497,
      "learning_rate": 1.1619433198380567e-05,
      "loss": 14.4922,
      "mean_token_accuracy": 0.5839249491691589,
      "num_tokens": 13900559.0,
      "step": 208
    },
    {
      "epoch": 0.8461538461538461,
      "grad_norm": 7.369160590617854,
      "learning_rate": 1.1578947368421053e-05,
      "loss": 13.5703,
      "mean_token_accuracy": 0.6064897775650024,
      "num_tokens": 13970733.0,
      "step": 209
    },
    {
      "epoch": 0.8502024291497976,
      "grad_norm": 7.851482544759256,
      "learning_rate": 1.1538461538461538e-05,
      "loss": 13.1406,
      "mean_token_accuracy": 0.6058458089828491,
      "num_tokens": 14039496.0,
      "step": 210
    },
    {
      "epoch": 0.8502024291497976,
      "eval_loss": 1.7237499952316284,
      "eval_mean_token_accuracy": 0.5978213995695114,
      "eval_num_tokens": 14039496.0,
      "eval_runtime": 0.6629,
      "eval_samples_per_second": 301.705,
      "eval_steps_per_second": 6.034,
      "step": 210
    },
    {
      "epoch": 0.854251012145749,
      "grad_norm": 7.716996728334529,
      "learning_rate": 1.1497975708502026e-05,
      "loss": 13.8281,
      "mean_token_accuracy": 0.5954342484474182,
      "num_tokens": 14102200.0,
      "step": 211
    },
    {
      "epoch": 0.8582995951417004,
      "grad_norm": 8.015987460003101,
      "learning_rate": 1.1457489878542512e-05,
      "loss": 14.3672,
      "mean_token_accuracy": 0.5886507034301758,
      "num_tokens": 14170374.0,
      "step": 212
    },
    {
      "epoch": 0.8623481781376519,
      "grad_norm": 7.8139654343716565,
      "learning_rate": 1.1417004048582996e-05,
      "loss": 14.0312,
      "mean_token_accuracy": 0.5954753756523132,
      "num_tokens": 14231040.0,
      "step": 213
    },
    {
      "epoch": 0.8663967611336032,
      "grad_norm": 7.595765550190981,
      "learning_rate": 1.1376518218623482e-05,
      "loss": 13.9375,
      "mean_token_accuracy": 0.5962504744529724,
      "num_tokens": 14296979.0,
      "step": 214
    },
    {
      "epoch": 0.8704453441295547,
      "grad_norm": 8.046423435434846,
      "learning_rate": 1.1336032388663969e-05,
      "loss": 14.6328,
      "mean_token_accuracy": 0.5841693878173828,
      "num_tokens": 14372997.0,
      "step": 215
    },
    {
      "epoch": 0.8704453441295547,
      "eval_loss": 1.720312476158142,
      "eval_mean_token_accuracy": 0.5980706065893173,
      "eval_num_tokens": 14372997.0,
      "eval_runtime": 0.6391,
      "eval_samples_per_second": 312.957,
      "eval_steps_per_second": 6.259,
      "step": 215
    },
    {
      "epoch": 0.8744939271255061,
      "grad_norm": 8.7705422913405,
      "learning_rate": 1.1295546558704453e-05,
      "loss": 14.0859,
      "mean_token_accuracy": 0.5949749946594238,
      "num_tokens": 14438573.0,
      "step": 216
    },
    {
      "epoch": 0.8785425101214575,
      "grad_norm": 7.832794988425434,
      "learning_rate": 1.1255060728744939e-05,
      "loss": 13.1484,
      "mean_token_accuracy": 0.613569438457489,
      "num_tokens": 14501484.0,
      "step": 217
    },
    {
      "epoch": 0.8825910931174089,
      "grad_norm": 7.669341614248951,
      "learning_rate": 1.1214574898785427e-05,
      "loss": 14.7031,
      "mean_token_accuracy": 0.5858564972877502,
      "num_tokens": 14574868.0,
      "step": 218
    },
    {
      "epoch": 0.8866396761133604,
      "grad_norm": 7.973334988452109,
      "learning_rate": 1.1174089068825913e-05,
      "loss": 13.8047,
      "mean_token_accuracy": 0.5929080247879028,
      "num_tokens": 14635874.0,
      "step": 219
    },
    {
      "epoch": 0.8906882591093117,
      "grad_norm": 9.162769594031285,
      "learning_rate": 1.1133603238866398e-05,
      "loss": 14.6016,
      "mean_token_accuracy": 0.5782532095909119,
      "num_tokens": 14698568.0,
      "step": 220
    },
    {
      "epoch": 0.8906882591093117,
      "eval_loss": 1.7178125381469727,
      "eval_mean_token_accuracy": 0.5981580018997192,
      "eval_num_tokens": 14698568.0,
      "eval_runtime": 0.6351,
      "eval_samples_per_second": 314.936,
      "eval_steps_per_second": 6.299,
      "step": 220
    },
    {
      "epoch": 0.8947368421052632,
      "grad_norm": 7.206939771632334,
      "learning_rate": 1.1093117408906884e-05,
      "loss": 13.6406,
      "mean_token_accuracy": 0.5951396226882935,
      "num_tokens": 14762701.0,
      "step": 221
    },
    {
      "epoch": 0.8987854251012146,
      "grad_norm": 8.38441745268265,
      "learning_rate": 1.105263157894737e-05,
      "loss": 14.3203,
      "mean_token_accuracy": 0.5884565114974976,
      "num_tokens": 14836157.0,
      "step": 222
    },
    {
      "epoch": 0.902834008097166,
      "grad_norm": 7.099567588103198,
      "learning_rate": 1.1012145748987854e-05,
      "loss": 13.6875,
      "mean_token_accuracy": 0.6000562310218811,
      "num_tokens": 14903783.0,
      "step": 223
    },
    {
      "epoch": 0.9068825910931174,
      "grad_norm": 7.639650741767317,
      "learning_rate": 1.0971659919028342e-05,
      "loss": 13.1719,
      "mean_token_accuracy": 0.6127068996429443,
      "num_tokens": 14964145.0,
      "step": 224
    },
    {
      "epoch": 0.9109311740890689,
      "grad_norm": 7.438961030040257,
      "learning_rate": 1.0931174089068828e-05,
      "loss": 14.3281,
      "mean_token_accuracy": 0.5837487578392029,
      "num_tokens": 15042849.0,
      "step": 225
    },
    {
      "epoch": 0.9109311740890689,
      "eval_loss": 1.7165625095367432,
      "eval_mean_token_accuracy": 0.5983149856328964,
      "eval_num_tokens": 15042849.0,
      "eval_runtime": 0.6312,
      "eval_samples_per_second": 316.878,
      "eval_steps_per_second": 6.338,
      "step": 225
    },
    {
      "epoch": 0.9149797570850202,
      "grad_norm": 8.491938043424565,
      "learning_rate": 1.0890688259109313e-05,
      "loss": 13.75,
      "mean_token_accuracy": 0.5973576903343201,
      "num_tokens": 15107705.0,
      "step": 226
    },
    {
      "epoch": 0.9190283400809717,
      "grad_norm": 7.63291712789061,
      "learning_rate": 1.0850202429149799e-05,
      "loss": 13.9375,
      "mean_token_accuracy": 0.5994954109191895,
      "num_tokens": 15179511.0,
      "step": 227
    },
    {
      "epoch": 0.9230769230769231,
      "grad_norm": 7.056222313978797,
      "learning_rate": 1.0809716599190285e-05,
      "loss": 13.75,
      "mean_token_accuracy": 0.6037131547927856,
      "num_tokens": 15240170.0,
      "step": 228
    },
    {
      "epoch": 0.9271255060728745,
      "grad_norm": 7.894406360447493,
      "learning_rate": 1.076923076923077e-05,
      "loss": 13.4609,
      "mean_token_accuracy": 0.5994337201118469,
      "num_tokens": 15300626.0,
      "step": 229
    },
    {
      "epoch": 0.9311740890688259,
      "grad_norm": 8.960700403802914,
      "learning_rate": 1.0728744939271255e-05,
      "loss": 13.7578,
      "mean_token_accuracy": 0.5973832607269287,
      "num_tokens": 15366420.0,
      "step": 230
    },
    {
      "epoch": 0.9311740890688259,
      "eval_loss": 1.713437557220459,
      "eval_mean_token_accuracy": 0.5988392233848572,
      "eval_num_tokens": 15366420.0,
      "eval_runtime": 0.6821,
      "eval_samples_per_second": 293.227,
      "eval_steps_per_second": 5.865,
      "step": 230
    },
    {
      "epoch": 0.9352226720647774,
      "grad_norm": 8.458253193834494,
      "learning_rate": 1.0688259109311743e-05,
      "loss": 12.6172,
      "mean_token_accuracy": 0.6170424222946167,
      "num_tokens": 15421793.0,
      "step": 231
    },
    {
      "epoch": 0.9392712550607287,
      "grad_norm": 7.0686015225039665,
      "learning_rate": 1.0647773279352228e-05,
      "loss": 13.5625,
      "mean_token_accuracy": 0.607692301273346,
      "num_tokens": 15485817.0,
      "step": 232
    },
    {
      "epoch": 0.9433198380566802,
      "grad_norm": 7.576993960188947,
      "learning_rate": 1.0607287449392714e-05,
      "loss": 13.8984,
      "mean_token_accuracy": 0.5907003879547119,
      "num_tokens": 15564078.0,
      "step": 233
    },
    {
      "epoch": 0.9473684210526315,
      "grad_norm": 6.97349951130989,
      "learning_rate": 1.05668016194332e-05,
      "loss": 13.2969,
      "mean_token_accuracy": 0.601026713848114,
      "num_tokens": 15633099.0,
      "step": 234
    },
    {
      "epoch": 0.951417004048583,
      "grad_norm": 7.03291868031142,
      "learning_rate": 1.0526315789473684e-05,
      "loss": 13.9062,
      "mean_token_accuracy": 0.5920431613922119,
      "num_tokens": 15710279.0,
      "step": 235
    },
    {
      "epoch": 0.951417004048583,
      "eval_loss": 1.7118749618530273,
      "eval_mean_token_accuracy": 0.5990229398012161,
      "eval_num_tokens": 15710279.0,
      "eval_runtime": 0.6566,
      "eval_samples_per_second": 304.588,
      "eval_steps_per_second": 6.092,
      "step": 235
    },
    {
      "epoch": 0.9554655870445344,
      "grad_norm": 6.669888447299149,
      "learning_rate": 1.048582995951417e-05,
      "loss": 14.1641,
      "mean_token_accuracy": 0.5870757102966309,
      "num_tokens": 15788970.0,
      "step": 236
    },
    {
      "epoch": 0.9595141700404858,
      "grad_norm": 7.4727551311500635,
      "learning_rate": 1.0445344129554658e-05,
      "loss": 13.3203,
      "mean_token_accuracy": 0.6026445031166077,
      "num_tokens": 15854074.0,
      "step": 237
    },
    {
      "epoch": 0.9635627530364372,
      "grad_norm": 7.907361177198574,
      "learning_rate": 1.0404858299595141e-05,
      "loss": 12.4453,
      "mean_token_accuracy": 0.6250936388969421,
      "num_tokens": 15915540.0,
      "step": 238
    },
    {
      "epoch": 0.9676113360323887,
      "grad_norm": 8.43850635314641,
      "learning_rate": 1.0364372469635629e-05,
      "loss": 13.1562,
      "mean_token_accuracy": 0.6030080318450928,
      "num_tokens": 15976973.0,
      "step": 239
    },
    {
      "epoch": 0.97165991902834,
      "grad_norm": 7.4122240402160875,
      "learning_rate": 1.0323886639676115e-05,
      "loss": 13.0938,
      "mean_token_accuracy": 0.6093297004699707,
      "num_tokens": 16032923.0,
      "step": 240
    },
    {
      "epoch": 0.97165991902834,
      "eval_loss": 1.709375023841858,
      "eval_mean_token_accuracy": 0.5996900647878647,
      "eval_num_tokens": 16032923.0,
      "eval_runtime": 0.671,
      "eval_samples_per_second": 298.041,
      "eval_steps_per_second": 5.961,
      "step": 240
    },
    {
      "epoch": 0.9757085020242915,
      "grad_norm": 7.354557905255342,
      "learning_rate": 1.02834008097166e-05,
      "loss": 14.0156,
      "mean_token_accuracy": 0.6022666692733765,
      "num_tokens": 16094928.0,
      "step": 241
    },
    {
      "epoch": 0.979757085020243,
      "grad_norm": 6.816456710472039,
      "learning_rate": 1.0242914979757086e-05,
      "loss": 13.2266,
      "mean_token_accuracy": 0.6128579378128052,
      "num_tokens": 16164039.0,
      "step": 242
    },
    {
      "epoch": 0.9838056680161943,
      "grad_norm": 7.422654481239377,
      "learning_rate": 1.0202429149797572e-05,
      "loss": 14.2344,
      "mean_token_accuracy": 0.5937775373458862,
      "num_tokens": 16241339.0,
      "step": 243
    },
    {
      "epoch": 0.9878542510121457,
      "grad_norm": 6.404505160904721,
