{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 510,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00984009840098401,
      "grad_norm": 3.7944442389822073,
      "learning_rate": 0.0,
      "loss": 1.2501,
      "num_tokens": 456505.0,
      "step": 1
    },
    {
      "epoch": 0.01968019680196802,
      "grad_norm": 3.778041640972742,
      "learning_rate": 6.25e-07,
      "loss": 1.2343,
      "num_tokens": 915160.0,
      "step": 2
    },
    {
      "epoch": 0.02952029520295203,
      "grad_norm": 3.8325814879608386,
      "learning_rate": 1.25e-06,
      "loss": 1.254,
      "num_tokens": 1365315.0,
      "step": 3
    },
    {
      "epoch": 0.03936039360393604,
      "grad_norm": 3.582565733113683,
      "learning_rate": 1.8750000000000003e-06,
      "loss": 1.1869,
      "num_tokens": 1841763.0,
      "step": 4
    },
    {
      "epoch": 0.04920049200492005,
      "grad_norm": 3.5604969753172315,
      "learning_rate": 2.5e-06,
      "loss": 1.2394,
      "num_tokens": 2301606.0,
      "step": 5
    },
    {
      "epoch": 0.05904059040590406,
      "grad_norm": 3.105374395878177,
      "learning_rate": 3.125e-06,
      "loss": 1.2366,
      "num_tokens": 2755825.0,
      "step": 6
    },
    {
      "epoch": 0.06888068880688807,
      "grad_norm": 2.316426838717515,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 1.1101,
      "num_tokens": 3196409.0,
      "step": 7
    },
    {
      "epoch": 0.07872078720787208,
      "grad_norm": 2.281060366927676,
      "learning_rate": 4.3750000000000005e-06,
      "loss": 1.1078,
      "num_tokens": 3622733.0,
      "step": 8
    },
    {
      "epoch": 0.08856088560885608,
      "grad_norm": 1.934577354985982,
      "learning_rate": 5e-06,
      "loss": 0.9014,
      "num_tokens": 4055914.0,
      "step": 9
    },
    {
      "epoch": 0.0984009840098401,
      "grad_norm": 1.9176079459138344,
      "learning_rate": 5.625e-06,
      "loss": 0.8745,
      "num_tokens": 4485159.0,
      "step": 10
    },
    {
      "epoch": 0.10824108241082411,
      "grad_norm": 1.786754010375736,
      "learning_rate": 6.25e-06,
      "loss": 0.7922,
      "num_tokens": 4933514.0,
      "step": 11
    },
    {
      "epoch": 0.11808118081180811,
      "grad_norm": 1.9933574737759214,
      "learning_rate": 6.875e-06,
      "loss": 0.4878,
      "num_tokens": 5383658.0,
      "step": 12
    },
    {
      "epoch": 0.12792127921279212,
      "grad_norm": 2.123289554302906,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.428,
      "num_tokens": 5839838.0,
      "step": 13
    },
    {
      "epoch": 0.13776137761377613,
      "grad_norm": 1.7562448014521572,
      "learning_rate": 8.125000000000001e-06,
      "loss": 0.3337,
      "num_tokens": 6286175.0,
      "step": 14
    },
    {
      "epoch": 0.14760147601476015,
      "grad_norm": 1.4384357290512548,
      "learning_rate": 8.750000000000001e-06,
      "loss": 0.2497,
      "num_tokens": 6725821.0,
      "step": 15
    },
    {
      "epoch": 0.15744157441574416,
      "grad_norm": 0.8232923354453182,
      "learning_rate": 9.375000000000001e-06,
      "loss": 0.1317,
      "num_tokens": 7169854.0,
      "step": 16
    },
    {
      "epoch": 0.16728167281672818,
      "grad_norm": 0.5262014955492348,
      "learning_rate": 1e-05,
      "loss": 0.1158,
      "num_tokens": 7602261.0,
      "step": 17
    },
    {
      "epoch": 0.17712177121771217,
      "grad_norm": 0.47218735378476806,
      "learning_rate": 9.999909003036192e-06,
      "loss": 0.098,
      "num_tokens": 8040457.0,
      "step": 18
    },
    {
      "epoch": 0.18696186961869618,
      "grad_norm": 0.22950756131023575,
      "learning_rate": 9.99963601582496e-06,
      "loss": 0.0827,
      "num_tokens": 8484210.0,
      "step": 19
    },
    {
      "epoch": 0.1968019680196802,
      "grad_norm": 0.24069667769460337,
      "learning_rate": 9.999181049406756e-06,
      "loss": 0.0733,
      "num_tokens": 8913622.0,
      "step": 20
    },
    {
      "epoch": 0.2066420664206642,
      "grad_norm": 0.19079800364724872,
      "learning_rate": 9.998544122181829e-06,
      "loss": 0.0851,
      "num_tokens": 9379389.0,
      "step": 21
    },
    {
      "epoch": 0.21648216482164823,
      "grad_norm": 0.17418458793254618,
      "learning_rate": 9.997725259909487e-06,
      "loss": 0.0734,
      "num_tokens": 9803100.0,
      "step": 22
    },
    {
      "epoch": 0.22632226322263221,
      "grad_norm": 0.19184343474298712,
      "learning_rate": 9.996724495707056e-06,
      "loss": 0.083,
      "num_tokens": 10247767.0,
      "step": 23
    },
    {
      "epoch": 0.23616236162361623,
      "grad_norm": 0.1475443750251538,
      "learning_rate": 9.995541870048537e-06,
      "loss": 0.0738,
      "num_tokens": 10691516.0,
      "step": 24
    },
    {
      "epoch": 0.24600246002460024,
      "grad_norm": 0.14419500643952865,
      "learning_rate": 9.994177430762971e-06,
      "loss": 0.0646,
      "num_tokens": 11149524.0,
      "step": 25
    },
    {
      "epoch": 0.25584255842558423,
      "grad_norm": 1.2190794556868674,
      "learning_rate": 9.992631233032507e-06,
      "loss": 0.0721,
      "num_tokens": 11589958.0,
      "step": 26
    },
    {
      "epoch": 0.2656826568265683,
      "grad_norm": 0.49893871677873436,
      "learning_rate": 9.990903339390164e-06,
      "loss": 0.0691,
      "num_tokens": 12050102.0,
      "step": 27
    },
    {
      "epoch": 0.27552275522755226,
      "grad_norm": 0.4422022706582718,
      "learning_rate": 9.988993819717312e-06,
      "loss": 0.0605,
      "num_tokens": 12508433.0,
      "step": 28
    },
    {
      "epoch": 0.2853628536285363,
      "grad_norm": 0.13441092661822238,
      "learning_rate": 9.986902751240836e-06,
      "loss": 0.0692,
      "num_tokens": 12939960.0,
      "step": 29
    },
    {
      "epoch": 0.2952029520295203,
      "grad_norm": 0.11858771432621444,
      "learning_rate": 9.984630218530014e-06,
      "loss": 0.0492,
      "num_tokens": 13387850.0,
      "step": 30
    },
    {
      "epoch": 0.3050430504305043,
      "grad_norm": 0.12696361103470127,
      "learning_rate": 9.982176313493108e-06,
      "loss": 0.0624,
      "num_tokens": 13866894.0,
      "step": 31
    },
    {
      "epoch": 0.3148831488314883,
      "grad_norm": 0.14678245947256616,
      "learning_rate": 9.979541135373628e-06,
      "loss": 0.0483,
      "num_tokens": 14314553.0,
      "step": 32
    },
    {
      "epoch": 0.3247232472324723,
      "grad_norm": 0.12403518011628543,
      "learning_rate": 9.976724790746333e-06,
      "loss": 0.0488,
      "num_tokens": 14747330.0,
      "step": 33
    },
    {
      "epoch": 0.33456334563345635,
      "grad_norm": 0.10283847292091904,
      "learning_rate": 9.973727393512921e-06,
      "loss": 0.0582,
      "num_tokens": 15215873.0,
      "step": 34
    },
    {
      "epoch": 0.34440344403444034,
      "grad_norm": 0.09860235260078455,
      "learning_rate": 9.970549064897407e-06,
      "loss": 0.0446,
      "num_tokens": 15653849.0,
      "step": 35
    },
    {
      "epoch": 0.35424354243542433,
      "grad_norm": 0.10274919661024226,
      "learning_rate": 9.967189933441243e-06,
      "loss": 0.0439,
      "num_tokens": 16112913.0,
      "step": 36
    },
    {
      "epoch": 0.3640836408364084,
      "grad_norm": 0.0918843632134462,
      "learning_rate": 9.9636501349981e-06,
      "loss": 0.0585,
      "num_tokens": 16570588.0,
      "step": 37
    },
    {
      "epoch": 0.37392373923739236,
      "grad_norm": 0.08618611894284056,
      "learning_rate": 9.95992981272838e-06,
      "loss": 0.0477,
      "num_tokens": 17028395.0,
      "step": 38
    },
    {
      "epoch": 0.3837638376383764,
      "grad_norm": 0.0915069403325355,
      "learning_rate": 9.956029117093432e-06,
      "loss": 0.045,
      "num_tokens": 17477681.0,
      "step": 39
    },
    {
      "epoch": 0.3936039360393604,
      "grad_norm": 0.09093140650787605,
      "learning_rate": 9.951948205849457e-06,
      "loss": 0.0444,
      "num_tokens": 17940049.0,
      "step": 40
    },
    {
      "epoch": 0.4034440344403444,
      "grad_norm": 0.08271507354884283,
      "learning_rate": 9.947687244041143e-06,
      "loss": 0.0401,
      "num_tokens": 18360868.0,
      "step": 41
    },
    {
      "epoch": 0.4132841328413284,
      "grad_norm": 0.08588968137159211,
      "learning_rate": 9.943246403994969e-06,
      "loss": 0.0358,
      "num_tokens": 18811281.0,
      "step": 42
    },
    {
      "epoch": 0.4231242312423124,
      "grad_norm": 0.08965565515357603,
      "learning_rate": 9.938625865312252e-06,
      "loss": 0.044,
      "num_tokens": 19236998.0,
      "step": 43
    },
    {
      "epoch": 0.43296432964329645,
      "grad_norm": 0.09636661290222473,
      "learning_rate": 9.933825814861877e-06,
      "loss": 0.0431,
      "num_tokens": 19689363.0,
      "step": 44
    },
    {
      "epoch": 0.44280442804428044,
      "grad_norm": 0.08912873391359938,
      "learning_rate": 9.928846446772737e-06,
      "loss": 0.0377,
      "num_tokens": 20129602.0,
      "step": 45
    },
    {
      "epoch": 0.45264452644526443,
      "grad_norm": 0.09271503002492597,
      "learning_rate": 9.923687962425895e-06,
      "loss": 0.0365,
      "num_tokens": 20566055.0,
      "step": 46
    },
    {
      "epoch": 0.46248462484624847,
      "grad_norm": 0.08617267288782972,
      "learning_rate": 9.91835057044642e-06,
      "loss": 0.0582,
      "num_tokens": 21035837.0,
      "step": 47
    },
    {
      "epoch": 0.47232472324723246,
      "grad_norm": 0.07942181409157618,
      "learning_rate": 9.912834486694963e-06,
      "loss": 0.0341,
      "num_tokens": 21490681.0,
      "step": 48
    },
    {
      "epoch": 0.4821648216482165,
      "grad_norm": 0.08409285833406879,
      "learning_rate": 9.907139934259025e-06,
      "loss": 0.0464,
      "num_tokens": 21949736.0,
      "step": 49
    },
    {
      "epoch": 0.4920049200492005,
      "grad_norm": 0.08981746101624732,
      "learning_rate": 9.90126714344393e-06,
      "loss": 0.0479,
      "num_tokens": 22408345.0,
      "step": 50
    },
    {
      "epoch": 0.5018450184501845,
      "grad_norm": 0.08557538109120558,
      "learning_rate": 9.895216351763515e-06,
      "loss": 0.04,
      "num_tokens": 22869507.0,
      "step": 51
    },
    {
      "epoch": 0.5116851168511685,
      "grad_norm": 0.08873060518107122,
      "learning_rate": 9.888987803930523e-06,
      "loss": 0.0359,
      "num_tokens": 23337492.0,
      "step": 52
    },
    {
      "epoch": 0.5215252152521526,
      "grad_norm": 0.08508195964995854,
      "learning_rate": 9.882581751846707e-06,
      "loss": 0.0338,
      "num_tokens": 23788038.0,
      "step": 53
    },
    {
      "epoch": 0.5313653136531366,
      "grad_norm": 0.076418318161816,
      "learning_rate": 9.87599845459264e-06,
      "loss": 0.0344,
      "num_tokens": 24233994.0,
      "step": 54
    },
    {
      "epoch": 0.5412054120541205,
      "grad_norm": 0.2889818789713905,
      "learning_rate": 9.869238178417235e-06,
      "loss": 0.2599,
      "num_tokens": 24697351.0,
      "step": 55
    },
    {
      "epoch": 0.5510455104551045,
      "grad_norm": 0.08884780995830746,
      "learning_rate": 9.862301196726988e-06,
      "loss": 0.0465,
      "num_tokens": 25183095.0,
      "step": 56
    },
    {
      "epoch": 0.5608856088560885,
      "grad_norm": 0.07990815808329678,
      "learning_rate": 9.855187790074906e-06,
      "loss": 0.0353,
      "num_tokens": 25651971.0,
      "step": 57
    },
    {
      "epoch": 0.5707257072570726,
      "grad_norm": 0.06894407429892842,
      "learning_rate": 9.847898246149173e-06,
      "loss": 0.0316,
      "num_tokens": 26129683.0,
      "step": 58
    },
    {
      "epoch": 0.5805658056580566,
      "grad_norm": 0.08216971413705307,
      "learning_rate": 9.840432859761504e-06,
      "loss": 0.0306,
      "num_tokens": 26548348.0,
      "step": 59
    },
    {
      "epoch": 0.5904059040590406,
      "grad_norm": 0.079031679127037,
      "learning_rate": 9.832791932835232e-06,
      "loss": 0.0362,
      "num_tokens": 26977631.0,
      "step": 60
    },
    {
      "epoch": 0.6002460024600246,
      "grad_norm": 0.07450412090133855,
      "learning_rate": 9.824975774393089e-06,
      "loss": 0.0276,
      "num_tokens": 27421323.0,
      "step": 61
    },
    {
      "epoch": 0.6100861008610086,
      "grad_norm": 0.08014735253624648,
      "learning_rate": 9.816984700544714e-06,
      "loss": 0.0286,
      "num_tokens": 27882356.0,
      "step": 62
    },
    {
      "epoch": 0.6199261992619927,
      "grad_norm": 0.08455294660438158,
      "learning_rate": 9.808819034473869e-06,
      "loss": 0.0407,
      "num_tokens": 28343854.0,
      "step": 63
    },
    {
      "epoch": 0.6297662976629766,
      "grad_norm": 0.08019778537515825,
      "learning_rate": 9.800479106425356e-06,
      "loss": 0.0299,
      "num_tokens": 28790695.0,
      "step": 64
    },
    {
      "epoch": 0.6396063960639606,
      "grad_norm": 0.08340888167507048,
      "learning_rate": 9.791965253691687e-06,
      "loss": 0.0353,
      "num_tokens": 29220825.0,
      "step": 65
    },
    {
      "epoch": 0.6494464944649446,
      "grad_norm": 0.08252486402936965,
      "learning_rate": 9.783277820599408e-06,
      "loss": 0.0367,
      "num_tokens": 29686358.0,
      "step": 66
    },
    {
      "epoch": 0.6592865928659286,
      "grad_norm": 0.08632773276059842,
      "learning_rate": 9.774417158495208e-06,
      "loss": 0.0331,
      "num_tokens": 30120521.0,
      "step": 67
    },
    {
      "epoch": 0.6691266912669127,
      "grad_norm": 0.082343171890358,
      "learning_rate": 9.765383625731683e-06,
      "loss": 0.0329,
      "num_tokens": 30573947.0,
      "step": 68
    },
    {
      "epoch": 0.6789667896678967,
      "grad_norm": 0.08874468637210653,
      "learning_rate": 9.756177587652857e-06,
      "loss": 0.0329,
      "num_tokens": 30999244.0,
      "step": 69
    },
    {
      "epoch": 0.6888068880688807,
      "grad_norm": 0.07673402020991506,
      "learning_rate": 9.746799416579403e-06,
      "loss": 0.0306,
      "num_tokens": 31468786.0,
      "step": 70
    },
    {
      "epoch": 0.6986469864698647,
      "grad_norm": 0.09204922624438575,
      "learning_rate": 9.737249491793587e-06,
      "loss": 0.0273,
      "num_tokens": 31905019.0,
      "step": 71
    },
    {
      "epoch": 0.7084870848708487,
      "grad_norm": 0.08145687118724444,
      "learning_rate": 9.727528199523923e-06,
      "loss": 0.029,
      "num_tokens": 32340154.0,
      "step": 72
    },
    {
      "epoch": 0.7183271832718328,
      "grad_norm": 0.09506872052374568,
      "learning_rate": 9.717635932929556e-06,
      "loss": 0.0373,
      "num_tokens": 32789598.0,
      "step": 73
    },
    {
      "epoch": 0.7281672816728167,
      "grad_norm": 0.08326889230017241,
      "learning_rate": 9.707573092084368e-06,
      "loss": 0.0286,
      "num_tokens": 33239225.0,
      "step": 74
    },
    {
      "epoch": 0.7380073800738007,
      "grad_norm": 0.07636964575035168,
      "learning_rate": 9.697340083960785e-06,
      "loss": 0.0291,
      "num_tokens": 33718797.0,
      "step": 75
    },
    {
      "epoch": 0.7478474784747847,
      "grad_norm": 0.09488168094776525,
      "learning_rate": 9.686937322413325e-06,
      "loss": 0.0328,
      "num_tokens": 34155674.0,
      "step": 76
    },
    {
      "epoch": 0.7576875768757687,
      "grad_norm": 0.0778086138359463,
      "learning_rate": 9.676365228161869e-06,
      "loss": 0.0252,
      "num_tokens": 34584921.0,
      "step": 77
    },
    {
      "epoch": 0.7675276752767528,
      "grad_norm": 0.08557737550120906,
      "learning_rate": 9.66562422877462e-06,
      "loss": 0.0338,
      "num_tokens": 35049146.0,
      "step": 78
    },
    {
      "epoch": 0.7773677736777368,
      "grad_norm": 0.09181023650151289,
      "learning_rate": 9.654714758650844e-06,
      "loss": 0.0299,
      "num_tokens": 35519987.0,
      "step": 79
    },
    {
      "epoch": 0.7872078720787208,
      "grad_norm": 0.07639914292637208,
      "learning_rate": 9.643637259003276e-06,
      "loss": 0.0242,
      "num_tokens": 35959127.0,