      "learning_rate": 1.0161943319838056e-05,
      "loss": 12.7969,
      "mean_token_accuracy": 0.6247119903564453,
      "num_tokens": 16315626.0,
      "step": 244
    },
    {
      "epoch": 0.9919028340080972,
      "grad_norm": 6.767676657987019,
      "learning_rate": 1.0121457489878544e-05,
      "loss": 13.0391,
      "mean_token_accuracy": 0.6165784597396851,
      "num_tokens": 16380702.0,
      "step": 245
    },
    {
      "epoch": 0.9919028340080972,
      "eval_loss": 1.7059375047683716,
      "eval_mean_token_accuracy": 0.5998604446649551,
      "eval_num_tokens": 16380702.0,
      "eval_runtime": 0.6318,
      "eval_samples_per_second": 316.572,
      "eval_steps_per_second": 6.331,
      "step": 245
    },
    {
      "epoch": 0.9959514170040485,
      "grad_norm": 8.40463068782111,
      "learning_rate": 1.008097165991903e-05,
      "loss": 13.5469,
      "mean_token_accuracy": 0.6080746054649353,
      "num_tokens": 16451754.0,
      "step": 246
    },
    {
      "epoch": 1.0,
      "grad_norm": 8.465098958585465,
      "learning_rate": 1.0040485829959515e-05,
      "loss": 13.3047,
      "mean_token_accuracy": 0.609171450138092,
      "num_tokens": 16516606.0,
      "step": 247
    },
    {
      "epoch": 1.0040485829959513,
      "grad_norm": 6.50385357417948,
      "learning_rate": 1e-05,
      "loss": 13.4297,
      "mean_token_accuracy": 0.6065150499343872,
      "num_tokens": 16584820.0,
      "step": 248
    },
    {
      "epoch": 1.008097165991903,
      "grad_norm": 7.977746144483265,
      "learning_rate": 9.959514170040487e-06,
      "loss": 13.5391,
      "mean_token_accuracy": 0.6004602313041687,
      "num_tokens": 16652244.0,
      "step": 249
    },
    {
      "epoch": 1.0121457489878543,
      "grad_norm": 7.116151586707309,
      "learning_rate": 9.919028340080973e-06,
      "loss": 13.3516,
      "mean_token_accuracy": 0.599148154258728,
      "num_tokens": 16715698.0,
      "step": 250
    },
    {
      "epoch": 1.0121457489878543,
      "eval_loss": 1.704687476158142,
      "eval_mean_token_accuracy": 0.5997532159090042,
      "eval_num_tokens": 16715698.0,
      "eval_runtime": 0.6344,
      "eval_samples_per_second": 315.241,
      "eval_steps_per_second": 6.305,
      "step": 250
    },
    {
      "epoch": 1.0161943319838056,
      "grad_norm": 7.204779568487317,
      "learning_rate": 9.878542510121458e-06,
      "loss": 13.4922,
      "mean_token_accuracy": 0.6105347275733948,
      "num_tokens": 16782076.0,
      "step": 251
    },
    {
      "epoch": 1.0202429149797572,
      "grad_norm": 6.934398308018714,
      "learning_rate": 9.838056680161944e-06,
      "loss": 12.2969,
      "mean_token_accuracy": 0.6307468414306641,
      "num_tokens": 16844000.0,
      "step": 252
    },
    {
      "epoch": 1.0242914979757085,
      "grad_norm": 6.262390941001453,
      "learning_rate": 9.79757085020243e-06,
      "loss": 12.1875,
      "mean_token_accuracy": 0.6304177045822144,
      "num_tokens": 16913997.0,
      "step": 253
    },
    {
      "epoch": 1.0283400809716599,
      "grad_norm": 7.26353805403556,
      "learning_rate": 9.757085020242916e-06,
      "loss": 13.0234,
      "mean_token_accuracy": 0.6181889176368713,
      "num_tokens": 16975505.0,
      "step": 254
    },
    {
      "epoch": 1.0323886639676114,
      "grad_norm": 7.253371839731625,
      "learning_rate": 9.7165991902834e-06,
      "loss": 12.9297,
      "mean_token_accuracy": 0.6192678809165955,
      "num_tokens": 17045831.0,
      "step": 255
    },
    {
      "epoch": 1.0323886639676114,
      "eval_loss": 1.7037500143051147,
      "eval_mean_token_accuracy": 0.600011944770813,
      "eval_num_tokens": 17045831.0,
      "eval_runtime": 0.6321,
      "eval_samples_per_second": 316.386,
      "eval_steps_per_second": 6.328,
      "step": 255
    },
    {
      "epoch": 1.0364372469635628,
      "grad_norm": 7.103579854054674,
      "learning_rate": 9.676113360323888e-06,
      "loss": 13.3125,
      "mean_token_accuracy": 0.6100465655326843,
      "num_tokens": 17114834.0,
      "step": 256
    },
    {
      "epoch": 1.040485829959514,
      "grad_norm": 7.116930652514039,
      "learning_rate": 9.635627530364373e-06,
      "loss": 13.9219,
      "mean_token_accuracy": 0.5969142317771912,
      "num_tokens": 17185739.0,
      "step": 257
    },
    {
      "epoch": 1.0445344129554657,
      "grad_norm": 7.377322519725867,
      "learning_rate": 9.595141700404859e-06,
      "loss": 12.9922,
      "mean_token_accuracy": 0.6139745712280273,
      "num_tokens": 17239157.0,
      "step": 258
    },
    {
      "epoch": 1.048582995951417,
      "grad_norm": 6.975650481941647,
      "learning_rate": 9.554655870445345e-06,
      "loss": 13.4297,
      "mean_token_accuracy": 0.6126266121864319,
      "num_tokens": 17303593.0,
      "step": 259
    },
    {
      "epoch": 1.0526315789473684,
      "grad_norm": 6.359499519445163,
      "learning_rate": 9.514170040485831e-06,
      "loss": 13.2734,
      "mean_token_accuracy": 0.6133294105529785,
      "num_tokens": 17368146.0,
      "step": 260
    },
    {
      "epoch": 1.0526315789473684,
      "eval_loss": 1.703125,
      "eval_mean_token_accuracy": 0.6001382917165756,
      "eval_num_tokens": 17368146.0,
      "eval_runtime": 0.6446,
      "eval_samples_per_second": 310.258,
      "eval_steps_per_second": 6.205,
      "step": 260
    },
    {
      "epoch": 1.05668016194332,
      "grad_norm": 6.671493038575028,
      "learning_rate": 9.473684210526315e-06,
      "loss": 12.0781,
      "mean_token_accuracy": 0.6352812647819519,
      "num_tokens": 17428599.0,
      "step": 261
    },
    {
      "epoch": 1.0607287449392713,
      "grad_norm": 7.378756810172295,
      "learning_rate": 9.433198380566803e-06,
      "loss": 12.2734,
      "mean_token_accuracy": 0.6308179497718811,
      "num_tokens": 17491185.0,
      "step": 262
    },
    {
      "epoch": 1.0647773279352226,
      "grad_norm": 7.720411271822566,
      "learning_rate": 9.392712550607288e-06,
      "loss": 13.9922,
      "mean_token_accuracy": 0.5976829528808594,
      "num_tokens": 17558922.0,
      "step": 263
    },
    {
      "epoch": 1.0688259109311742,
      "grad_norm": 7.148823701274059,
      "learning_rate": 9.352226720647774e-06,
      "loss": 12.6562,
      "mean_token_accuracy": 0.6221409440040588,
      "num_tokens": 17625704.0,
      "step": 264
    },
    {
      "epoch": 1.0728744939271255,
      "grad_norm": 6.952211161763132,
      "learning_rate": 9.31174089068826e-06,
      "loss": 13.0,
      "mean_token_accuracy": 0.6169036030769348,
      "num_tokens": 17693375.0,
      "step": 265
    },
    {
      "epoch": 1.0728744939271255,
      "eval_loss": 1.7009375095367432,
      "eval_mean_token_accuracy": 0.6005319058895111,
      "eval_num_tokens": 17693375.0,
      "eval_runtime": 0.6522,
      "eval_samples_per_second": 306.677,
      "eval_steps_per_second": 6.134,
      "step": 265
    },
    {
      "epoch": 1.0769230769230769,
      "grad_norm": 7.532405610943497,
      "learning_rate": 9.271255060728746e-06,
      "loss": 13.4609,
      "mean_token_accuracy": 0.6039229035377502,
      "num_tokens": 17756505.0,
      "step": 266
    },
    {
      "epoch": 1.0809716599190284,
      "grad_norm": 6.463674669796744,
      "learning_rate": 9.230769230769232e-06,
      "loss": 12.7422,
      "mean_token_accuracy": 0.6189759969711304,
      "num_tokens": 17827037.0,
      "step": 267
    },
    {
      "epoch": 1.0850202429149798,
      "grad_norm": 7.120027379484065,
      "learning_rate": 9.190283400809717e-06,
      "loss": 13.7656,
      "mean_token_accuracy": 0.5981738567352295,
      "num_tokens": 17892704.0,
      "step": 268
    },
    {
      "epoch": 1.0890688259109311,
      "grad_norm": 6.2456984016212065,
      "learning_rate": 9.149797570850203e-06,
      "loss": 12.7344,
      "mean_token_accuracy": 0.6219497919082642,
      "num_tokens": 17968213.0,
      "step": 269
    },
    {
      "epoch": 1.0931174089068827,
      "grad_norm": 6.297910754288368,
      "learning_rate": 9.109311740890689e-06,
      "loss": 12.7188,
      "mean_token_accuracy": 0.6200737357139587,
      "num_tokens": 18039604.0,
      "step": 270
    },
    {
      "epoch": 1.0931174089068827,
      "eval_loss": 1.7000000476837158,
      "eval_mean_token_accuracy": 0.6004240810871124,
      "eval_num_tokens": 18039604.0,
      "eval_runtime": 0.6437,
      "eval_samples_per_second": 310.685,
      "eval_steps_per_second": 6.214,
      "step": 270
    },
    {
      "epoch": 1.097165991902834,
      "grad_norm": 7.198610948868953,
      "learning_rate": 9.068825910931175e-06,
      "loss": 12.2812,
      "mean_token_accuracy": 0.6254372000694275,
      "num_tokens": 18096278.0,
      "step": 271
    },
    {
      "epoch": 1.1012145748987854,
      "grad_norm": 6.306684772679187,
      "learning_rate": 9.02834008097166e-06,
      "loss": 13.7031,
      "mean_token_accuracy": 0.6017244458198547,
      "num_tokens": 18175556.0,
      "step": 272
    },
    {
      "epoch": 1.1052631578947367,
      "grad_norm": 6.50769292978356,
      "learning_rate": 8.987854251012147e-06,
      "loss": 12.2578,
      "mean_token_accuracy": 0.6318495273590088,
      "num_tokens": 18243307.0,
      "step": 273
    },
    {
      "epoch": 1.1093117408906883,
      "grad_norm": 6.913532871754586,
      "learning_rate": 8.947368421052632e-06,
      "loss": 12.9922,
      "mean_token_accuracy": 0.61337810754776,
      "num_tokens": 18309195.0,
      "step": 274
    },
    {
      "epoch": 1.1133603238866396,
      "grad_norm": 6.884920896555758,
      "learning_rate": 8.906882591093118e-06,
      "loss": 13.875,
      "mean_token_accuracy": 0.6053746342658997,
      "num_tokens": 18376761.0,
      "step": 275
    },
    {
      "epoch": 1.1133603238866396,
      "eval_loss": 1.6981250047683716,
      "eval_mean_token_accuracy": 0.6006451994180679,
      "eval_num_tokens": 18376761.0,
      "eval_runtime": 0.6441,
      "eval_samples_per_second": 310.493,
      "eval_steps_per_second": 6.21,
      "step": 275
    },
    {
      "epoch": 1.117408906882591,
      "grad_norm": 7.596057850426039,
      "learning_rate": 8.866396761133604e-06,
      "loss": 13.5703,
      "mean_token_accuracy": 0.5959470868110657,
      "num_tokens": 18438213.0,
      "step": 276
    },
    {
      "epoch": 1.1214574898785425,
      "grad_norm": 7.244460879105897,
      "learning_rate": 8.82591093117409e-06,
      "loss": 13.6094,
      "mean_token_accuracy": 0.6075268983840942,
      "num_tokens": 18508957.0,
      "step": 277
    },
    {
      "epoch": 1.125506072874494,
      "grad_norm": 7.695579384649985,
      "learning_rate": 8.785425101214575e-06,
      "loss": 13.1641,
      "mean_token_accuracy": 0.6176554560661316,
      "num_tokens": 18565083.0,
      "step": 278
    },
    {
      "epoch": 1.1295546558704452,
      "grad_norm": 6.522870352716677,
      "learning_rate": 8.744939271255063e-06,
      "loss": 12.6953,
      "mean_token_accuracy": 0.6225432753562927,
      "num_tokens": 18633531.0,
      "step": 279
    },
    {
      "epoch": 1.1336032388663968,
      "grad_norm": 7.604533792660402,
      "learning_rate": 8.704453441295547e-06,
      "loss": 12.4688,
      "mean_token_accuracy": 0.6240532398223877,