      "step": 80
    },
    {
      "epoch": 0.7970479704797048,
      "grad_norm": 0.08200922089613671,
      "learning_rate": 9.632392177840286e-06,
      "loss": 0.0317,
      "num_tokens": 36416651.0,
      "step": 81
    },
    {
      "epoch": 0.8068880688806888,
      "grad_norm": 0.07954028434263948,
      "learning_rate": 9.620979969947759e-06,
      "loss": 0.0293,
      "num_tokens": 36864154.0,
      "step": 82
    },
    {
      "epoch": 0.8167281672816729,
      "grad_norm": 0.07878375949867687,
      "learning_rate": 9.609401096870707e-06,
      "loss": 0.0237,
      "num_tokens": 37310281.0,
      "step": 83
    },
    {
      "epoch": 0.8265682656826568,
      "grad_norm": 0.07728168843840597,
      "learning_rate": 9.597656026894591e-06,
      "loss": 0.0322,
      "num_tokens": 37746606.0,
      "step": 84
    },
    {
      "epoch": 0.8364083640836408,
      "grad_norm": 0.07855221188672869,
      "learning_rate": 9.585745235026391e-06,
      "loss": 0.0258,
      "num_tokens": 38189615.0,
      "step": 85
    },
    {
      "epoch": 0.8462484624846248,
      "grad_norm": 0.07691630967258262,
      "learning_rate": 9.5736692029754e-06,
      "loss": 0.0293,
      "num_tokens": 38637318.0,
      "step": 86
    },
    {
      "epoch": 0.8560885608856088,
      "grad_norm": 0.07209047793755496,
      "learning_rate": 9.561428419133723e-06,
      "loss": 0.0235,
      "num_tokens": 39102853.0,
      "step": 87
    },
    {
      "epoch": 0.8659286592865929,
      "grad_norm": 0.0802072339239599,
      "learning_rate": 9.549023378556548e-06,
      "loss": 0.0311,
      "num_tokens": 39538535.0,
      "step": 88
    },
    {
      "epoch": 0.8757687576875769,
      "grad_norm": 0.09334524313401625,
      "learning_rate": 9.53645458294211e-06,
      "loss": 0.0484,
      "num_tokens": 40020296.0,
      "step": 89
    },
    {
      "epoch": 0.8856088560885609,
      "grad_norm": 2.3961901610996605,
      "learning_rate": 9.523722540611403e-06,
      "loss": 0.3276,
      "num_tokens": 40506093.0,
      "step": 90
    },
    {
      "epoch": 0.8954489544895449,
      "grad_norm": 0.09376957957757263,
      "learning_rate": 9.510827766487625e-06,
      "loss": 0.0288,
      "num_tokens": 40937880.0,
      "step": 91
    },
    {
      "epoch": 0.9052890528905289,
      "grad_norm": 0.08607984794603309,
      "learning_rate": 9.497770782075353e-06,
      "loss": 0.0247,
      "num_tokens": 41374337.0,
      "step": 92
    },
    {
      "epoch": 0.915129151291513,
      "grad_norm": 0.07253858203781333,
      "learning_rate": 9.484552115439445e-06,
      "loss": 0.0293,
      "num_tokens": 41811558.0,
      "step": 93
    },
    {
      "epoch": 0.9249692496924969,
      "grad_norm": 0.07768364358007782,
      "learning_rate": 9.471172301183695e-06,
      "loss": 0.0257,
      "num_tokens": 42259726.0,
      "step": 94
    },
    {
      "epoch": 0.9348093480934809,
      "grad_norm": 0.0769153663260077,
      "learning_rate": 9.4576318804292e-06,
      "loss": 0.0232,
      "num_tokens": 42684319.0,
      "step": 95
    },
    {
      "epoch": 0.9446494464944649,
      "grad_norm": 0.08163342042509363,
      "learning_rate": 9.443931400792486e-06,
      "loss": 0.0256,
      "num_tokens": 43113589.0,
      "step": 96
    },
    {
      "epoch": 0.9544895448954489,
      "grad_norm": 0.06536764982172343,
      "learning_rate": 9.430071416363352e-06,
      "loss": 0.0218,
      "num_tokens": 43575488.0,
      "step": 97
    },
    {
      "epoch": 0.964329643296433,
      "grad_norm": 0.08195099679978833,
      "learning_rate": 9.416052487682465e-06,
      "loss": 0.0254,
      "num_tokens": 44016216.0,
      "step": 98
    },
    {
      "epoch": 0.974169741697417,
      "grad_norm": 0.1266005657397246,
      "learning_rate": 9.401875181718686e-06,
      "loss": 0.0454,
      "num_tokens": 44497742.0,
      "step": 99
    },
    {
      "epoch": 0.984009840098401,
      "grad_norm": 0.07988798247506342,
      "learning_rate": 9.387540071846155e-06,
      "loss": 0.024,
      "num_tokens": 44935936.0,
      "step": 100
    },
    {
      "epoch": 0.993849938499385,
      "grad_norm": 0.07277763654694067,
      "learning_rate": 9.373047737821078e-06,
      "loss": 0.0216,
      "num_tokens": 45381042.0,
      "step": 101
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.07277763654694067,
      "learning_rate": 9.358398765758296e-06,
      "loss": 0.0229,
      "num_tokens": 45593876.0,
      "step": 102
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.07811997085809708,
      "eval_num_tokens": 45593876.0,
      "eval_runtime": 54.709,
      "eval_samples_per_second": 41.054,
      "eval_steps_per_second": 5.136,
      "step": 102
    },
    {
      "epoch": 1.009840098400984,
      "grad_norm": 0.10795878798324991,
      "learning_rate": 9.34359374810758e-06,
      "loss": 0.0201,
      "num_tokens": 46020335.0,
      "step": 103
    },
    {
      "epoch": 1.019680196801968,
      "grad_norm": 0.07593949135329942,
      "learning_rate": 9.328633283629666e-06,
      "loss": 0.0222,
      "num_tokens": 46466853.0,
      "step": 104
    },
    {
      "epoch": 1.029520295202952,
      "grad_norm": 0.07596980345063492,
      "learning_rate": 9.31351797737204e-06,
      "loss": 0.0253,
      "num_tokens": 46900993.0,
      "step": 105
    },
    {
      "epoch": 1.039360393603936,
      "grad_norm": 0.08317964089954727,
      "learning_rate": 9.29824844064447e-06,
      "loss": 0.0206,
      "num_tokens": 47334869.0,
      "step": 106
    },
    {
      "epoch": 1.04920049200492,
      "grad_norm": 0.0805362815127939,
      "learning_rate": 9.282825290994282e-06,
      "loss": 0.0213,
      "num_tokens": 47797630.0,
      "step": 107
    },
    {
      "epoch": 1.0590405904059041,
      "grad_norm": 0.07839099238240128,
      "learning_rate": 9.267249152181379e-06,
      "loss": 0.0454,
      "num_tokens": 48281974.0,
      "step": 108
    },
    {
      "epoch": 1.068880688806888,
      "grad_norm": 0.0757738535866923,
      "learning_rate": 9.251520654153028e-06,
      "loss": 0.022,
      "num_tokens": 48730118.0,
      "step": 109
    },
    {
      "epoch": 1.0787207872078721,
      "grad_norm": 0.08256710571520359,
      "learning_rate": 9.235640433018363e-06,
      "loss": 0.0195,
      "num_tokens": 49197576.0,
      "step": 110
    },
    {
      "epoch": 1.088560885608856,
      "grad_norm": 0.07849933177459094,
      "learning_rate": 9.219609131022684e-06,
      "loss": 0.0203,
      "num_tokens": 49673054.0,
      "step": 111
    },
    {
      "epoch": 1.09840098400984,
      "grad_norm": 0.08067924302373455,
      "learning_rate": 9.203427396521454e-06,
      "loss": 0.0219,
      "num_tokens": 50130569.0,
      "step": 112
    },
    {
      "epoch": 1.1082410824108242,
      "grad_norm": 0.07527801624664898,
      "learning_rate": 9.187095883954104e-06,
      "loss": 0.0195,
      "num_tokens": 50574721.0,
      "step": 113
    },
    {
      "epoch": 1.118081180811808,
      "grad_norm": 0.08229755724299215,
      "learning_rate": 9.170615253817547e-06,
      "loss": 0.0193,
      "num_tokens": 51010865.0,
      "step": 114
    },
    {
      "epoch": 1.1279212792127922,
      "grad_norm": 0.07673721236222701,
      "learning_rate": 9.153986172639474e-06,
      "loss": 0.0211,
      "num_tokens": 51469765.0,
      "step": 115
    },
    {
      "epoch": 1.137761377613776,
      "grad_norm": 0.0845900192373935,
      "learning_rate": 9.137209312951395e-06,
      "loss": 0.0226,
      "num_tokens": 51906114.0,
      "step": 116
    },
    {
      "epoch": 1.1476014760147601,
      "grad_norm": 0.08215860044207468,
      "learning_rate": 9.12028535326144e-06,
      "loss": 0.022,
      "num_tokens": 52354068.0,
      "step": 117
    },
    {
      "epoch": 1.1574415744157442,
      "grad_norm": 0.07420368746928867,
      "learning_rate": 9.103214978026922e-06,
      "loss": 0.0188,
      "num_tokens": 52836346.0,
      "step": 118
    },
    {
      "epoch": 1.1672816728167281,
      "grad_norm": 0.07450541307438634,
      "learning_rate": 9.085998877626644e-06,