      "num_tokens": 18689313.0,
      "step": 280
    },
    {
      "epoch": 1.1336032388663968,
      "eval_loss": 1.6974999904632568,
      "eval_mean_token_accuracy": 0.6007698774337769,
      "eval_num_tokens": 18689313.0,
      "eval_runtime": 0.6368,
      "eval_samples_per_second": 314.051,
      "eval_steps_per_second": 6.281,
      "step": 280
    },
    {
      "epoch": 1.1376518218623481,
      "grad_norm": 6.192506768790122,
      "learning_rate": 8.663967611336033e-06,
      "loss": 13.0312,
      "mean_token_accuracy": 0.6120434999465942,
      "num_tokens": 18756467.0,
      "step": 281
    },
    {
      "epoch": 1.1417004048582995,
      "grad_norm": 5.735298339214011,
      "learning_rate": 8.62348178137652e-06,
      "loss": 12.3672,
      "mean_token_accuracy": 0.6290688514709473,
      "num_tokens": 18830631.0,
      "step": 282
    },
    {
      "epoch": 1.145748987854251,
      "grad_norm": 6.8807292772906985,
      "learning_rate": 8.582995951417005e-06,
      "loss": 12.9688,
      "mean_token_accuracy": 0.6203005313873291,
      "num_tokens": 18902033.0,
      "step": 283
    },
    {
      "epoch": 1.1497975708502024,
      "grad_norm": 6.299127839988924,
      "learning_rate": 8.54251012145749e-06,
      "loss": 11.7344,
      "mean_token_accuracy": 0.6383894085884094,
      "num_tokens": 18965552.0,
      "step": 284
    },
    {
      "epoch": 1.1538461538461537,
      "grad_norm": 7.602932263969704,
      "learning_rate": 8.502024291497976e-06,
      "loss": 13.7656,
      "mean_token_accuracy": 0.6015710830688477,
      "num_tokens": 19032067.0,
      "step": 285
    },
    {
      "epoch": 1.1538461538461537,
      "eval_loss": 1.6959375143051147,
      "eval_mean_token_accuracy": 0.6004112809896469,
      "eval_num_tokens": 19032067.0,
      "eval_runtime": 0.6551,
      "eval_samples_per_second": 305.283,
      "eval_steps_per_second": 6.106,
      "step": 285
    },
    {
      "epoch": 1.1578947368421053,
      "grad_norm": 6.900267004886904,
      "learning_rate": 8.461538461538462e-06,
      "loss": 12.1562,
      "mean_token_accuracy": 0.6340711116790771,
      "num_tokens": 19091771.0,
      "step": 286
    },
    {
      "epoch": 1.1619433198380567,
      "grad_norm": 7.015232796047964,
      "learning_rate": 8.421052631578948e-06,
      "loss": 12.7578,
      "mean_token_accuracy": 0.6194669604301453,
      "num_tokens": 19160875.0,
      "step": 287
    },
    {
      "epoch": 1.165991902834008,
      "grad_norm": 8.07776854225312,
      "learning_rate": 8.380566801619434e-06,
      "loss": 12.8672,
      "mean_token_accuracy": 0.621356189250946,
      "num_tokens": 19234248.0,
      "step": 288
    },
    {
      "epoch": 1.1700404858299596,
      "grad_norm": 6.886729565425438,
      "learning_rate": 8.340080971659919e-06,
      "loss": 13.3125,
      "mean_token_accuracy": 0.6096766591072083,
      "num_tokens": 19300037.0,
      "step": 289
    },
    {
      "epoch": 1.174089068825911,
      "grad_norm": 7.451714517536367,
      "learning_rate": 8.299595141700405e-06,
      "loss": 12.9062,
      "mean_token_accuracy": 0.6140288710594177,
      "num_tokens": 19359507.0,
      "step": 290
    },
    {
      "epoch": 1.174089068825911,
      "eval_loss": 1.6950000524520874,
      "eval_mean_token_accuracy": 0.600911945104599,
      "eval_num_tokens": 19359507.0,
      "eval_runtime": 0.6622,
      "eval_samples_per_second": 302.012,
      "eval_steps_per_second": 6.04,
      "step": 290
    },
    {
      "epoch": 1.1781376518218623,
      "grad_norm": 6.19859060055071,
      "learning_rate": 8.259109311740891e-06,
      "loss": 12.9922,
      "mean_token_accuracy": 0.6148094534873962,
      "num_tokens": 19426042.0,
      "step": 291
    },
    {
      "epoch": 1.1821862348178138,
      "grad_norm": 6.3346510938910345,
      "learning_rate": 8.218623481781377e-06,
      "loss": 12.8516,
      "mean_token_accuracy": 0.615065336227417,
      "num_tokens": 19494767.0,
      "step": 292
    },
    {
      "epoch": 1.1862348178137652,
      "grad_norm": 7.185444383014579,
      "learning_rate": 8.178137651821862e-06,
      "loss": 12.8281,
      "mean_token_accuracy": 0.6221780776977539,
      "num_tokens": 19561762.0,
      "step": 293
    },
    {
      "epoch": 1.1902834008097165,
      "grad_norm": 7.763951709420421,
      "learning_rate": 8.13765182186235e-06,
      "loss": 13.5312,
      "mean_token_accuracy": 0.6010787487030029,
      "num_tokens": 19629313.0,
      "step": 294
    },
    {
      "epoch": 1.194331983805668,
      "grad_norm": 6.6829771026826785,
      "learning_rate": 8.097165991902834e-06,
      "loss": 12.6875,
      "mean_token_accuracy": 0.6254391074180603,
      "num_tokens": 19699694.0,
      "step": 295
    },
    {
      "epoch": 1.194331983805668,
      "eval_loss": 1.6931250095367432,
      "eval_mean_token_accuracy": 0.6008877456188202,
      "eval_num_tokens": 19699694.0,
      "eval_runtime": 0.6362,
      "eval_samples_per_second": 314.38,
      "eval_steps_per_second": 6.288,
      "step": 295
    },
    {
      "epoch": 1.1983805668016194,
      "grad_norm": 6.897120986568694,
      "learning_rate": 8.056680161943322e-06,
      "loss": 12.2109,
      "mean_token_accuracy": 0.6307302713394165,
      "num_tokens": 19757434.0,
      "step": 296
    },
    {
      "epoch": 1.2024291497975708,
      "grad_norm": 7.44384813842545,
      "learning_rate": 8.016194331983806e-06,
      "loss": 12.7656,
      "mean_token_accuracy": 0.6187056303024292,
      "num_tokens": 19820570.0,
      "step": 297
    },
    {
      "epoch": 1.2064777327935223,
      "grad_norm": 6.4203676546627015,
      "learning_rate": 7.975708502024292e-06,
      "loss": 13.2812,
      "mean_token_accuracy": 0.6042314767837524,
      "num_tokens": 19890728.0,
      "step": 298
    },
    {
      "epoch": 1.2105263157894737,
      "grad_norm": 6.848564846162855,
      "learning_rate": 7.935222672064778e-06,
      "loss": 13.0703,
      "mean_token_accuracy": 0.6080402731895447,
      "num_tokens": 19959322.0,
      "step": 299
    },
    {
      "epoch": 1.214574898785425,
      "grad_norm": 6.862334865838493,
      "learning_rate": 7.894736842105265e-06,
      "loss": 13.3516,
      "mean_token_accuracy": 0.6005032062530518,
      "num_tokens": 20031722.0,
      "step": 300
    },
    {
      "epoch": 1.214574898785425,
      "eval_loss": 1.691562533378601,
      "eval_mean_token_accuracy": 0.6006248891353607,
      "eval_num_tokens": 20031722.0,
      "eval_runtime": 0.6358,
      "eval_samples_per_second": 314.55,
      "eval_steps_per_second": 6.291,
      "step": 300
    },
    {
      "epoch": 1.2186234817813766,
      "grad_norm": 7.001796551105443,
      "learning_rate": 7.854251012145749e-06,
      "loss": 12.7109,
      "mean_token_accuracy": 0.6257545948028564,
      "num_tokens": 20104837.0,
      "step": 301
    },
    {
      "epoch": 1.222672064777328,
      "grad_norm": 7.289919019857029,
      "learning_rate": 7.813765182186235e-06,
      "loss": 12.8828,
      "mean_token_accuracy": 0.6084967851638794,
      "num_tokens": 20172668.0,
      "step": 302
    },
    {
      "epoch": 1.2267206477732793,
      "grad_norm": 6.345630914099327,
      "learning_rate": 7.773279352226721e-06,
      "loss": 12.9141,
      "mean_token_accuracy": 0.6153846383094788,
      "num_tokens": 20241372.0,
      "step": 303
    },
    {
      "epoch": 1.2307692307692308,
      "grad_norm": 7.5520644152330245,
      "learning_rate": 7.732793522267207e-06,
      "loss": 13.6797,
      "mean_token_accuracy": 0.607166588306427,
      "num_tokens": 20310730.0,
      "step": 304
    },
    {
      "epoch": 1.2348178137651822,
      "grad_norm": 8.642720338892964,
      "learning_rate": 7.692307692307694e-06,
      "loss": 12.9219,
      "mean_token_accuracy": 0.6174276471138,
      "num_tokens": 20378755.0,
      "step": 305
    },
    {
      "epoch": 1.2348178137651822,
      "eval_loss": 1.6906249523162842,
      "eval_mean_token_accuracy": 0.601470410823822,
      "eval_num_tokens": 20378755.0,
      "eval_runtime": 0.6338,
      "eval_samples_per_second": 315.573,
      "eval_steps_per_second": 6.311,
      "step": 305
    },
    {
      "epoch": 1.2388663967611335,
      "grad_norm": 6.901290072516897,
      "learning_rate": 7.651821862348178e-06,
      "loss": 13.1094,
      "mean_token_accuracy": 0.6087667942047119,
      "num_tokens": 20449130.0,
      "step": 306
    },
    {
      "epoch": 1.242914979757085,
      "grad_norm": 5.8437165539153995,
      "learning_rate": 7.611336032388664e-06,
      "loss": 12.1797,
      "mean_token_accuracy": 0.6274487972259521,
      "num_tokens": 20511878.0,
      "step": 307
    },
    {
      "epoch": 1.2469635627530364,
      "grad_norm": 6.275119502029468,
      "learning_rate": 7.570850202429151e-06,
      "loss": 13.0,
      "mean_token_accuracy": 0.6140267252922058,
      "num_tokens": 20580939.0,
      "step": 308
    },
    {
      "epoch": 1.2510121457489878,
      "grad_norm": 6.632528297788201,
      "learning_rate": 7.5303643724696364e-06,
      "loss": 12.2969,
      "mean_token_accuracy": 0.6276517510414124,
      "num_tokens": 20646337.0,
      "step": 309
    },
    {
      "epoch": 1.2550607287449393,
      "grad_norm": 6.802264706016231,
      "learning_rate": 7.489878542510122e-06,
      "loss": 12.5312,
      "mean_token_accuracy": 0.6189518570899963,
      "num_tokens": 20706089.0,
      "step": 310
    },
    {
      "epoch": 1.2550607287449393,
      "eval_loss": 1.6903125047683716,
      "eval_mean_token_accuracy": 0.6015071421861649,
      "eval_num_tokens": 20706089.0,
      "eval_runtime": 0.6486,
      "eval_samples_per_second": 308.342,
      "eval_steps_per_second": 6.167,
      "step": 310
    },
    {
      "epoch": 1.2591093117408907,
      "grad_norm": 6.618283699689559,
      "learning_rate": 7.449392712550608e-06,
      "loss": 12.2422,
      "mean_token_accuracy": 0.6300427317619324,
      "num_tokens": 20776353.0,
      "step": 311
    },
    {
      "epoch": 1.263157894736842,
      "grad_norm": 6.385834652503417,
      "learning_rate": 7.408906882591094e-06,
      "loss": 13.1641,
      "mean_token_accuracy": 0.6120158433914185,
      "num_tokens": 20839051.0,
      "step": 312
    },
    {
      "epoch": 1.2672064777327936,
      "grad_norm": 6.604656501463769,
      "learning_rate": 7.368421052631579e-06,
      "loss": 12.3516,
      "mean_token_accuracy": 0.6265976428985596,
      "num_tokens": 20907888.0,
      "step": 313
    },
    {
      "epoch": 1.271255060728745,
      "grad_norm": 6.333208429557312,
      "learning_rate": 7.327935222672065e-06,
      "loss": 12.2422,
      "mean_token_accuracy": 0.6266438961029053,
      "num_tokens": 20973573.0,
      "step": 314
    },
    {
      "epoch": 1.2753036437246963,
      "grad_norm": 6.851872002967978,
      "learning_rate": 7.2874493927125516e-06,
      "loss": 13.3438,
      "mean_token_accuracy": 0.6068040132522583,
      "num_tokens": 21041626.0,
      "step": 315
    },
    {
      "epoch": 1.2753036437246963,
      "eval_loss": 1.690000057220459,
      "eval_mean_token_accuracy": 0.6009943783283234,
      "eval_num_tokens": 21041626.0,
      "eval_runtime": 0.6597,
      "eval_samples_per_second": 303.162,