      "loss": 0.0192,
      "num_tokens": 53299172.0,
      "step": 119
    },
    {
      "epoch": 1.1771217712177122,
      "grad_norm": 0.07878886229739003,
      "learning_rate": 9.068637748332993e-06,
      "loss": 0.0215,
      "num_tokens": 53759861.0,
      "step": 120
    },
    {
      "epoch": 1.186961869618696,
      "grad_norm": 0.08311056334441597,
      "learning_rate": 9.051132292283772e-06,
      "loss": 0.0208,
      "num_tokens": 54228512.0,
      "step": 121
    },
    {
      "epoch": 1.1968019680196802,
      "grad_norm": 0.07068781735081182,
      "learning_rate": 9.033483217453801e-06,
      "loss": 0.0812,
      "num_tokens": 54692852.0,
      "step": 122
    },
    {
      "epoch": 1.2066420664206643,
      "grad_norm": 0.596254901083269,
      "learning_rate": 9.015691237626292e-06,
      "loss": 0.0199,
      "num_tokens": 55139782.0,
      "step": 123
    },
    {
      "epoch": 1.2164821648216482,
      "grad_norm": 0.08202279255895727,
      "learning_rate": 8.997757072363976e-06,
      "loss": 0.0342,
      "num_tokens": 55604658.0,
      "step": 124
    },
    {
      "epoch": 1.2263222632226323,
      "grad_norm": 0.09057478290667956,
      "learning_rate": 8.979681446980002e-06,
      "loss": 0.0227,
      "num_tokens": 56030690.0,
      "step": 125
    },
    {
      "epoch": 1.2361623616236161,
      "grad_norm": 0.07661103115531635,
      "learning_rate": 8.961465092508607e-06,
      "loss": 0.0339,
      "num_tokens": 56492821.0,
      "step": 126
    },
    {
      "epoch": 1.2460024600246002,
      "grad_norm": 0.08310739437969392,
      "learning_rate": 8.943108745675542e-06,
      "loss": 0.0249,
      "num_tokens": 56927699.0,
      "step": 127
    },
    {
      "epoch": 1.2558425584255843,
      "grad_norm": 0.08009221352147507,
      "learning_rate": 8.92461314886829e-06,
      "loss": 0.0209,
      "num_tokens": 57365827.0,
      "step": 128
    },
    {
      "epoch": 1.2656826568265682,
      "grad_norm": 0.07973094836265254,
      "learning_rate": 8.905979050106029e-06,
      "loss": 0.0251,
      "num_tokens": 57821453.0,
      "step": 129
    },
    {
      "epoch": 1.2755227552275523,
      "grad_norm": 0.07600070319773061,
      "learning_rate": 8.887207203009385e-06,
      "loss": 0.0213,
      "num_tokens": 58267867.0,
      "step": 130
    },
    {
      "epoch": 1.2853628536285364,
      "grad_norm": 0.07258837000806613,
      "learning_rate": 8.868298366769956e-06,
      "loss": 0.0198,
      "num_tokens": 58715078.0,
      "step": 131
    },
    {
      "epoch": 1.2952029520295203,
      "grad_norm": 0.07826062337656157,
      "learning_rate": 8.849253306119601e-06,
      "loss": 0.0199,
      "num_tokens": 59159310.0,
      "step": 132
    },
    {
      "epoch": 1.3050430504305042,
      "grad_norm": 0.07770042002025847,
      "learning_rate": 8.83007279129952e-06,
      "loss": 0.027,
      "num_tokens": 59594031.0,
      "step": 133
    },
    {
      "epoch": 1.3148831488314883,
      "grad_norm": 0.07607344407726713,
      "learning_rate": 8.810757598029094e-06,
      "loss": 0.0342,
      "num_tokens": 60038506.0,
      "step": 134
    },
    {
      "epoch": 1.3247232472324724,
      "grad_norm": 0.08771686774228402,
      "learning_rate": 8.79130850747452e-06,
      "loss": 0.0234,
      "num_tokens": 60492486.0,
      "step": 135
    },
    {
      "epoch": 1.3345633456334562,
      "grad_norm": 0.07482147000786651,
      "learning_rate": 8.771726306217217e-06,
      "loss": 0.0196,
      "num_tokens": 60925341.0,
      "step": 136
    },
    {
      "epoch": 1.3444034440344403,
      "grad_norm": 0.07171750614547971,
      "learning_rate": 8.752011786222011e-06,
      "loss": 0.0224,
      "num_tokens": 61401128.0,
      "step": 137
    },
    {
      "epoch": 1.3542435424354244,
      "grad_norm": 0.07289189868770962,
      "learning_rate": 8.732165744805107e-06,
      "loss": 0.0198,
      "num_tokens": 61845691.0,
      "step": 138
    },
    {
      "epoch": 1.3640836408364083,
      "grad_norm": 0.07907747558023923,
      "learning_rate": 8.712188984601845e-06,
      "loss": 0.0185,
      "num_tokens": 62286361.0,
      "step": 139
    },
    {
      "epoch": 1.3739237392373924,
      "grad_norm": 0.06910414114179665,
      "learning_rate": 8.692082313534233e-06,
      "loss": 0.0179,
      "num_tokens": 62727406.0,
      "step": 140
    },
    {
      "epoch": 1.3837638376383765,
      "grad_norm": 0.07791959325829377,
      "learning_rate": 8.671846544778284e-06,
      "loss": 0.0204,
      "num_tokens": 63182141.0,
      "step": 141
    },
    {
      "epoch": 1.3936039360393604,
      "grad_norm": 0.0741558195977179,
      "learning_rate": 8.651482496731116e-06,
      "loss": 0.0178,
      "num_tokens": 63600729.0,
      "step": 142
    },
    {
      "epoch": 1.4034440344403443,
      "grad_norm": 0.07283375136096223,
      "learning_rate": 8.630990992977854e-06,
      "loss": 0.0198,
      "num_tokens": 64066267.0,
      "step": 143
    },
    {
      "epoch": 1.4132841328413284,
      "grad_norm": 0.0731783816547012,
      "learning_rate": 8.61037286225834e-06,
      "loss": 0.2547,
      "num_tokens": 64515946.0,
      "step": 144
    },
    {
      "epoch": 1.4231242312423125,
      "grad_norm": 1.0212050791856901,
      "learning_rate": 8.589628938433587e-06,
      "loss": 0.0192,
      "num_tokens": 64949958.0,
      "step": 145
    },
    {
      "epoch": 1.4329643296432963,
      "grad_norm": 0.09844320658741419,
      "learning_rate": 8.56876006045208e-06,
      "loss": 0.0176,
      "num_tokens": 65381018.0,
      "step": 146
    },
    {
      "epoch": 1.4428044280442804,
      "grad_norm": 0.07030907656382593,
      "learning_rate": 8.547767072315835e-06,
      "loss": 0.0241,
      "num_tokens": 65814016.0,
      "step": 147
    },
    {
      "epoch": 1.4526445264452645,
      "grad_norm": 0.0779412275694533,
      "learning_rate": 8.526650823046266e-06,
      "loss": 0.0265,
      "num_tokens": 66252980.0,
      "step": 148
    },
    {
      "epoch": 1.4624846248462484,
      "grad_norm": 0.09570533939331194,
      "learning_rate": 8.505412166649847e-06,
      "loss": 0.0199,
      "num_tokens": 66718111.0,
      "step": 149
    },
    {
      "epoch": 1.4723247232472325,
      "grad_norm": 0.07915246167438994,
      "learning_rate": 8.484051962083579e-06,
      "loss": 0.0204,
      "num_tokens": 67163762.0,
      "step": 150
    },
    {
      "epoch": 1.4821648216482166,
      "grad_norm": 0.07935176799416567,
      "learning_rate": 8.462571073220243e-06,
      "loss": 0.0225,
      "num_tokens": 67624386.0,
      "step": 151
    },
    {
      "epoch": 1.4920049200492005,
      "grad_norm": 0.07841589822630919,
      "learning_rate": 8.44097036881347e-06,
      "loss": 0.0392,
      "num_tokens": 68065290.0,
      "step": 152
    },
    {
      "epoch": 1.5018450184501844,
      "grad_norm": 0.3517146293571387,
      "learning_rate": 8.419250722462603e-06,
      "loss": 0.0178,
      "num_tokens": 68519107.0,
      "step": 153
    },
    {
      "epoch": 1.5116851168511685,
      "grad_norm": 0.0764909788834621,
      "learning_rate": 8.39741301257736e-06,
      "loss": 0.0194,
      "num_tokens": 68971128.0,
      "step": 154
    },
    {
      "epoch": 1.5215252152521526,
      "grad_norm": 0.08078822036852527,
      "learning_rate": 8.375458122342317e-06,
      "loss": 0.0206,
      "num_tokens": 69403792.0,
      "step": 155
    },
    {
      "epoch": 1.5313653136531364,
      "grad_norm": 0.08235320219175549,
      "learning_rate": 8.353386939681186e-06,
      "loss": 0.0175,
      "num_tokens": 69836602.0,
      "step": 156
    },
    {
      "epoch": 1.5412054120541205,
      "grad_norm": 0.0735540837139594,
      "learning_rate": 8.331200357220908e-06,
      "loss": 0.0194,
      "num_tokens": 70283814.0,
      "step": 157
    },
    {
      "epoch": 1.5510455104551046,
      "grad_norm": 0.07322399084658018,
      "learning_rate": 8.308899272255542e-06,
      "loss": 0.0184,
      "num_tokens": 70726284.0,
      "step": 158
    },
    {
      "epoch": 1.5608856088560885,
      "grad_norm": 0.07790348390650517,
      "learning_rate": 8.286484586709989e-06,
      "loss": 0.0183,
      "num_tokens": 71155169.0,
      "step": 159
    },
    {
      "epoch": 1.5707257072570726,
      "grad_norm": 0.08611809383964489,
      "learning_rate": 8.263957207103506e-06,
      "loss": 0.0205,
      "num_tokens": 71591204.0,
      "step": 160
    },
    {
      "epoch": 1.5805658056580567,
      "grad_norm": 0.0706229845173915,
      "learning_rate": 8.241318044513046e-06,
      "loss": 0.0277,
      "num_tokens": 72032119.0,
      "step": 161
    },
    {
      "epoch": 1.5904059040590406,
      "grad_norm": 0.09019039164269532,
      "learning_rate": 8.218568014536414e-06,
      "loss": 0.0176,
      "num_tokens": 72492164.0,
      "step": 162
    },
    {
      "epoch": 1.6002460024600245,
      "grad_norm": 0.07947315916491103,
      "learning_rate": 8.195708037255233e-06,
      "loss": 0.0202,
      "num_tokens": 72962752.0,
      "step": 163
    },
    {
      "epoch": 1.6100861008610086,
      "grad_norm": 0.06840189166732885,
      "learning_rate": 8.172739037197739e-06,
      "loss": 0.018,
      "num_tokens": 73415974.0,
      "step": 164
    },
    {
      "epoch": 1.6199261992619927,
      "grad_norm": 0.07366616747573093,
      "learning_rate": 8.149661943301382e-06,
      "loss": 0.0181,
      "num_tokens": 73882834.0,
      "step": 165
    },
    {
      "epoch": 1.6297662976629765,
      "grad_norm": 0.07081012920317416,
      "learning_rate": 8.126477688875262e-06,
      "loss": 0.0204,
      "num_tokens": 74321580.0,
      "step": 166
    },
    {
      "epoch": 1.6396063960639606,
      "grad_norm": 0.07863097311534642,
      "learning_rate": 8.103187211562386e-06,
      "loss": 0.0229,
      "num_tokens": 74781751.0,
      "step": 167
    },
    {
      "epoch": 1.6494464944649447,
      "grad_norm": 0.10797044478776457,
      "learning_rate": 8.079791453301742e-06,
      "loss": 0.0287,
      "num_tokens": 75219935.0,
      "step": 168
    },
    {
      "epoch": 1.6592865928659286,
      "grad_norm": 0.07041534985061697,
      "learning_rate": 8.056291360290202e-06,
      "loss": 0.0248,
      "num_tokens": 75665232.0,
      "step": 169
    },
    {
      "epoch": 1.6691266912669127,
      "grad_norm": 0.08695303118518641,
      "learning_rate": 8.032687882944264e-06,
      "loss": 0.0193,
      "num_tokens": 76087411.0,
      "step": 170
    },
    {
      "epoch": 1.6789667896678968,
      "grad_norm": 0.06704813880798238,
      "learning_rate": 8.0089819758616e-06,
      "loss": 0.0169,
      "num_tokens": 76529931.0,
      "step": 171
    },
    {
      "epoch": 1.6888068880688807,
      "grad_norm": 0.06935996975041725,
      "learning_rate": 7.985174597782469e-06,
      "loss": 0.0197,
      "num_tokens": 76974869.0,
      "step": 172
    },
    {
      "epoch": 1.6986469864698646,
      "grad_norm": 0.0812644475398725,
      "learning_rate": 7.961266711550922e-06,
      "loss": 0.0259,
      "num_tokens": 77413009.0,
      "step": 173
    },
    {
      "epoch": 1.7084870848708487,
      "grad_norm": 0.07469198601302375,
      "learning_rate": 7.937259284075872e-06,
      "loss": 0.0191,
      "num_tokens": 77854298.0,
      "step": 174
    },
    {
      "epoch": 1.7183271832718328,
      "grad_norm": 0.07554209425696685,
      "learning_rate": 7.913153286291995e-06,
      "loss": 0.025,
      "num_tokens": 78299682.0,
      "step": 175
    },
    {
      "epoch": 1.7281672816728166,
      "grad_norm": 0.07564661483692575,
      "learning_rate": 7.888949693120443e-06,
      "loss": 0.0172,
      "num_tokens": 78723460.0,
      "step": 176
    },
    {
      "epoch": 1.7380073800738007,
      "grad_norm": 0.6264202015289688,
      "learning_rate": 7.864649483429442e-06,
      "loss": 0.0402,
      "num_tokens": 79151526.0,
      "step": 177
    },
    {
      "epoch": 1.7478474784747848,
      "grad_norm": 0.07431323606896861,
      "learning_rate": 7.840253639994676e-06,
      "loss": 0.0182,
      "num_tokens": 79591692.0,
      "step": 178
    },
    {
      "epoch": 1.7576875768757687,
      "grad_norm": 0.07199128250127072,
      "learning_rate": 7.815763149459563e-06,
      "loss": 0.018,
      "num_tokens": 80054397.0,
      "step": 179
    },
    {
      "epoch": 1.7675276752767528,
      "grad_norm": 0.0736771332831437,
      "learning_rate": 7.791179002295334e-06,
      "loss": 0.0182,
      "num_tokens": 80527436.0,
      "step": 180
    },
    {
      "epoch": 1.777367773677737,
      "grad_norm": 0.0722896910687323,
      "learning_rate": 7.766502192760995e-06,
      "loss": 0.0299,
      "num_tokens": 80984085.0,
      "step": 181
    },
    {
      "epoch": 1.7872078720787208,
      "grad_norm": 0.13146348676004535,
      "learning_rate": 7.741733718863096e-06,
      "loss": 0.0172,
      "num_tokens": 81417093.0,
      "step": 182
    },
    {
      "epoch": 1.7970479704797047,
      "grad_norm": 0.07559775090622188,
      "learning_rate": 7.71687458231538e-06,
      "loss": 0.0173,
      "num_tokens": 81857802.0,
      "step": 183
    },
    {
      "epoch": 1.8068880688806888,
      "grad_norm": 0.07625026619956689,
      "learning_rate": 7.69192578849827e-06,
      "loss": 0.0174,
      "num_tokens": 82314635.0,
      "step": 184
    },
    {
      "epoch": 1.8167281672816729,
      "grad_norm": 0.07079163666898536,
      "learning_rate": 7.666888346418205e-06,
      "loss": 0.0255,
      "num_tokens": 82774404.0,
      "step": 185
    },
    {
      "epoch": 1.8265682656826567,
      "grad_norm": 0.07862230056744444,
      "learning_rate": 7.641763268666832e-06,
      "loss": 0.0166,
      "num_tokens": 83224858.0,
      "step": 186
    },
    {
      "epoch": 1.8364083640836408,
      "grad_norm": 0.07767548895299481,
      "learning_rate": 7.616551571380061e-06,
      "loss": 0.0303,
      "num_tokens": 83685638.0,
      "step": 187
    },
    {
      "epoch": 1.846248462484625,
      "grad_norm": 0.0767555813557926,
      "learning_rate": 7.5912542741969585e-06,
      "loss": 0.0173,
      "num_tokens": 84118329.0,
      "step": 188
    },
    {
      "epoch": 1.8560885608856088,
      "grad_norm": 0.06505326217418561,
      "learning_rate": 7.5658724002185215e-06,
      "loss": 0.2302,
      "num_tokens": 84642441.0,
      "step": 189
    },
    {
      "epoch": 1.865928659286593,
      "grad_norm": 0.9831912884395022,
      "learning_rate": 7.54040697596629e-06,
      "loss": 0.0173,
      "num_tokens": 85075013.0,
      "step": 190
    },
    {
      "epoch": 1.875768757687577,
      "grad_norm": 0.0852074767092427,
      "learning_rate": 7.514859031340835e-06,
      "loss": 0.0197,
      "num_tokens": 85539398.0,
      "step": 191
    },
    {
      "epoch": 1.8856088560885609,
      "grad_norm": 0.07502455159038045,
      "learning_rate": 7.489229599580111e-06,
      "loss": 0.0167,
      "num_tokens": 85976652.0,
      "step": 192
    },
    {
      "epoch": 1.8954489544895448,
      "grad_norm": 0.07796568336104527,
      "learning_rate": 7.463519717217663e-06,
      "loss": 0.0253,
      "num_tokens": 86404836.0,
      "step": 193
    },
    {
      "epoch": 1.9052890528905289,
      "grad_norm": 0.07733304316410633,
      "learning_rate": 7.437730424040702e-06,
      "loss": 0.0232,
      "num_tokens": 86871021.0,
      "step": 194
    },
    {
      "epoch": 1.915129151291513,
      "grad_norm": 0.07837311923363188,
      "learning_rate": 7.411862763048068e-06,
      "loss": 0.0228,
      "num_tokens": 87328297.0,
      "step": 195
    },
    {
      "epoch": 1.9249692496924968,
      "grad_norm": 0.07159308881612252,
      "learning_rate": 7.38591778040803e-06,
      "loss": 0.0178,
      "num_tokens": 87780478.0,
      "step": 196
    },
    {
      "epoch": 1.934809348093481,
      "grad_norm": 0.06995284279442164,
      "learning_rate": 7.359896525415986e-06,
      "loss": 0.0166,
      "num_tokens": 88245218.0,
      "step": 197
    },
    {
      "epoch": 1.944649446494465,
      "grad_norm": 0.074185946727602,
      "learning_rate": 7.333800050452024e-06,
      "loss": 0.0335,