      "eval_steps_per_second": 6.063,
      "step": 315
    },
    {
      "epoch": 1.2793522267206479,
      "grad_norm": 7.312872948522801,
      "learning_rate": 7.246963562753037e-06,
      "loss": 13.3906,
      "mean_token_accuracy": 0.6119928359985352,
      "num_tokens": 21109864.0,
      "step": 316
    },
    {
      "epoch": 1.2834008097165992,
      "grad_norm": 7.258406742426028,
      "learning_rate": 7.206477732793523e-06,
      "loss": 13.6797,
      "mean_token_accuracy": 0.6022233963012695,
      "num_tokens": 21180631.0,
      "step": 317
    },
    {
      "epoch": 1.2874493927125505,
      "grad_norm": 7.236283798071484,
      "learning_rate": 7.165991902834008e-06,
      "loss": 13.5156,
      "mean_token_accuracy": 0.607459545135498,
      "num_tokens": 21249144.0,
      "step": 318
    },
    {
      "epoch": 1.291497975708502,
      "grad_norm": 6.443354672755848,
      "learning_rate": 7.125506072874494e-06,
      "loss": 12.5156,
      "mean_token_accuracy": 0.6236591339111328,
      "num_tokens": 21315677.0,
      "step": 319
    },
    {
      "epoch": 1.2955465587044535,
      "grad_norm": 6.703854250267342,
      "learning_rate": 7.0850202429149805e-06,
      "loss": 12.0312,
      "mean_token_accuracy": 0.6358534693717957,
      "num_tokens": 21370804.0,
      "step": 320
    },
    {
      "epoch": 1.2955465587044535,
      "eval_loss": 1.6896874904632568,
      "eval_mean_token_accuracy": 0.6014674603939056,
      "eval_num_tokens": 21370804.0,
      "eval_runtime": 0.6354,
      "eval_samples_per_second": 314.742,
      "eval_steps_per_second": 6.295,
      "step": 320
    },
    {
      "epoch": 1.2995951417004048,
      "grad_norm": 6.223799662239967,
      "learning_rate": 7.044534412955466e-06,
      "loss": 13.5234,
      "mean_token_accuracy": 0.6001178622245789,
      "num_tokens": 21442134.0,
      "step": 321
    },
    {
      "epoch": 1.3036437246963564,
      "grad_norm": 6.245172102328749,
      "learning_rate": 7.004048582995951e-06,
      "loss": 13.3438,
      "mean_token_accuracy": 0.6050729751586914,
      "num_tokens": 21505987.0,
      "step": 322
    },
    {
      "epoch": 1.3076923076923077,
      "grad_norm": 7.640593911902389,
      "learning_rate": 6.963562753036438e-06,
      "loss": 13.125,
      "mean_token_accuracy": 0.614285945892334,
      "num_tokens": 21571444.0,
      "step": 323
    },
    {
      "epoch": 1.311740890688259,
      "grad_norm": 6.182617066137327,
      "learning_rate": 6.923076923076923e-06,
      "loss": 12.9219,
      "mean_token_accuracy": 0.6171441078186035,
      "num_tokens": 21631622.0,
      "step": 324
    },
    {
      "epoch": 1.3157894736842106,
      "grad_norm": 6.612642501099769,
      "learning_rate": 6.882591093117409e-06,
      "loss": 13.1328,
      "mean_token_accuracy": 0.6197838187217712,
      "num_tokens": 21699600.0,
      "step": 325
    },
    {
      "epoch": 1.3157894736842106,
      "eval_loss": 1.6884374618530273,
      "eval_mean_token_accuracy": 0.6019182503223419,
      "eval_num_tokens": 21699600.0,
      "eval_runtime": 0.6487,
      "eval_samples_per_second": 308.327,
      "eval_steps_per_second": 6.167,
      "step": 325
    },
    {
      "epoch": 1.319838056680162,
      "grad_norm": 6.268645760381665,
      "learning_rate": 6.842105263157896e-06,
      "loss": 13.0859,
      "mean_token_accuracy": 0.6107904314994812,
      "num_tokens": 21770004.0,
      "step": 326
    },
    {
      "epoch": 1.3238866396761133,
      "grad_norm": 6.480172792037266,
      "learning_rate": 6.801619433198381e-06,
      "loss": 13.4766,
      "mean_token_accuracy": 0.6045525670051575,
      "num_tokens": 21835174.0,
      "step": 327
    },
    {
      "epoch": 1.3279352226720649,
      "grad_norm": 6.89015735331616,
      "learning_rate": 6.761133603238867e-06,
      "loss": 12.0156,
      "mean_token_accuracy": 0.6316675543785095,
      "num_tokens": 21901773.0,
      "step": 328
    },
    {
      "epoch": 1.3319838056680162,
      "grad_norm": 6.136078812928923,
      "learning_rate": 6.720647773279353e-06,
      "loss": 12.8594,
      "mean_token_accuracy": 0.6160688400268555,
      "num_tokens": 21973454.0,
      "step": 329
    },
    {
      "epoch": 1.3360323886639676,
      "grad_norm": 6.530535217583928,
      "learning_rate": 6.6801619433198385e-06,
      "loss": 13.3281,
      "mean_token_accuracy": 0.608657717704773,
      "num_tokens": 22043421.0,
      "step": 330
    },
    {
      "epoch": 1.3360323886639676,
      "eval_loss": 1.6871875524520874,
      "eval_mean_token_accuracy": 0.6020240485668182,
      "eval_num_tokens": 22043421.0,
      "eval_runtime": 0.6344,
      "eval_samples_per_second": 315.283,
      "eval_steps_per_second": 6.306,
      "step": 330
    },
    {
      "epoch": 1.3400809716599191,
      "grad_norm": 7.032557524136393,
      "learning_rate": 6.639676113360325e-06,
      "loss": 13.6797,
      "mean_token_accuracy": 0.5984200835227966,
      "num_tokens": 22112983.0,
      "step": 331
    },
    {
      "epoch": 1.3441295546558705,
      "grad_norm": 6.878568332018199,
      "learning_rate": 6.599190283400811e-06,
      "loss": 13.625,
      "mean_token_accuracy": 0.60277259349823,
      "num_tokens": 22174362.0,
      "step": 332
    },
    {
      "epoch": 1.3481781376518218,
      "grad_norm": 6.2805599900614615,
      "learning_rate": 6.558704453441296e-06,
      "loss": 14.2031,
      "mean_token_accuracy": 0.592296838760376,
      "num_tokens": 22245929.0,
      "step": 333
    },
    {
      "epoch": 1.3522267206477734,
      "grad_norm": 6.851335318270516,
      "learning_rate": 6.518218623481782e-06,
      "loss": 13.8516,
      "mean_token_accuracy": 0.6001754403114319,
      "num_tokens": 22315536.0,
      "step": 334
    },
    {
      "epoch": 1.3562753036437247,
      "grad_norm": 6.690083307280702,
      "learning_rate": 6.4777327935222675e-06,
      "loss": 13.9688,
      "mean_token_accuracy": 0.5958229303359985,
      "num_tokens": 22381388.0,
      "step": 335
    },
    {
      "epoch": 1.3562753036437247,
      "eval_loss": 1.6865625381469727,
      "eval_mean_token_accuracy": 0.601932093501091,
      "eval_num_tokens": 22381388.0,
      "eval_runtime": 0.6419,
      "eval_samples_per_second": 311.575,
      "eval_steps_per_second": 6.232,
      "step": 335
    },
    {
      "epoch": 1.360323886639676,
      "grad_norm": 7.1780394505714336,
      "learning_rate": 6.437246963562754e-06,
      "loss": 12.5156,
      "mean_token_accuracy": 0.6246969699859619,
      "num_tokens": 22448683.0,
      "step": 336
    },
    {
      "epoch": 1.3643724696356276,
      "grad_norm": 7.289906239024176,
      "learning_rate": 6.39676113360324e-06,
      "loss": 13.5469,
      "mean_token_accuracy": 0.6012796759605408,
      "num_tokens": 22510793.0,
      "step": 337
    },
    {
      "epoch": 1.368421052631579,
      "grad_norm": 6.901589521281514,
      "learning_rate": 6.356275303643725e-06,
      "loss": 13.5234,
      "mean_token_accuracy": 0.6072494983673096,
      "num_tokens": 22579193.0,
      "step": 338
    },
    {
      "epoch": 1.3724696356275303,
      "grad_norm": 6.256711193594864,
      "learning_rate": 6.31578947368421e-06,
      "loss": 12.7188,
      "mean_token_accuracy": 0.622461199760437,
      "num_tokens": 22649271.0,
      "step": 339
    },
    {
      "epoch": 1.376518218623482,
      "grad_norm": 5.928363958207777,
      "learning_rate": 6.275303643724697e-06,
      "loss": 12.8203,
      "mean_token_accuracy": 0.6171985268592834,
      "num_tokens": 22721597.0,
      "step": 340
    },
    {
      "epoch": 1.376518218623482,
      "eval_loss": 1.6862499713897705,
      "eval_mean_token_accuracy": 0.6023992896080017,
      "eval_num_tokens": 22721597.0,
      "eval_runtime": 0.6436,
      "eval_samples_per_second": 310.773,
      "eval_steps_per_second": 6.215,
      "step": 340
    },
    {
      "epoch": 1.3805668016194332,
      "grad_norm": 7.046994398153724,
      "learning_rate": 6.234817813765183e-06,
      "loss": 13.5625,
      "mean_token_accuracy": 0.6081273555755615,
      "num_tokens": 22788300.0,
      "step": 341
    },
    {
      "epoch": 1.3846153846153846,
      "grad_norm": 5.679939194503434,
      "learning_rate": 6.194331983805668e-06,
      "loss": 12.5,
      "mean_token_accuracy": 0.6215187907218933,
      "num_tokens": 22859065.0,
      "step": 342
    },
    {
      "epoch": 1.3886639676113361,
      "grad_norm": 6.613152950818566,
      "learning_rate": 6.153846153846155e-06,
      "loss": 12.8125,
      "mean_token_accuracy": 0.6154517531394958,
      "num_tokens": 22923264.0,
      "step": 343
    },
    {
      "epoch": 1.3927125506072875,
      "grad_norm": 6.167835687979181,
      "learning_rate": 6.11336032388664e-06,
      "loss": 12.8047,
      "mean_token_accuracy": 0.6202464699745178,
      "num_tokens": 22980128.0,
      "step": 344
    },
    {
      "epoch": 1.3967611336032388,
      "grad_norm": 6.861084643476399,
      "learning_rate": 6.0728744939271254e-06,
      "loss": 12.0469,
      "mean_token_accuracy": 0.6357682347297668,
      "num_tokens": 23038976.0,
      "step": 345
    },
    {
      "epoch": 1.3967611336032388,
      "eval_loss": 1.6846874952316284,
      "eval_mean_token_accuracy": 0.6027306467294693,
      "eval_num_tokens": 23038976.0,
      "eval_runtime": 0.6479,
      "eval_samples_per_second": 308.696,
      "eval_steps_per_second": 6.174,
      "step": 345
    },
    {
      "epoch": 1.4008097165991904,
      "grad_norm": 6.0083674637757065,
      "learning_rate": 6.0323886639676124e-06,
      "loss": 12.8516,
      "mean_token_accuracy": 0.6229951977729797,
      "num_tokens": 23113673.0,
      "step": 346
    },
    {
      "epoch": 1.4048582995951417,
      "grad_norm": 6.91334940259032,
      "learning_rate": 5.991902834008098e-06,
      "loss": 13.6016,
      "mean_token_accuracy": 0.6017605066299438,
      "num_tokens": 23182354.0,
      "step": 347
    },
    {
      "epoch": 1.408906882591093,
      "grad_norm": 7.225053680436044,
      "learning_rate": 5.951417004048583e-06,
      "loss": 13.2812,
      "mean_token_accuracy": 0.6096216440200806,
      "num_tokens": 23244819.0,
      "step": 348
    },
    {
      "epoch": 1.4129554655870447,
      "grad_norm": 6.910314738528923,
      "learning_rate": 5.91093117408907e-06,
      "loss": 12.5781,
      "mean_token_accuracy": 0.6233025193214417,
      "num_tokens": 23301143.0,
      "step": 349
    },
    {
      "epoch": 1.417004048582996,
      "grad_norm": 13.172174214194689,
      "learning_rate": 5.870445344129555e-06,
      "loss": 13.7188,
      "mean_token_accuracy": 0.5989026427268982,
      "num_tokens": 23373743.0,
      "step": 350
    },
    {
      "epoch": 1.417004048582996,
      "eval_loss": 1.683750033378601,
      "eval_mean_token_accuracy": 0.6025040000677109,
      "eval_num_tokens": 23373743.0,
      "eval_runtime": 0.638,
      "eval_samples_per_second": 313.493,
      "eval_steps_per_second": 6.27,
      "step": 350
    },
    {
      "epoch": 1.4210526315789473,
      "grad_norm": 7.059247196559215,
      "learning_rate": 5.8299595141700406e-06,
      "loss": 12.2812,
      "mean_token_accuracy": 0.6239601373672485,
      "num_tokens": 23441002.0,
      "step": 351