      "num_tokens": 88720048.0,
      "step": 198
    },
    {
      "epoch": 1.954489544895449,
      "grad_norm": 0.0936664061322253,
      "learning_rate": 7.307629410938364e-06,
      "loss": 0.0156,
      "num_tokens": 89171687.0,
      "step": 199
    },
    {
      "epoch": 1.964329643296433,
      "grad_norm": 0.06592479834851843,
      "learning_rate": 7.281385665296663e-06,
      "loss": 0.0162,
      "num_tokens": 89636320.0,
      "step": 200
    },
    {
      "epoch": 1.974169741697417,
      "grad_norm": 0.08486840853612633,
      "learning_rate": 7.255069874905221e-06,
      "loss": 0.0177,
      "num_tokens": 90074778.0,
      "step": 201
    },
    {
      "epoch": 1.984009840098401,
      "grad_norm": 0.06923307537599123,
      "learning_rate": 7.228683104056051e-06,
      "loss": 0.0168,
      "num_tokens": 90519743.0,
      "step": 202
    },
    {
      "epoch": 1.9938499384993849,
      "grad_norm": 0.09250588119689185,
      "learning_rate": 7.202226419911832e-06,
      "loss": 0.0266,
      "num_tokens": 90971202.0,
      "step": 203
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.08932358174959376,
      "learning_rate": 7.175700892462757e-06,
      "loss": 0.0167,
      "num_tokens": 91183681.0,
      "step": 204
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.08808860927820206,
      "eval_num_tokens": 91183681.0,
      "eval_runtime": 53.9315,
      "eval_samples_per_second": 41.645,
      "eval_steps_per_second": 5.21,
      "step": 204
    },
    {
      "epoch": 2.009840098400984,
      "grad_norm": 0.07874869315833909,
      "learning_rate": 7.149107594483251e-06,
      "loss": 0.0142,
      "num_tokens": 91625671.0,
      "step": 205
    },
    {
      "epoch": 2.019680196801968,
      "grad_norm": 0.06385620551213778,
      "learning_rate": 7.122447601488592e-06,
      "loss": 0.0132,
      "num_tokens": 92071488.0,
      "step": 206
    },
    {
      "epoch": 2.029520295202952,
      "grad_norm": 0.06846197400142105,
      "learning_rate": 7.095721991691411e-06,
      "loss": 0.0149,
      "num_tokens": 92542156.0,
      "step": 207
    },
    {
      "epoch": 2.039360393603936,
      "grad_norm": 0.07424945414823086,
      "learning_rate": 7.0689318459580845e-06,
      "loss": 0.0156,
      "num_tokens": 93002703.0,
      "step": 208
    },
    {
      "epoch": 2.0492004920049203,
      "grad_norm": 0.06687580312011086,
      "learning_rate": 7.042078247765019e-06,
      "loss": 0.0135,
      "num_tokens": 93436834.0,
      "step": 209
    },
    {
      "epoch": 2.059040590405904,
      "grad_norm": 0.07720021453648518,
      "learning_rate": 7.015162283154843e-06,
      "loss": 0.0137,
      "num_tokens": 93871635.0,
      "step": 210
    },
    {
      "epoch": 2.068880688806888,
      "grad_norm": 0.13453391743262458,
      "learning_rate": 6.988185040692469e-06,
      "loss": 0.0221,
      "num_tokens": 94314058.0,
      "step": 211
    },
    {
      "epoch": 2.078720787207872,
      "grad_norm": 0.07982223152072775,
      "learning_rate": 6.961147611421076e-06,
      "loss": 0.017,
      "num_tokens": 94750976.0,
      "step": 212
    },
    {
      "epoch": 2.088560885608856,
      "grad_norm": 0.06995730861373262,
      "learning_rate": 6.934051088817988e-06,
      "loss": 0.0137,
      "num_tokens": 95193789.0,
      "step": 213
    },
    {
      "epoch": 2.09840098400984,
      "grad_norm": 0.07438600726959783,
      "learning_rate": 6.906896568750441e-06,
      "loss": 0.0193,
      "num_tokens": 95676386.0,
      "step": 214
    },
    {
      "epoch": 2.108241082410824,
      "grad_norm": 0.09331884860488432,
      "learning_rate": 6.87968514943127e-06,
      "loss": 0.0154,
      "num_tokens": 96137917.0,
      "step": 215
    },
    {
      "epoch": 2.1180811808118083,
      "grad_norm": 0.06703452835053635,
      "learning_rate": 6.852417931374494e-06,
      "loss": 0.0134,
      "num_tokens": 96568059.0,
      "step": 216
    },
    {
      "epoch": 2.127921279212792,
      "grad_norm": 0.07093081986870549,
      "learning_rate": 6.825096017350807e-06,
      "loss": 0.0138,
      "num_tokens": 97019588.0,
      "step": 217
    },
    {
      "epoch": 2.137761377613776,
      "grad_norm": 0.0650948479503258,
      "learning_rate": 6.797720512342967e-06,
      "loss": 0.0137,
      "num_tokens": 97456418.0,
      "step": 218
    },
    {
      "epoch": 2.14760147601476,
      "grad_norm": 0.06693139683273135,
      "learning_rate": 6.77029252350113e-06,
      "loss": 0.0142,
      "num_tokens": 97874765.0,
      "step": 219
    },
    {
      "epoch": 2.1574415744157442,
      "grad_norm": 0.07881816970778455,
      "learning_rate": 6.742813160098054e-06,
      "loss": 0.0188,
      "num_tokens": 98322373.0,
      "step": 220
    },
    {
      "epoch": 2.167281672816728,
      "grad_norm": 0.07381706020969016,
      "learning_rate": 6.715283533484242e-06,
      "loss": 0.0125,
      "num_tokens": 98762055.0,
      "step": 221
    },
    {
      "epoch": 2.177121771217712,
      "grad_norm": 0.06829050170688594,
      "learning_rate": 6.6877047570430044e-06,
      "loss": 0.0147,
      "num_tokens": 99212257.0,
      "step": 222
    },
    {
      "epoch": 2.1869618696186963,
      "grad_norm": 0.0726323898489312,
      "learning_rate": 6.660077946145412e-06,
      "loss": 0.0149,
      "num_tokens": 99651696.0,
      "step": 223
    },
    {
      "epoch": 2.19680196801968,
      "grad_norm": 0.06996376101830218,
      "learning_rate": 6.632404218105205e-06,
      "loss": 0.014,
      "num_tokens": 100115333.0,
      "step": 224
    },
    {
      "epoch": 2.206642066420664,
      "grad_norm": 0.07058857975728597,
      "learning_rate": 6.604684692133597e-06,
      "loss": 0.0128,
      "num_tokens": 100567168.0,
      "step": 225
    },
    {
      "epoch": 2.2164821648216484,
      "grad_norm": 0.06705830086377462,
      "learning_rate": 6.576920489294011e-06,
      "loss": 0.014,
      "num_tokens": 101017414.0,
      "step": 226
    },
    {
      "epoch": 2.2263222632226323,
      "grad_norm": 0.08216121325842957,
      "learning_rate": 6.549112732456739e-06,
      "loss": 0.0244,
      "num_tokens": 101478653.0,
      "step": 227
    },
    {
      "epoch": 2.236162361623616,
      "grad_norm": 0.06604918422838713,
      "learning_rate": 6.5212625462535365e-06,
      "loss": 0.0133,
      "num_tokens": 101922998.0,
      "step": 228
    },
    {
      "epoch": 2.2460024600246005,
      "grad_norm": 0.06450225948970358,
      "learning_rate": 6.493371057032129e-06,
      "loss": 0.0149,
      "num_tokens": 102357947.0,
      "step": 229
    },
    {
      "epoch": 2.2558425584255843,
      "grad_norm": 0.07514996917424294,
      "learning_rate": 6.465439392810664e-06,
      "loss": 0.0167,
      "num_tokens": 102803832.0,
      "step": 230
    },
    {
      "epoch": 2.265682656826568,
      "grad_norm": 0.06462428507734051,
      "learning_rate": 6.4374686832320944e-06,
      "loss": 0.0142,
      "num_tokens": 103241692.0,
      "step": 231
    },
    {
      "epoch": 2.275522755227552,
      "grad_norm": 0.06485952063828938,
      "learning_rate": 6.409460059518482e-06,
      "loss": 0.0136,
      "num_tokens": 103688326.0,
      "step": 232
    },
    {
      "epoch": 2.2853628536285364,
      "grad_norm": 0.06533997999817706,
      "learning_rate": 6.381414654425261e-06,
      "loss": 0.0131,
      "num_tokens": 104139997.0,
      "step": 233
    },
    {
      "epoch": 2.2952029520295203,
      "grad_norm": 0.06878268907753365,
      "learning_rate": 6.353333602195414e-06,
      "loss": 0.0138,
      "num_tokens": 104583247.0,
      "step": 234
    },
    {
      "epoch": 2.305043050430504,
      "grad_norm": 0.061527579151490784,
      "learning_rate": 6.325218038513604e-06,
      "loss": 0.0129,
      "num_tokens": 105013546.0,
      "step": 235
    },
    {
      "epoch": 2.3148831488314885,
      "grad_norm": 0.0688594189041464,
      "learning_rate": 6.2970691004602425e-06,
      "loss": 0.0147,
      "num_tokens": 105469533.0,
      "step": 236
    },
    {
      "epoch": 2.3247232472324724,
      "grad_norm": 0.07212293085873876,
      "learning_rate": 6.26888792646551e-06,
      "loss": 0.0138,
      "num_tokens": 105902012.0,
      "step": 237
    },
    {
      "epoch": 2.3345633456334562,
      "grad_norm": 0.07097729248579715,
      "learning_rate": 6.240675656263303e-06,
      "loss": 0.0133,
      "num_tokens": 106319708.0,
      "step": 238
    },
    {
      "epoch": 2.34440344403444,
      "grad_norm": 0.0702207231329528,
      "learning_rate": 6.212433430845145e-06,
      "loss": 0.0136,
      "num_tokens": 106767770.0,
      "step": 239
    },
    {
      "epoch": 2.3542435424354244,
      "grad_norm": 0.06717197740035392,
      "learning_rate": 6.184162392414044e-06,
      "loss": 0.0127,
      "num_tokens": 107230010.0,
      "step": 240
    },
    {
      "epoch": 2.3640836408364083,
      "grad_norm": 0.09206853570190297,
      "learning_rate": 6.155863684338294e-06,
      "loss": 0.0182,
      "num_tokens": 107696665.0,
      "step": 241
    },
    {
      "epoch": 2.373923739237392,
      "grad_norm": 0.07931539686074184,
      "learning_rate": 6.127538451105232e-06,
      "loss": 0.0156,
      "num_tokens": 108145998.0,
      "step": 242
    },
    {
      "epoch": 2.3837638376383765,
      "grad_norm": 0.0845167365221342,
      "learning_rate": 6.099187838274959e-06,
      "loss": 0.0304,
      "num_tokens": 108605347.0,
      "step": 243
    },
    {
      "epoch": 2.3936039360393604,
      "grad_norm": 0.8319925155014395,
      "learning_rate": 6.070812992434003e-06,
      "loss": 0.077,
      "num_tokens": 109053120.0,
      "step": 244
    },
    {
      "epoch": 2.4034440344403443,
      "grad_norm": 0.08254084053779843,
      "learning_rate": 6.042415061148954e-06,
      "loss": 0.0153,
      "num_tokens": 109511574.0,
      "step": 245
    },
    {
      "epoch": 2.4132841328413286,
      "grad_norm": 0.07621464852457635,
      "learning_rate": 6.013995192920044e-06,
      "loss": 0.013,
      "num_tokens": 109961861.0,
      "step": 246
    },
    {
      "epoch": 2.4231242312423125,
      "grad_norm": 0.06290755400921484,
      "learning_rate": 5.985554537134702e-06,
      "loss": 0.0133,
      "num_tokens": 110439530.0,
      "step": 247
    },
    {
      "epoch": 2.4329643296432963,
      "grad_norm": 0.06549923207889226,
      "learning_rate": 5.957094244021071e-06,
      "loss": 0.0133,
      "num_tokens": 110902468.0,
      "step": 248
    },
    {
      "epoch": 2.4428044280442807,
      "grad_norm": 0.06398296126869986,
      "learning_rate": 5.928615464601497e-06,
      "loss": 0.0128,
      "num_tokens": 111361759.0,
      "step": 249
    },
    {
      "epoch": 2.4526445264452645,
      "grad_norm": 0.062244715362799644,
      "learning_rate": 5.900119350645956e-06,
      "loss": 0.0128,
      "num_tokens": 111799435.0,
      "step": 250
    },
    {
      "epoch": 2.4624846248462484,
      "grad_norm": 0.06503161600374163,
      "learning_rate": 5.871607054625497e-06,
      "loss": 0.0128,
      "num_tokens": 112244747.0,
      "step": 251
    },
    {
      "epoch": 2.4723247232472323,
      "grad_norm": 0.08086590997362891,
      "learning_rate": 5.8430797296656125e-06,
      "loss": 0.0184,
      "num_tokens": 112678903.0,
      "step": 252
    },
    {
      "epoch": 2.4821648216482166,
      "grad_norm": 0.07239451855920867,
      "learning_rate": 5.814538529499622e-06,
      "loss": 0.0149,
      "num_tokens": 113132832.0,
      "step": 253
    },
    {
      "epoch": 2.4920049200492005,
      "grad_norm": 0.06030312987290577,
      "learning_rate": 5.785984608421993e-06,
      "loss": 0.0127,
      "num_tokens": 113568429.0,
      "step": 254
    },
    {
      "epoch": 2.5018450184501844,
      "grad_norm": 0.06349775541516244,
      "learning_rate": 5.757419121241667e-06,
      "loss": 0.0125,
      "num_tokens": 114042240.0,
      "step": 255
    },
    {
      "epoch": 2.5116851168511687,
      "grad_norm": 0.06952013750985335,
      "learning_rate": 5.7288432232353615e-06,
      "loss": 0.0204,
      "num_tokens": 114496441.0,
      "step": 256
    },
    {
      "epoch": 2.5215252152521526,
      "grad_norm": 0.0958262233433174,
      "learning_rate": 5.7002580701008325e-06,
      "loss": 0.0149,
      "num_tokens": 114936236.0,
      "step": 257
    },
    {
      "epoch": 2.5313653136531364,
      "grad_norm": 0.06572975411347728,
      "learning_rate": 5.6716648179101445e-06,
      "loss": 0.0123,
      "num_tokens": 115365529.0,
      "step": 258
    },
    {
      "epoch": 2.5412054120541203,
      "grad_norm": 0.07287254897275752,
      "learning_rate": 5.64306462306291e-06,
      "loss": 0.0177,
      "num_tokens": 115812361.0,
      "step": 259
    },
    {
      "epoch": 2.5510455104551046,
      "grad_norm": 0.0677506186552676,
      "learning_rate": 5.614458642239534e-06,
      "loss": 0.0126,
      "num_tokens": 116269752.0,
      "step": 260
    },
    {
      "epoch": 2.5608856088560885,
      "grad_norm": 0.07088790175345892,
      "learning_rate": 5.585848032354411e-06,
      "loss": 0.0139,
      "num_tokens": 116739082.0,
      "step": 261
    },
    {
      "epoch": 2.570725707257073,
      "grad_norm": 2.483507979054926,
      "learning_rate": 5.557233950509159e-06,
      "loss": 0.3298,
      "num_tokens": 117236975.0,
      "step": 262
    },
    {
      "epoch": 2.5805658056580567,
      "grad_norm": 0.6712341553033803,
      "learning_rate": 5.528617553945807e-06,
      "loss": 0.0131,
      "num_tokens": 117701799.0,
      "step": 263
    },
    {
      "epoch": 2.5904059040590406,
      "grad_norm": 0.070379027103792,
      "learning_rate": 5.500000000000001e-06,
      "loss": 0.019,
      "num_tokens": 118190544.0,
      "step": 264
    },
    {
      "epoch": 2.6002460024600245,
      "grad_norm": 0.09944926431551483,
      "learning_rate": 5.4713824460541964e-06,
      "loss": 0.0153,
      "num_tokens": 118625146.0,
      "step": 265
    },
    {
      "epoch": 2.6100861008610083,
      "grad_norm": 0.07370939155932825,
      "learning_rate": 5.442766049490843e-06,
      "loss": 0.0138,
      "num_tokens": 119077739.0,
      "step": 266
    },
    {
      "epoch": 2.6199261992619927,
      "grad_norm": 0.06555516765204612,
      "learning_rate": 5.414151967645591e-06,
      "loss": 0.0136,
      "num_tokens": 119502701.0,
      "step": 267
    },
    {
      "epoch": 2.6297662976629765,
      "grad_norm": 0.060577987544993946,
      "learning_rate": 5.385541357760469e-06,
      "loss": 0.0121,
      "num_tokens": 119956823.0,
      "step": 268
    },
    {
      "epoch": 2.639606396063961,
      "grad_norm": 0.06969958736256228,
      "learning_rate": 5.35693537693709e-06,
      "loss": 0.0131,
      "num_tokens": 120410284.0,
      "step": 269
    },
    {
      "epoch": 2.6494464944649447,
      "grad_norm": 0.08178808292429539,
      "learning_rate": 5.3283351820898586e-06,
      "loss": 0.0183,
      "num_tokens": 120837514.0,
      "step": 270
    },
    {
      "epoch": 2.6592865928659286,
      "grad_norm": 0.12228602708630738,
      "learning_rate": 5.299741929899171e-06,
      "loss": 0.0206,
      "num_tokens": 121266377.0,
      "step": 271
    },
    {
      "epoch": 2.6691266912669125,
      "grad_norm": 0.07647057417070459,
      "learning_rate": 5.27115677676464e-06,
      "loss": 0.0154,
      "num_tokens": 121730907.0,
      "step": 272
    },
    {
      "epoch": 2.678966789667897,
      "grad_norm": 0.07263570161343703,
      "learning_rate": 5.242580878758334e-06,
      "loss": 0.0138,
      "num_tokens": 122162564.0,
      "step": 273
    },
    {
      "epoch": 2.6888068880688807,
      "grad_norm": 0.07390794347850005,
      "learning_rate": 5.21401539157801e-06,
      "loss": 0.0131,
      "num_tokens": 122644233.0,
      "step": 274
    },
    {
      "epoch": 2.6986469864698646,
      "grad_norm": 0.05624120433704004,
      "learning_rate": 5.1854614705003796e-06,
      "loss": 0.0114,
      "num_tokens": 123070674.0,
      "step": 275
    },
    {
      "epoch": 2.708487084870849,
      "grad_norm": 0.07371873132309133,
      "learning_rate": 5.156920270334389e-06,