    },
    {
      "epoch": 1.425101214574899,
      "grad_norm": 6.949818046383843,
      "learning_rate": 5.789473684210527e-06,
      "loss": 12.3125,
      "mean_token_accuracy": 0.6313892006874084,
      "num_tokens": 23495161.0,
      "step": 352
    },
    {
      "epoch": 1.4291497975708503,
      "grad_norm": 6.059843894633797,
      "learning_rate": 5.748987854251013e-06,
      "loss": 13.2344,
      "mean_token_accuracy": 0.6091973781585693,
      "num_tokens": 23564636.0,
      "step": 353
    },
    {
      "epoch": 1.4331983805668016,
      "grad_norm": 6.849778295075088,
      "learning_rate": 5.708502024291498e-06,
      "loss": 14.1484,
      "mean_token_accuracy": 0.5898481011390686,
      "num_tokens": 23634618.0,
      "step": 354
    },
    {
      "epoch": 1.4372469635627532,
      "grad_norm": 7.196515102955155,
      "learning_rate": 5.668016194331984e-06,
      "loss": 13.4844,
      "mean_token_accuracy": 0.608712911605835,
      "num_tokens": 23697715.0,
      "step": 355
    },
    {
      "epoch": 1.4372469635627532,
      "eval_loss": 1.683437466621399,
      "eval_mean_token_accuracy": 0.6026097536087036,
      "eval_num_tokens": 23697715.0,
      "eval_runtime": 0.6368,
      "eval_samples_per_second": 314.046,
      "eval_steps_per_second": 6.281,
      "step": 355
    },
    {
      "epoch": 1.4412955465587045,
      "grad_norm": 6.083851158318087,
      "learning_rate": 5.6275303643724695e-06,
      "loss": 13.1719,
      "mean_token_accuracy": 0.6111401319503784,
      "num_tokens": 23762805.0,
      "step": 356
    },
    {
      "epoch": 1.4453441295546559,
      "grad_norm": 6.3854204394565865,
      "learning_rate": 5.5870445344129565e-06,
      "loss": 14.1484,
      "mean_token_accuracy": 0.5870954990386963,
      "num_tokens": 23824801.0,
      "step": 357
    },
    {
      "epoch": 1.4493927125506074,
      "grad_norm": 6.205029195096519,
      "learning_rate": 5.546558704453442e-06,
      "loss": 12.5078,
      "mean_token_accuracy": 0.6245027184486389,
      "num_tokens": 23893486.0,
      "step": 358
    },
    {
      "epoch": 1.4534412955465588,
      "grad_norm": 6.498352605010108,
      "learning_rate": 5.506072874493927e-06,
      "loss": 13.1406,
      "mean_token_accuracy": 0.6103119254112244,
      "num_tokens": 23956739.0,
      "step": 359
    },
    {
      "epoch": 1.45748987854251,
      "grad_norm": 7.863658695213157,
      "learning_rate": 5.465587044534414e-06,
      "loss": 12.2969,
      "mean_token_accuracy": 0.6250467896461487,
      "num_tokens": 24015613.0,
      "step": 360
    },
    {
      "epoch": 1.45748987854251,
      "eval_loss": 1.6828124523162842,
      "eval_mean_token_accuracy": 0.6025930494070053,
      "eval_num_tokens": 24015613.0,
      "eval_runtime": 0.6414,
      "eval_samples_per_second": 311.802,
      "eval_steps_per_second": 6.236,
      "step": 360
    },
    {
      "epoch": 1.4615384615384617,
      "grad_norm": 6.004778979711357,
      "learning_rate": 5.425101214574899e-06,
      "loss": 13.375,
      "mean_token_accuracy": 0.6013794541358948,
      "num_tokens": 24093533.0,
      "step": 361
    },
    {
      "epoch": 1.465587044534413,
      "grad_norm": 6.524314866552431,
      "learning_rate": 5.384615384615385e-06,
      "loss": 13.1094,
      "mean_token_accuracy": 0.6128621101379395,
      "num_tokens": 24172029.0,
      "step": 362
    },
    {
      "epoch": 1.4696356275303644,
      "grad_norm": 5.757817859771559,
      "learning_rate": 5.344129554655872e-06,
      "loss": 12.7188,
      "mean_token_accuracy": 0.6142829060554504,
      "num_tokens": 24238844.0,
      "step": 363
    },
    {
      "epoch": 1.4736842105263157,
      "grad_norm": 5.843888667434192,
      "learning_rate": 5.303643724696357e-06,
      "loss": 12.4688,
      "mean_token_accuracy": 0.6290068030357361,
      "num_tokens": 24308602.0,
      "step": 364
    },
    {
      "epoch": 1.4777327935222673,
      "grad_norm": 7.095669449455518,
      "learning_rate": 5.263157894736842e-06,
      "loss": 13.2891,
      "mean_token_accuracy": 0.6107263565063477,
      "num_tokens": 24381011.0,
      "step": 365
    },
    {
      "epoch": 1.4777327935222673,
      "eval_loss": 1.6828124523162842,
      "eval_mean_token_accuracy": 0.6028460413217545,
      "eval_num_tokens": 24381011.0,
      "eval_runtime": 0.6426,
      "eval_samples_per_second": 311.246,
      "eval_steps_per_second": 6.225,
      "step": 365
    },
    {
      "epoch": 1.4817813765182186,
      "grad_norm": 7.089500617007727,
      "learning_rate": 5.222672064777329e-06,
      "loss": 13.2422,
      "mean_token_accuracy": 0.6082656383514404,
      "num_tokens": 24447301.0,
      "step": 366
    },
    {
      "epoch": 1.48582995951417,
      "grad_norm": 5.759008858515851,
      "learning_rate": 5.1821862348178145e-06,
      "loss": 11.7812,
      "mean_token_accuracy": 0.6373518109321594,
      "num_tokens": 24521514.0,
      "step": 367
    },
    {
      "epoch": 1.4898785425101215,
      "grad_norm": 7.257765350400875,
      "learning_rate": 5.1417004048583e-06,
      "loss": 13.4219,
      "mean_token_accuracy": 0.6083607077598572,
      "num_tokens": 24587266.0,
      "step": 368
    },
    {
      "epoch": 1.4939271255060729,
      "grad_norm": 6.617774961612806,
      "learning_rate": 5.101214574898786e-06,
      "loss": 11.9922,
      "mean_token_accuracy": 0.6397205591201782,
      "num_tokens": 24647883.0,
      "step": 369
    },
    {
      "epoch": 1.4979757085020242,
      "grad_norm": 6.034070184034884,
      "learning_rate": 5.060728744939272e-06,
      "loss": 14.0781,
      "mean_token_accuracy": 0.5942122936248779,
      "num_tokens": 24723728.0,
      "step": 370
    },
    {
      "epoch": 1.4979757085020242,
      "eval_loss": 1.682187557220459,
      "eval_mean_token_accuracy": 0.6028562039136887,
      "eval_num_tokens": 24723728.0,
      "eval_runtime": 0.6595,
      "eval_samples_per_second": 303.264,
      "eval_steps_per_second": 6.065,
      "step": 370
    },
    {
      "epoch": 1.5020242914979756,
      "grad_norm": 7.413748338165612,
      "learning_rate": 5.020242914979757e-06,
      "loss": 13.0234,
      "mean_token_accuracy": 0.6105521321296692,
      "num_tokens": 24786206.0,
      "step": 371
    },
    {
      "epoch": 1.5060728744939271,
      "grad_norm": 7.4171774757668745,
      "learning_rate": 4.9797570850202435e-06,
      "loss": 13.1641,
      "mean_token_accuracy": 0.6107207536697388,
      "num_tokens": 24851452.0,
      "step": 372
    },
    {
      "epoch": 1.5101214574898787,
      "grad_norm": 6.678125850261295,
      "learning_rate": 4.939271255060729e-06,
      "loss": 12.1406,
      "mean_token_accuracy": 0.6283853650093079,
      "num_tokens": 24912625.0,
      "step": 373
    },
    {
      "epoch": 1.5141700404858298,
      "grad_norm": 6.7202926787891375,
      "learning_rate": 4.898785425101215e-06,
      "loss": 13.0469,
      "mean_token_accuracy": 0.6010765433311462,
      "num_tokens": 24976596.0,
      "step": 374
    },
    {
      "epoch": 1.5182186234817814,
      "grad_norm": 7.189454405906928,
      "learning_rate": 4.8582995951417e-06,
      "loss": 11.9141,
      "mean_token_accuracy": 0.6381216049194336,
      "num_tokens": 25030493.0,
      "step": 375
    },
    {
      "epoch": 1.5182186234817814,
      "eval_loss": 1.6809375286102295,
      "eval_mean_token_accuracy": 0.6026807576417923,
      "eval_num_tokens": 25030493.0,
      "eval_runtime": 0.6393,
      "eval_samples_per_second": 312.855,
      "eval_steps_per_second": 6.257,
      "step": 375
    },
    {
      "epoch": 1.522267206477733,
      "grad_norm": 6.309792255182094,
      "learning_rate": 4.817813765182186e-06,
      "loss": 12.0938,
      "mean_token_accuracy": 0.6288850903511047,
      "num_tokens": 25094133.0,
      "step": 376
    },
    {
      "epoch": 1.526315789473684,
      "grad_norm": 7.06660860300255,
      "learning_rate": 4.7773279352226725e-06,
      "loss": 12.7812,
      "mean_token_accuracy": 0.6158961057662964,
      "num_tokens": 25155860.0,
      "step": 377
    },
    {
      "epoch": 1.5303643724696356,
      "grad_norm": 8.076900211471674,
      "learning_rate": 4.736842105263158e-06,
      "loss": 12.5859,
      "mean_token_accuracy": 0.6151171922683716,
      "num_tokens": 25211146.0,
      "step": 378
    },
    {
      "epoch": 1.5344129554655872,
      "grad_norm": 6.221731683405277,
      "learning_rate": 4.696356275303644e-06,
      "loss": 12.4609,
      "mean_token_accuracy": 0.6272507905960083,
      "num_tokens": 25280353.0,
      "step": 379
    },
    {
      "epoch": 1.5384615384615383,
      "grad_norm": 7.4685425851793665,
      "learning_rate": 4.65587044534413e-06,
      "loss": 13.1562,
      "mean_token_accuracy": 0.6171966195106506,
      "num_tokens": 25335393.0,
      "step": 380
    },
    {
      "epoch": 1.5384615384615383,
      "eval_loss": 1.6803125143051147,
      "eval_mean_token_accuracy": 0.6025642305612564,
      "eval_num_tokens": 25335393.0,
      "eval_runtime": 0.6333,
      "eval_samples_per_second": 315.815,
      "eval_steps_per_second": 6.316,
      "step": 380
    },
    {
      "epoch": 1.54251012145749,
      "grad_norm": 6.476918972266226,
      "learning_rate": 4.615384615384616e-06,
      "loss": 12.8594,
      "mean_token_accuracy": 0.6220955848693848,
      "num_tokens": 25413527.0,
      "step": 381
    },
    {
      "epoch": 1.5465587044534415,
      "grad_norm": 6.648186663706087,
      "learning_rate": 4.5748987854251014e-06,
      "loss": 12.5703,
      "mean_token_accuracy": 0.6231284141540527,
      "num_tokens": 25480245.0,
      "step": 382
    },
    {
      "epoch": 1.5506072874493926,
      "grad_norm": 7.367613086661741,
      "learning_rate": 4.534412955465588e-06,
      "loss": 13.6406,
      "mean_token_accuracy": 0.6029731631278992,
      "num_tokens": 25547914.0,
      "step": 383
    },
    {
      "epoch": 1.5546558704453441,
      "grad_norm": 5.503766366101462,
      "learning_rate": 4.493927125506074e-06,
      "loss": 13.8125,
      "mean_token_accuracy": 0.5945603847503662,
      "num_tokens": 25631660.0,
      "step": 384
    },
    {
      "epoch": 1.5587044534412957,
      "grad_norm": 6.8977689320409885,
      "learning_rate": 4.453441295546559e-06,
      "loss": 13.5156,
      "mean_token_accuracy": 0.6095985174179077,
      "num_tokens": 25696588.0,
      "step": 385
    },
    {
      "epoch": 1.5587044534412957,
      "eval_loss": 1.6799999475479126,
      "eval_mean_token_accuracy": 0.6023529767990112,
      "eval_num_tokens": 25696588.0,
      "eval_runtime": 0.636,
      "eval_samples_per_second": 314.454,
      "eval_steps_per_second": 6.289,
      "step": 385
    },
    {
      "epoch": 1.5627530364372468,
      "grad_norm": 6.227043197214259,
      "learning_rate": 4.412955465587045e-06,
      "loss": 13.2656,
      "mean_token_accuracy": 0.6109288930892944,
      "num_tokens": 25764637.0,
      "step": 386
    },
    {
      "epoch": 1.5668016194331984,
      "grad_norm": 6.437168132260638,
      "learning_rate": 4.372469635627531e-06,
      "loss": 13.0156,
      "mean_token_accuracy": 0.6073063611984253,
      "num_tokens": 25828262.0,
      "step": 387
    },
    {
      "epoch": 1.5708502024291497,
      "grad_norm": 6.443176214251158,
      "learning_rate": 4.3319838056680166e-06,
      "loss": 12.5391,
      "mean_token_accuracy": 0.6221445798873901,
      "num_tokens": 25894997.0,
      "step": 388
    },
    {
      "epoch": 1.574898785425101,
      "grad_norm": 6.541270347260077,
      "learning_rate": 4.291497975708503e-06,
      "loss": 12.5625,
      "mean_token_accuracy": 0.6267231106758118,
      "num_tokens": 25956940.0,
      "step": 389
    },
    {
      "epoch": 1.5789473684210527,
      "grad_norm": 6.622129593963462,
      "learning_rate": 4.251012145748988e-06,
      "loss": 12.5,
      "mean_token_accuracy": 0.620368242263794,
      "num_tokens": 26022017.0,
      "step": 390
    },
    {
      "epoch": 1.5789473684210527,
      "eval_loss": 1.6799999475479126,
      "eval_mean_token_accuracy": 0.6024383455514908,
      "eval_num_tokens": 26022017.0,
      "eval_runtime": 0.6355,
      "eval_samples_per_second": 314.693,
      "eval_steps_per_second": 6.294,
      "step": 390
    },
    {
      "epoch": 1.582995951417004,
      "grad_norm": 5.683943779857191,
      "learning_rate": 4.210526315789474e-06,
      "loss": 12.7578,
      "mean_token_accuracy": 0.6183803677558899,
      "num_tokens": 26089381.0,
      "step": 391
    },
    {
      "epoch": 1.5870445344129553,
      "grad_norm": 6.4439629905284,
      "learning_rate": 4.170040485829959e-06,
      "loss": 13.5859,
      "mean_token_accuracy": 0.6022599339485168,
      "num_tokens": 26160421.0,
      "step": 392
    },
    {
      "epoch": 1.591093117408907,
      "grad_norm": 6.482307740702656,
      "learning_rate": 4.1295546558704455e-06,
      "loss": 13.0938,
      "mean_token_accuracy": 0.6097571849822998,
      "num_tokens": 26227164.0,
      "step": 393
    },
    {
      "epoch": 1.5951417004048583,
      "grad_norm": 5.6243349055097775,
      "learning_rate": 4.089068825910931e-06,
      "loss": 12.7969,
      "mean_token_accuracy": 0.6154806613922119,
      "num_tokens": 26297716.0,
      "step": 394
    },
    {
      "epoch": 1.5991902834008096,
      "grad_norm": 7.1839751833286725,
      "learning_rate": 4.048582995951417e-06,
      "loss": 12.5312,
      "mean_token_accuracy": 0.627181351184845,
      "num_tokens": 26358579.0,
      "step": 395
    },
    {
      "epoch": 1.5991902834008096,
      "eval_loss": 1.6775000095367432,
      "eval_mean_token_accuracy": 0.6028738915920258,
      "eval_num_tokens": 26358579.0,
      "eval_runtime": 0.6349,
      "eval_samples_per_second": 314.994,
      "eval_steps_per_second": 6.3,
      "step": 395
    },
    {
      "epoch": 1.6032388663967612,
      "grad_norm": 6.286540348376537,
      "learning_rate": 4.008097165991903e-06,
      "loss": 12.8438,
      "mean_token_accuracy": 0.618080198764801,
      "num_tokens": 26423963.0,
      "step": 396
    },
    {
      "epoch": 1.6072874493927125,
      "grad_norm": 8.094171440991735,
      "learning_rate": 3.967611336032389e-06,
      "loss": 12.6172,
      "mean_token_accuracy": 0.6203653216362,
      "num_tokens": 26473680.0,
      "step": 397
    },
    {
      "epoch": 1.6113360323886639,
      "grad_norm": 6.734637207480726,
      "learning_rate": 3.9271255060728745e-06,
      "loss": 13.1016,
      "mean_token_accuracy": 0.6167324185371399,
      "num_tokens": 26531787.0,
      "step": 398
    },
    {
      "epoch": 1.6153846153846154,
      "grad_norm": 6.248322358568084,
      "learning_rate": 3.886639676113361e-06,
      "loss": 12.0234,
      "mean_token_accuracy": 0.6334991455078125,
      "num_tokens": 26589739.0,
      "step": 399
    },
    {
      "epoch": 1.6194331983805668,
      "grad_norm": 7.003524821469698,
      "learning_rate": 3.846153846153847e-06,
      "loss": 12.5938,
      "mean_token_accuracy": 0.6244420409202576,
      "num_tokens": 26650518.0,
      "step": 400
    },
    {
      "epoch": 1.6194331983805668,
      "eval_loss": 1.6775000095367432,
      "eval_mean_token_accuracy": 0.6030516028404236,
      "eval_num_tokens": 26650518.0,
      "eval_runtime": 0.6363,
      "eval_samples_per_second": 314.294,
      "eval_steps_per_second": 6.286,
      "step": 400
    },
    {
      "epoch": 1.623481781376518,
      "grad_norm": 6.575500256754098,
      "learning_rate": 3.805668016194332e-06,
      "loss": 14.2031,
      "mean_token_accuracy": 0.5973594784736633,
      "num_tokens": 26721628.0,
      "step": 401
    },
    {
      "epoch": 1.6275303643724697,
      "grad_norm": 6.399615806731187,
      "learning_rate": 3.7651821862348182e-06,
      "loss": 13.0859,
      "mean_token_accuracy": 0.6143900752067566,
      "num_tokens": 26795479.0,
      "step": 402
    },
    {
      "epoch": 1.631578947368421,
      "grad_norm": 6.108759765220742,
      "learning_rate": 3.724696356275304e-06,
      "loss": 12.5,
      "mean_token_accuracy": 0.6269562244415283,
      "num_tokens": 26858994.0,
      "step": 403
    },
    {
      "epoch": 1.6356275303643724,
      "grad_norm": 6.378724273369664,
      "learning_rate": 3.6842105263157896e-06,
      "loss": 12.7422,
      "mean_token_accuracy": 0.6249410510063171,
      "num_tokens": 26924821.0,
      "step": 404
    },
    {
      "epoch": 1.639676113360324,
      "grad_norm": 6.284255806434736,
      "learning_rate": 3.6437246963562758e-06,
      "loss": 12.1797,
      "mean_token_accuracy": 0.6245778203010559,
      "num_tokens": 26981139.0,
      "step": 405
    },
    {
      "epoch": 1.639676113360324,
      "eval_loss": 1.6775000095367432,
      "eval_mean_token_accuracy": 0.6031172126531601,
      "eval_num_tokens": 26981139.0,
      "eval_runtime": 0.6612,
      "eval_samples_per_second": 302.482,
      "eval_steps_per_second": 6.05,
      "step": 405
    },
    {
      "epoch": 1.6437246963562753,
      "grad_norm": 6.284046381701035,
      "learning_rate": 3.6032388663967615e-06,
      "loss": 12.8438,
      "mean_token_accuracy": 0.6137465238571167,
      "num_tokens": 27036897.0,
      "step": 406
    },
    {
      "epoch": 1.6477732793522266,
      "grad_norm": 6.192144168674015,
      "learning_rate": 3.562753036437247e-06,
      "loss": 13.4453,
      "mean_token_accuracy": 0.6057220697402954,
      "num_tokens": 27113856.0,
      "step": 407
    },
    {
      "epoch": 1.6518218623481782,
      "grad_norm": 7.747379888410599,
      "learning_rate": 3.522267206477733e-06,
      "loss": 13.2812,
      "mean_token_accuracy": 0.6113630533218384,
      "num_tokens": 27187122.0,
      "step": 408
    },
    {
      "epoch": 1.6558704453441295,
      "grad_norm": 7.619435387018495,
      "learning_rate": 3.481781376518219e-06,
      "loss": 12.625,
      "mean_token_accuracy": 0.6203478574752808,
      "num_tokens": 27245197.0,
      "step": 409
    },
    {
      "epoch": 1.6599190283400809,
      "grad_norm": 6.774444594833395,
      "learning_rate": 3.4412955465587043e-06,
      "loss": 13.4766,
      "mean_token_accuracy": 0.6078824996948242,
      "num_tokens": 27310976.0,
      "step": 410
    },
    {
      "epoch": 1.6599190283400809,
      "eval_loss": 1.6768749952316284,
      "eval_mean_token_accuracy": 0.6029177159070969,
      "eval_num_tokens": 27310976.0,
      "eval_runtime": 0.652,
      "eval_samples_per_second": 306.731,
      "eval_steps_per_second": 6.135,
      "step": 410
    },
    {
      "epoch": 1.6639676113360324,
      "grad_norm": 6.546428537195409,
      "learning_rate": 3.4008097165991905e-06,
      "loss": 13.0,
      "mean_token_accuracy": 0.6162744760513306,
      "num_tokens": 27381838.0,
      "step": 411
    },
    {
      "epoch": 1.6680161943319838,
      "grad_norm": 5.661705927031424,
      "learning_rate": 3.3603238866396766e-06,
      "loss": 12.9844,
      "mean_token_accuracy": 0.6101025938987732,
      "num_tokens": 27457427.0,
      "step": 412
    },
    {
      "epoch": 1.6720647773279351,
      "grad_norm": 6.634404347857771,
      "learning_rate": 3.3198380566801623e-06,
      "loss": 13.2344,
      "mean_token_accuracy": 0.6103484034538269,
      "num_tokens": 27525598.0,
      "step": 413
    },
    {
      "epoch": 1.6761133603238867,
      "grad_norm": 6.4588054803695165,
      "learning_rate": 3.279352226720648e-06,
      "loss": 13.0859,
      "mean_token_accuracy": 0.6082247495651245,
      "num_tokens": 27586138.0,
      "step": 414
    },
    {
      "epoch": 1.680161943319838,
      "grad_norm": 6.229468320759774,
      "learning_rate": 3.2388663967611337e-06,
      "loss": 13.2578,
      "mean_token_accuracy": 0.6039714813232422,
      "num_tokens": 27656654.0,
      "step": 415
    },
    {
      "epoch": 1.680161943319838,
      "eval_loss": 1.6768749952316284,
      "eval_mean_token_accuracy": 0.603041797876358,
      "eval_num_tokens": 27656654.0,
      "eval_runtime": 0.6416,
      "eval_samples_per_second": 311.713,
      "eval_steps_per_second": 6.234,
      "step": 415
    },
    {
      "epoch": 1.6842105263157894,
      "grad_norm": 6.530706970746754,
      "learning_rate": 3.19838056680162e-06,
      "loss": 12.9609,
      "mean_token_accuracy": 0.616111159324646,
      "num_tokens": 27719395.0,
      "step": 416
    },
    {
      "epoch": 1.688259109311741,
      "grad_norm": 5.974041721296059,
      "learning_rate": 3.157894736842105e-06,
      "loss": 13.0781,
      "mean_token_accuracy": 0.6097447872161865,
      "num_tokens": 27805700.0,
      "step": 417
    },
    {
      "epoch": 1.6923076923076923,
      "grad_norm": 6.9520061051049655,
      "learning_rate": 3.1174089068825913e-06,
      "loss": 13.1328,
      "mean_token_accuracy": 0.6185556054115295,
      "num_tokens": 27870607.0,
      "step": 418
    },
    {
      "epoch": 1.6963562753036436,
      "grad_norm": 6.687485877625691,
      "learning_rate": 3.0769230769230774e-06,
      "loss": 13.125,
      "mean_token_accuracy": 0.6092858910560608,
      "num_tokens": 27935199.0,
      "step": 419
    },
    {
      "epoch": 1.7004048582995952,
      "grad_norm": 5.89529220246302,
      "learning_rate": 3.0364372469635627e-06,
      "loss": 12.7344,
      "mean_token_accuracy": 0.6122572422027588,
      "num_tokens": 27999633.0,
      "step": 420
    },
    {
      "epoch": 1.7004048582995952,
      "eval_loss": 1.6762499809265137,
      "eval_mean_token_accuracy": 0.6032748818397522,
      "eval_num_tokens": 27999633.0,
      "eval_runtime": 0.6365,
      "eval_samples_per_second": 314.198,
      "eval_steps_per_second": 6.284,
      "step": 420
    },
    {
      "epoch": 1.7044534412955465,
      "grad_norm": 5.97670859064412,
      "learning_rate": 2.995951417004049e-06,
      "loss": 13.6172,
      "mean_token_accuracy": 0.6090903878211975,
      "num_tokens": 28070629.0,
      "step": 421
    },
    {
      "epoch": 1.708502024291498,
      "grad_norm": 6.35673530964282,
      "learning_rate": 2.955465587044535e-06,
      "loss": 14.1406,
      "mean_token_accuracy": 0.5912197828292847,
      "num_tokens": 28143265.0,
      "step": 422
    },
    {
      "epoch": 1.7125506072874495,
      "grad_norm": 6.087049071022692,
      "learning_rate": 2.9149797570850203e-06,
      "loss": 12.7266,
      "mean_token_accuracy": 0.6160033345222473,
      "num_tokens": 28217426.0,
      "step": 423
    },
    {
      "epoch": 1.7165991902834008,
      "grad_norm": 6.5904480079951915,
      "learning_rate": 2.8744939271255064e-06,
      "loss": 12.9141,
      "mean_token_accuracy": 0.6187474727630615,
      "num_tokens": 28286022.0,
      "step": 424
    },
    {
      "epoch": 1.7206477732793521,
      "grad_norm": 6.195739080649372,
      "learning_rate": 2.834008097165992e-06,
      "loss": 13.5859,
      "mean_token_accuracy": 0.6009435057640076,
      "num_tokens": 28356036.0,
      "step": 425
    },
    {
      "epoch": 1.7206477732793521,
      "eval_loss": 1.6762499809265137,
      "eval_mean_token_accuracy": 0.6033691614866257,
      "eval_num_tokens": 28356036.0,
      "eval_runtime": 0.641,
      "eval_samples_per_second": 312.0,
      "eval_steps_per_second": 6.24,
      "step": 425
    },
    {
      "epoch": 1.7246963562753037,
      "grad_norm": 6.307993289907047,
      "learning_rate": 2.7935222672064783e-06,
      "loss": 13.1875,
      "mean_token_accuracy": 0.6147733926773071,
      "num_tokens": 28430355.0,
      "step": 426
    },
    {
      "epoch": 1.728744939271255,
      "grad_norm": 6.289447300448821,
      "learning_rate": 2.7530364372469636e-06,
      "loss": 12.5156,
      "mean_token_accuracy": 0.6260796785354614,
      "num_tokens": 28494906.0,
      "step": 427
    },
    {
      "epoch": 1.7327935222672064,
      "grad_norm": 6.826350614499696,
      "learning_rate": 2.7125506072874497e-06,
      "loss": 12.6094,
      "mean_token_accuracy": 0.6202298998832703,
      "num_tokens": 28562395.0,
      "step": 428
    },
    {
      "epoch": 1.736842105263158,
      "grad_norm": 6.595176524843769,
      "learning_rate": 2.672064777327936e-06,
      "loss": 12.8047,
      "mean_token_accuracy": 0.6235784292221069,
      "num_tokens": 28624889.0,
      "step": 429
    },
    {
      "epoch": 1.7408906882591093,
      "grad_norm": 5.691303853471213,
      "learning_rate": 2.631578947368421e-06,
      "loss": 13.2656,
      "mean_token_accuracy": 0.6069343686103821,
      "num_tokens": 28698788.0,
      "step": 430
    },
    {
      "epoch": 1.7408906882591093,
      "eval_loss": 1.675624966621399,
      "eval_mean_token_accuracy": 0.603339210152626,
      "eval_num_tokens": 28698788.0,
      "eval_runtime": 0.6338,
      "eval_samples_per_second": 315.581,
      "eval_steps_per_second": 6.312,
      "step": 430
    },
    {
      "epoch": 1.7449392712550607,
      "grad_norm": 5.997572345174048,
      "learning_rate": 2.5910931174089072e-06,
      "loss": 12.9844,
      "mean_token_accuracy": 0.6104946732521057,
      "num_tokens": 28773747.0,
      "step": 431
    },
    {
      "epoch": 1.7489878542510122,
      "grad_norm": 7.249965290425393,
      "learning_rate": 2.550607287449393e-06,
      "loss": 13.4297,
      "mean_token_accuracy": 0.6104491353034973,
      "num_tokens": 28842544.0,
      "step": 432
    },
    {
      "epoch": 1.7530364372469636,
      "grad_norm": 6.000414268983646,
      "learning_rate": 2.5101214574898787e-06,
      "loss": 12.5781,
      "mean_token_accuracy": 0.6204512119293213,
      "num_tokens": 28912462.0,
      "step": 433
    },
    {
      "epoch": 1.757085020242915,
      "grad_norm": 6.118904164123435,
      "learning_rate": 2.4696356275303644e-06,
      "loss": 12.7109,
      "mean_token_accuracy": 0.6210896968841553,
      "num_tokens": 28992347.0,
      "step": 434
    },
    {
      "epoch": 1.7611336032388665,
      "grad_norm": 6.1441994160988065,
      "learning_rate": 2.42914979757085e-06,
      "loss": 11.4531,
      "mean_token_accuracy": 0.6428948044776917,
      "num_tokens": 29066420.0,
      "step": 435
    },
    {
      "epoch": 1.7611336032388665,
      "eval_loss": 1.6749999523162842,
      "eval_mean_token_accuracy": 0.6034704595804214,
      "eval_num_tokens": 29066420.0,
      "eval_runtime": 0.6572,
      "eval_samples_per_second": 304.314,
      "eval_steps_per_second": 6.086,
      "step": 435
    },
    {
      "epoch": 1.7651821862348178,
      "grad_norm": 8.29981852170875,
      "learning_rate": 2.3886639676113362e-06,
      "loss": 12.7266,
      "mean_token_accuracy": 0.6216300129890442,
      "num_tokens": 29140409.0,
      "step": 436
    },
    {
      "epoch": 1.7692307692307692,
      "grad_norm": 7.051595151204688,
      "learning_rate": 2.348178137651822e-06,
      "loss": 12.9375,
      "mean_token_accuracy": 0.6193385124206543,
      "num_tokens": 29200219.0,
      "step": 437
    },
    {
      "epoch": 1.7732793522267207,
      "grad_norm": 6.304795255922526,
      "learning_rate": 2.307692307692308e-06,
      "loss": 12.4688,
      "mean_token_accuracy": 0.6236432194709778,
      "num_tokens": 29271314.0,
      "step": 438
    },
    {
      "epoch": 1.777327935222672,
      "grad_norm": 5.556856434206601,
      "learning_rate": 2.267206477732794e-06,
      "loss": 12.4609,
      "mean_token_accuracy": 0.6236319541931152,
      "num_tokens": 29342739.0,
      "step": 439
    },
    {
      "epoch": 1.7813765182186234,
      "grad_norm": 7.6181275587053445,
      "learning_rate": 2.2267206477732795e-06,
      "loss": 12.3359,
      "mean_token_accuracy": 0.6281608939170837,
      "num_tokens": 29399868.0,
      "step": 440
    },
    {
      "epoch": 1.7813765182186234,
      "eval_loss": 1.6746875047683716,
      "eval_mean_token_accuracy": 0.6035052537918091,
      "eval_num_tokens": 29399868.0,
      "eval_runtime": 0.6412,
      "eval_samples_per_second": 311.93,
      "eval_steps_per_second": 6.239,
      "step": 440
    },
    {
      "epoch": 1.785425101214575,
      "grad_norm": 7.163759421924407,
      "learning_rate": 2.1862348178137656e-06,
      "loss": 13.4844,
      "mean_token_accuracy": 0.6067489385604858,
      "num_tokens": 29471825.0,
      "step": 441
    },
    {
      "epoch": 1.7894736842105263,
      "grad_norm": 5.980379834577915,
      "learning_rate": 2.1457489878542513e-06,
      "loss": 12.6953,
      "mean_token_accuracy": 0.6232491135597229,
      "num_tokens": 29543780.0,
      "step": 442
    },
    {
      "epoch": 1.7935222672064777,
      "grad_norm": 5.658389056927855,
      "learning_rate": 2.105263157894737e-06,
      "loss": 12.9844,
      "mean_token_accuracy": 0.6110087633132935,
      "num_tokens": 29620910.0,
      "step": 443
    },
    {
      "epoch": 1.7975708502024292,
      "grad_norm": 5.632632433555104,
      "learning_rate": 2.0647773279352228e-06,
      "loss": 12.3203,
      "mean_token_accuracy": 0.6279122829437256,
      "num_tokens": 29695744.0,
      "step": 444
    },
    {
      "epoch": 1.8016194331983806,
      "grad_norm": 5.67469055605223,
      "learning_rate": 2.0242914979757085e-06,
      "loss": 12.5547,
      "mean_token_accuracy": 0.6283406615257263,
      "num_tokens": 29767874.0,
      "step": 445
    },
    {
      "epoch": 1.8016194331983806,
      "eval_loss": 1.6746875047683716,
      "eval_mean_token_accuracy": 0.6034050285816193,
      "eval_num_tokens": 29767874.0,
      "eval_runtime": 0.6422,
      "eval_samples_per_second": 311.443,
      "eval_steps_per_second": 6.229,
      "step": 445
    },
    {
      "epoch": 1.805668016194332,
      "grad_norm": 6.569962521896024,
      "learning_rate": 1.9838056680161946e-06,
      "loss": 11.7422,
      "mean_token_accuracy": 0.6383576393127441,
      "num_tokens": 29831675.0,
      "step": 446
    },
    {
      "epoch": 1.8097165991902835,
      "grad_norm": 6.86808486740946,
      "learning_rate": 1.9433198380566803e-06,
      "loss": 13.7891,
      "mean_token_accuracy": 0.5985156893730164,
      "num_tokens": 29900728.0,
      "step": 447
    },
    {
      "epoch": 1.8137651821862348,
      "grad_norm": 5.57356193397258,
      "learning_rate": 1.902834008097166e-06,
      "loss": 13.1406,
      "mean_token_accuracy": 0.6123852133750916,
      "num_tokens": 29970601.0,
      "step": 448
    },
    {
      "epoch": 1.8178137651821862,
      "grad_norm": 6.094864274620779,
      "learning_rate": 1.862348178137652e-06,
      "loss": 13.1562,
      "mean_token_accuracy": 0.613720178604126,
      "num_tokens": 30034673.0,
      "step": 449
    },
    {
      "epoch": 1.8218623481781377,
      "grad_norm": 7.163324826123694,
      "learning_rate": 1.8218623481781379e-06,
      "loss": 12.1641,
      "mean_token_accuracy": 0.6280580759048462,
      "num_tokens": 30091840.0,
      "step": 450
    },
    {
      "epoch": 1.8218623481781377,
      "eval_loss": 1.6737500429153442,
      "eval_mean_token_accuracy": 0.6033113747835159,
      "eval_num_tokens": 30091840.0,
      "eval_runtime": 0.6388,
      "eval_samples_per_second": 313.097,
      "eval_steps_per_second": 6.262,
      "step": 450
    },
    {
      "epoch": 1.825910931174089,
      "grad_norm": 5.721465202296356,
      "learning_rate": 1.7813765182186236e-06,
      "loss": 13.1172,
      "mean_token_accuracy": 0.6090490818023682,
      "num_tokens": 30160795.0,
      "step": 451
    },
    {
      "epoch": 1.8299595141700404,
      "grad_norm": 5.834496019461034,
      "learning_rate": 1.7408906882591095e-06,
      "loss": 11.6094,
      "mean_token_accuracy": 0.6369089484214783,
      "num_tokens": 30223207.0,
      "step": 452
    },
    {
      "epoch": 1.834008097165992,
      "grad_norm": 5.989944536955842,
      "learning_rate": 1.7004048582995952e-06,
      "loss": 12.8125,
      "mean_token_accuracy": 0.6168772578239441,
      "num_tokens": 30292998.0,
      "step": 453
    },
    {
      "epoch": 1.8380566801619433,
      "grad_norm": 6.175252808359519,
      "learning_rate": 1.6599190283400812e-06,
      "loss": 12.4141,
      "mean_token_accuracy": 0.6250400543212891,
      "num_tokens": 30358645.0,
      "step": 454
    },
    {
      "epoch": 1.8421052631578947,
      "grad_norm": 6.320123552822768,
      "learning_rate": 1.6194331983805669e-06,
      "loss": 12.7969,
      "mean_token_accuracy": 0.6187460422515869,
      "num_tokens": 30426126.0,
      "step": 455
    },
    {
      "epoch": 1.8421052631578947,
      "eval_loss": 1.6737500429153442,
      "eval_mean_token_accuracy": 0.6035889238119125,
      "eval_num_tokens": 30426126.0,
      "eval_runtime": 0.6462,
      "eval_samples_per_second": 309.508,
      "eval_steps_per_second": 6.19,
      "step": 455
    },
    {
      "epoch": 1.8461538461538463,
      "grad_norm": 7.054930335620368,
      "learning_rate": 1.5789473684210526e-06,
      "loss": 12.9453,
      "mean_token_accuracy": 0.6147574186325073,
      "num_tokens": 30490700.0,
      "step": 456
    },
    {
      "epoch": 1.8502024291497976,
      "grad_norm": 5.963169448840178,
      "learning_rate": 1.5384615384615387e-06,
      "loss": 12.7891,
      "mean_token_accuracy": 0.6221683025360107,
      "num_tokens": 30564085.0,
      "step": 457
    },
    {
      "epoch": 1.854251012145749,
      "grad_norm": 6.545239654505327,
      "learning_rate": 1.4979757085020244e-06,
      "loss": 13.8984,
      "mean_token_accuracy": 0.6013196110725403,
      "num_tokens": 30629773.0,
      "step": 458
    },
    {
      "epoch": 1.8582995951417005,
      "grad_norm": 7.19847490245891,
      "learning_rate": 1.4574898785425101e-06,
      "loss": 12.6953,
      "mean_token_accuracy": 0.6308658123016357,
      "num_tokens": 30685115.0,
      "step": 459
    },
    {
      "epoch": 1.8623481781376519,
      "grad_norm": 6.402339501112951,
      "learning_rate": 1.417004048582996e-06,
      "loss": 14.625,
      "mean_token_accuracy": 0.5789289474487305,
      "num_tokens": 30759499.0,
      "step": 460
    },
    {
      "epoch": 1.8623481781376519,
      "eval_loss": 1.6737500429153442,
      "eval_mean_token_accuracy": 0.6036617457866669,
      "eval_num_tokens": 30759499.0,
      "eval_runtime": 0.6471,
      "eval_samples_per_second": 309.048,
      "eval_steps_per_second": 6.181,
      "step": 460
    },
    {
      "epoch": 1.8663967611336032,
      "grad_norm": 6.554938664776423,
      "learning_rate": 1.3765182186234818e-06,
      "loss": 12.0312,
      "mean_token_accuracy": 0.6326593160629272,
      "num_tokens": 30821383.0,
      "step": 461
    },
    {
      "epoch": 1.8704453441295548,
      "grad_norm": 7.057225553782759,
      "learning_rate": 1.336032388663968e-06,
      "loss": 12.2266,
      "mean_token_accuracy": 0.6232849359512329,
      "num_tokens": 30881357.0,
      "step": 462
    },
    {
      "epoch": 1.874493927125506,
      "grad_norm": 5.899931807473347,
      "learning_rate": 1.2955465587044536e-06,
      "loss": 12.8906,
      "mean_token_accuracy": 0.6196330785751343,
      "num_tokens": 30944702.0,
      "step": 463
    },
    {
      "epoch": 1.8785425101214575,
      "grad_norm": 6.484707334873664,
      "learning_rate": 1.2550607287449393e-06,
      "loss": 12.6484,
      "mean_token_accuracy": 0.6225480437278748,
      "num_tokens": 31018431.0,
      "step": 464
    },
    {
      "epoch": 1.882591093117409,
      "grad_norm": 6.294801781620055,
      "learning_rate": 1.214574898785425e-06,
      "loss": 12.8203,
      "mean_token_accuracy": 0.6174430847167969,
      "num_tokens": 31081947.0,
      "step": 465
    },
    {
      "epoch": 1.882591093117409,
      "eval_loss": 1.6731250286102295,
      "eval_mean_token_accuracy": 0.6035991758108139,
      "eval_num_tokens": 31081947.0,
      "eval_runtime": 0.6369,
      "eval_samples_per_second": 314.028,
      "eval_steps_per_second": 6.281,
      "step": 465
    },
    {
      "epoch": 1.8866396761133604,
      "grad_norm": 7.0152791422709795,
      "learning_rate": 1.174089068825911e-06,
      "loss": 13.6797,
      "mean_token_accuracy": 0.5981454849243164,
      "num_tokens": 31147796.0,
      "step": 466
    },
    {
      "epoch": 1.8906882591093117,
      "grad_norm": 6.303822986663203,
      "learning_rate": 1.133603238866397e-06,
      "loss": 13.2266,
      "mean_token_accuracy": 0.6078663468360901,
      "num_tokens": 31212134.0,
      "step": 467
    },
    {
      "epoch": 1.8947368421052633,
      "grad_norm": 5.882937931956007,
      "learning_rate": 1.0931174089068828e-06,
      "loss": 13.8359,
      "mean_token_accuracy": 0.6062960624694824,
      "num_tokens": 31282591.0,
      "step": 468
    },
    {
      "epoch": 1.8987854251012146,
      "grad_norm": 5.937096953516108,
      "learning_rate": 1.0526315789473685e-06,
      "loss": 12.3516,
      "mean_token_accuracy": 0.6282551884651184,
      "num_tokens": 31354656.0,
      "step": 469
    },
    {
      "epoch": 1.902834008097166,
      "grad_norm": 6.3577908169539255,
      "learning_rate": 1.0121457489878542e-06,
      "loss": 12.9531,
      "mean_token_accuracy": 0.6160109043121338,
      "num_tokens": 31420675.0,
      "step": 470
    },
    {
      "epoch": 1.902834008097166,
      "eval_loss": 1.673437476158142,
      "eval_mean_token_accuracy": 0.6037902683019638,
      "eval_num_tokens": 31420675.0,
      "eval_runtime": 0.6715,
      "eval_samples_per_second": 297.827,
      "eval_steps_per_second": 5.957,
      "step": 470
    },
    {
      "epoch": 1.9068825910931175,
      "grad_norm": 6.62892566578751,
      "learning_rate": 9.716599190283402e-07,
      "loss": 13.2188,
      "mean_token_accuracy": 0.6141934990882874,
      "num_tokens": 31486290.0,
      "step": 471
    },
    {
      "epoch": 1.9109311740890689,
      "grad_norm": 6.1078260873922625,
      "learning_rate": 9.31174089068826e-07,
      "loss": 12.6641,
      "mean_token_accuracy": 0.6177834868431091,
      "num_tokens": 31557060.0,
      "step": 472
    },
    {
      "epoch": 1.9149797570850202,
      "grad_norm": 6.925934956753748,
      "learning_rate": 8.906882591093118e-07,
      "loss": 13.0938,
      "mean_token_accuracy": 0.613021969795227,
      "num_tokens": 31628035.0,
      "step": 473
    },
    {
      "epoch": 1.9190283400809718,
      "grad_norm": 5.645167673990774,
      "learning_rate": 8.502024291497976e-07,
      "loss": 13.0625,
      "mean_token_accuracy": 0.6135348677635193,
      "num_tokens": 31704450.0,
      "step": 474
    },
    {
      "epoch": 1.9230769230769231,
      "grad_norm": 6.3104167802865385,
      "learning_rate": 8.097165991902834e-07,
      "loss": 12.6016,
      "mean_token_accuracy": 0.6247060298919678,
      "num_tokens": 31771696.0,
      "step": 475
    },
    {
      "epoch": 1.9230769230769231,
      "eval_loss": 1.6740624904632568,
      "eval_mean_token_accuracy": 0.6038688272237778,
      "eval_num_tokens": 31771696.0,
      "eval_runtime": 0.6717,
      "eval_samples_per_second": 297.737,
      "eval_steps_per_second": 5.955,
      "step": 475
    },
    {
      "epoch": 1.9271255060728745,
      "grad_norm": 8.75546525050239,
      "learning_rate": 7.692307692307694e-07,
      "loss": 14.3125,
      "mean_token_accuracy": 0.6006154417991638,
      "num_tokens": 31833828.0,
      "step": 476
    },
    {
      "epoch": 1.931174089068826,
      "grad_norm": 6.568540834125108,
      "learning_rate": 7.287449392712551e-07,
      "loss": 12.4688,
      "mean_token_accuracy": 0.6260876655578613,
      "num_tokens": 31908019.0,
      "step": 477
    },
    {
      "epoch": 1.9352226720647774,
      "grad_norm": 5.992577753842725,
      "learning_rate": 6.882591093117409e-07,
      "loss": 11.6328,
      "mean_token_accuracy": 0.64389967918396,
      "num_tokens": 31968785.0,
      "step": 478
    },
    {
      "epoch": 1.9392712550607287,
      "grad_norm": 6.737090705162421,
      "learning_rate": 6.477732793522268e-07,
      "loss": 12.6562,
      "mean_token_accuracy": 0.629744291305542,
      "num_tokens": 32032478.0,
      "step": 479
    },
    {
      "epoch": 1.9433198380566803,
      "grad_norm": 6.273424964829595,
      "learning_rate": 6.072874493927125e-07,
      "loss": 12.1953,
      "mean_token_accuracy": 0.6316378116607666,
      "num_tokens": 32096023.0,
      "step": 480
    },
    {
      "epoch": 1.9433198380566803,
      "eval_loss": 1.6731250286102295,
      "eval_mean_token_accuracy": 0.6038562655448914,
      "eval_num_tokens": 32096023.0,
      "eval_runtime": 0.637,
      "eval_samples_per_second": 313.985,
      "eval_steps_per_second": 6.28,
      "step": 480
    },
    {
      "epoch": 1.9473684210526314,
      "grad_norm": 6.127164798465686,
      "learning_rate": 5.668016194331984e-07,
      "loss": 12.1719,
      "mean_token_accuracy": 0.6262784600257874,
      "num_tokens": 32164919.0,
      "step": 481
    },
    {
      "epoch": 1.951417004048583,
      "grad_norm": 5.756695804667141,
      "learning_rate": 5.263157894736843e-07,
      "loss": 12.6875,
      "mean_token_accuracy": 0.6200114488601685,
      "num_tokens": 32233114.0,
      "step": 482
    },
    {
      "epoch": 1.9554655870445345,
      "grad_norm": 6.564768636864393,
      "learning_rate": 4.858299595141701e-07,
      "loss": 13.1953,
      "mean_token_accuracy": 0.6086875796318054,
      "num_tokens": 32303117.0,
      "step": 483
    },
    {
      "epoch": 1.9595141700404857,
      "grad_norm": 7.252534772707513,
      "learning_rate": 4.453441295546559e-07,
      "loss": 13.25,
      "mean_token_accuracy": 0.6089531779289246,
      "num_tokens": 32373927.0,
      "step": 484
    },
    {
      "epoch": 1.9635627530364372,
      "grad_norm": 6.803603694268994,
      "learning_rate": 4.048582995951417e-07,
      "loss": 12.4609,
      "mean_token_accuracy": 0.624879777431488,
      "num_tokens": 32444699.0,
      "step": 485
    },
    {
      "epoch": 1.9635627530364372,
      "eval_loss": 1.6731250286102295,
      "eval_mean_token_accuracy": 0.6039727181196213,
      "eval_num_tokens": 32444699.0,
      "eval_runtime": 0.6475,
      "eval_samples_per_second": 308.869,
      "eval_steps_per_second": 6.177,
      "step": 485
    },
    {
      "epoch": 1.9676113360323888,
      "grad_norm": 6.049306687892726,
      "learning_rate": 3.6437246963562754e-07,
      "loss": 14.0625,
      "mean_token_accuracy": 0.593756914138794,
      "num_tokens": 32516907.0,
      "step": 486
    },
    {
      "epoch": 1.97165991902834,
      "grad_norm": 7.270949297430324,
      "learning_rate": 3.238866396761134e-07,
      "loss": 12.6406,
      "mean_token_accuracy": 0.6211447715759277,
      "num_tokens": 32583342.0,
      "step": 487
    },
    {
      "epoch": 1.9757085020242915,
      "grad_norm": 8.038406126817675,
      "learning_rate": 2.834008097165992e-07,
      "loss": 12.9766,
      "mean_token_accuracy": 0.6204248666763306,
      "num_tokens": 32639142.0,
      "step": 488
    },
    {
      "epoch": 1.979757085020243,
      "grad_norm": 6.211130232502858,
      "learning_rate": 2.4291497975708504e-07,
      "loss": 12.1953,
      "mean_token_accuracy": 0.6283571124076843,
      "num_tokens": 32705595.0,
      "step": 489
    },
    {
      "epoch": 1.9838056680161942,
      "grad_norm": 6.72573911112471,
      "learning_rate": 2.0242914979757086e-07,
      "loss": 13.8281,
      "mean_token_accuracy": 0.5990269184112549,
      "num_tokens": 32771222.0,
      "step": 490
    },
    {
      "epoch": 1.9838056680161942,
      "eval_loss": 1.6731250286102295,
      "eval_mean_token_accuracy": 0.6039326936006546,
      "eval_num_tokens": 32771222.0,
      "eval_runtime": 0.639,
      "eval_samples_per_second": 312.997,
      "eval_steps_per_second": 6.26,
      "step": 490
    },
    {
      "epoch": 1.9878542510121457,
      "grad_norm": 6.616901783015977,
      "learning_rate": 1.619433198380567e-07,
      "loss": 12.0078,
      "mean_token_accuracy": 0.6346266865730286,
      "num_tokens": 32827549.0,
      "step": 491
    },
    {
      "epoch": 1.9919028340080973,
      "grad_norm": 6.065995562899032,
      "learning_rate": 1.2145748987854252e-07,
      "loss": 13.1484,
      "mean_token_accuracy": 0.6112900972366333,
      "num_tokens": 32900562.0,
      "step": 492
    },
    {
      "epoch": 1.9959514170040484,
      "grad_norm": 6.877917660573328,
      "learning_rate": 8.097165991902835e-08,
      "loss": 13.1406,
      "mean_token_accuracy": 0.6131631135940552,
      "num_tokens": 32967282.0,
      "step": 493
    },
    {
      "epoch": 2.0,
      "grad_norm": 6.297817735745015,
      "learning_rate": 4.0485829959514176e-08,
      "loss": 12.4141,
      "mean_token_accuracy": 0.6238685250282288,
      "num_tokens": 33035067.0,
      "step": 494
    },
    {
      "epoch": 2.0,
      "step": 494,
      "total_flos": 39806562140160.0,
      "train_loss": 13.738502656882591,
      "train_runtime": 467.8481,
      "train_samples_per_second": 67.569,
      "train_steps_per_second": 1.056
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 494,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 39806562140160.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}