      "loss": 0.0194,
      "num_tokens": 123517476.0,
      "step": 276
    },
    {
      "epoch": 2.7183271832718328,
      "grad_norm": 0.06758978472435712,
      "learning_rate": 5.1283929453745055e-06,
      "loss": 0.0129,
      "num_tokens": 123957650.0,
      "step": 277
    },
    {
      "epoch": 2.7281672816728166,
      "grad_norm": 0.06857276382476074,
      "learning_rate": 5.099880649354044e-06,
      "loss": 0.0125,
      "num_tokens": 124423561.0,
      "step": 278
    },
    {
      "epoch": 2.7380073800738005,
      "grad_norm": 0.06198166285648246,
      "learning_rate": 5.071384535398505e-06,
      "loss": 0.0119,
      "num_tokens": 124871204.0,
      "step": 279
    },
    {
      "epoch": 2.747847478474785,
      "grad_norm": 0.05801997208341688,
      "learning_rate": 5.04290575597893e-06,
      "loss": 0.0119,
      "num_tokens": 125320936.0,
      "step": 280
    },
    {
      "epoch": 2.7576875768757687,
      "grad_norm": 0.09983800531852628,
      "learning_rate": 5.0144454628653015e-06,
      "loss": 0.0157,
      "num_tokens": 125785587.0,
      "step": 281
    },
    {
      "epoch": 2.767527675276753,
      "grad_norm": 0.05961861980322237,
      "learning_rate": 4.986004807079959e-06,
      "loss": 0.0119,
      "num_tokens": 126223799.0,
      "step": 282
    },
    {
      "epoch": 2.777367773677737,
      "grad_norm": 0.06887056012305312,
      "learning_rate": 4.957584938851048e-06,
      "loss": 0.0127,
      "num_tokens": 126674560.0,
      "step": 283
    },
    {
      "epoch": 2.787207872078721,
      "grad_norm": 0.06432285678662777,
      "learning_rate": 4.929187007565996e-06,
      "loss": 0.0124,
      "num_tokens": 127121758.0,
      "step": 284
    },
    {
      "epoch": 2.7970479704797047,
      "grad_norm": 0.06283306903955838,
      "learning_rate": 4.9008121617250425e-06,
      "loss": 0.0122,
      "num_tokens": 127564319.0,
      "step": 285
    },
    {
      "epoch": 2.8068880688806885,
      "grad_norm": 0.07395862495517919,
      "learning_rate": 4.87246154889477e-06,
      "loss": 0.0125,
      "num_tokens": 128014723.0,
      "step": 286
    },
    {
      "epoch": 2.816728167281673,
      "grad_norm": 0.06772968868173306,
      "learning_rate": 4.8441363156617085e-06,
      "loss": 0.026,
      "num_tokens": 128456573.0,
      "step": 287
    },
    {
      "epoch": 2.8265682656826567,
      "grad_norm": 0.2058477599150272,
      "learning_rate": 4.815837607585957e-06,
      "loss": 0.0313,
      "num_tokens": 128888085.0,
      "step": 288
    },
    {
      "epoch": 2.836408364083641,
      "grad_norm": 0.05983028509302605,
      "learning_rate": 4.787566569154855e-06,
      "loss": 0.0136,
      "num_tokens": 129344186.0,
      "step": 289
    },
    {
      "epoch": 2.846248462484625,
      "grad_norm": 0.1679165256737002,
      "learning_rate": 4.759324343736698e-06,
      "loss": 0.0268,
      "num_tokens": 129820337.0,
      "step": 290
    },
    {
      "epoch": 2.856088560885609,
      "grad_norm": 0.069693981729958,
      "learning_rate": 4.731112073534491e-06,
      "loss": 0.012,
      "num_tokens": 130264132.0,
      "step": 291
    },
    {
      "epoch": 2.8659286592865927,
      "grad_norm": 0.05673801969192786,
      "learning_rate": 4.70293089953976e-06,
      "loss": 0.237,
      "num_tokens": 130747367.0,
      "step": 292
    },
    {
      "epoch": 2.875768757687577,
      "grad_norm": 0.9244716369700087,
      "learning_rate": 4.674781961486399e-06,
      "loss": 0.0129,
      "num_tokens": 131189544.0,
      "step": 293
    },
    {
      "epoch": 2.885608856088561,
      "grad_norm": 0.0670539720853974,
      "learning_rate": 4.646666397804586e-06,
      "loss": 0.0127,
      "num_tokens": 131615817.0,
      "step": 294
    },
    {
      "epoch": 2.8954489544895448,
      "grad_norm": 0.07778029323101539,
      "learning_rate": 4.618585345574741e-06,
      "loss": 0.0136,
      "num_tokens": 132065833.0,
      "step": 295
    },
    {
      "epoch": 2.905289052890529,
      "grad_norm": 0.06633645417900966,
      "learning_rate": 4.5905399404815196e-06,
      "loss": 0.0119,
      "num_tokens": 132513181.0,
      "step": 296
    },
    {
      "epoch": 2.915129151291513,
      "grad_norm": 0.06604742202311176,
      "learning_rate": 4.562531316767908e-06,
      "loss": 0.0178,
      "num_tokens": 132975979.0,
      "step": 297
    },
    {
      "epoch": 2.924969249692497,
      "grad_norm": 0.06375772945002761,
      "learning_rate": 4.534560607189338e-06,
      "loss": 0.0121,
      "num_tokens": 133411946.0,
      "step": 298
    },
    {
      "epoch": 2.9348093480934807,
      "grad_norm": 0.0644873715390372,
      "learning_rate": 4.506628942967874e-06,
      "loss": 0.0226,
      "num_tokens": 133882037.0,
      "step": 299
    },
    {
      "epoch": 2.944649446494465,
      "grad_norm": 0.06122403707300358,
      "learning_rate": 4.478737453746464e-06,
      "loss": 0.0111,
      "num_tokens": 134338580.0,
      "step": 300
    },
    {
      "epoch": 2.954489544895449,
      "grad_norm": 0.06192995198797032,
      "learning_rate": 4.450887267543261e-06,
      "loss": 0.023,
      "num_tokens": 134806429.0,
      "step": 301
    },
    {
      "epoch": 2.9643296432964332,
      "grad_norm": 0.06577423487360488,
      "learning_rate": 4.423079510705992e-06,
      "loss": 0.0127,
      "num_tokens": 135253050.0,
      "step": 302
    },
    {
      "epoch": 2.974169741697417,
      "grad_norm": 0.061821762890230156,
      "learning_rate": 4.395315307866404e-06,
      "loss": 0.0118,
      "num_tokens": 135701900.0,
      "step": 303
    },
    {
      "epoch": 2.984009840098401,
      "grad_norm": 0.060295397517859534,
      "learning_rate": 4.3675957818947965e-06,
      "loss": 0.0112,
      "num_tokens": 136134539.0,
      "step": 304
    },
    {
      "epoch": 2.993849938499385,
      "grad_norm": 0.06204359834906306,
      "learning_rate": 4.33992205385459e-06,
      "loss": 0.0119,
      "num_tokens": 136581981.0,
      "step": 305
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.06204359834906306,
      "learning_rate": 4.312295242956998e-06,
      "loss": 0.0109,
      "num_tokens": 136774441.0,
      "step": 306
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.0963606908917427,
      "eval_num_tokens": 136774441.0,
      "eval_runtime": 53.9214,
      "eval_samples_per_second": 41.653,
      "eval_steps_per_second": 5.211,
      "step": 306
    },
    {
      "epoch": 3.009840098400984,
      "grad_norm": 0.08266586517900253,
      "learning_rate": 4.284716466515759e-06,
      "loss": 0.0218,
      "num_tokens": 137235846.0,
      "step": 307
    },
    {
      "epoch": 3.019680196801968,
      "grad_norm": 0.06025259361613064,
      "learning_rate": 4.257186839901948e-06,
      "loss": 0.01,
      "num_tokens": 137676575.0,
      "step": 308
    },
    {
      "epoch": 3.029520295202952,
      "grad_norm": 0.059520087712568295,
      "learning_rate": 4.229707476498871e-06,
      "loss": 0.0107,
      "num_tokens": 138127277.0,
      "step": 309
    },
    {
      "epoch": 3.039360393603936,
      "grad_norm": 0.060007105121960225,
      "learning_rate": 4.2022794876570335e-06,
      "loss": 0.0099,
      "num_tokens": 138558346.0,
      "step": 310
    },
    {
      "epoch": 3.0492004920049203,
      "grad_norm": 0.05765555936281279,
      "learning_rate": 4.1749039826491956e-06,
      "loss": 0.2021,
      "num_tokens": 139029117.0,
      "step": 311
    },
    {
      "epoch": 3.059040590405904,
      "grad_norm": 0.25549047851203505,
      "learning_rate": 4.1475820686255055e-06,
      "loss": 0.01,
      "num_tokens": 139465608.0,
      "step": 312
    },
    {
      "epoch": 3.068880688806888,
      "grad_norm": 0.05745397404349778,
      "learning_rate": 4.120314850568731e-06,
      "loss": 0.0291,
      "num_tokens": 139932040.0,
      "step": 313
    },
    {
      "epoch": 3.078720787207872,
      "grad_norm": 0.21571060654935606,
      "learning_rate": 4.093103431249563e-06,
      "loss": 0.011,
      "num_tokens": 140393810.0,
      "step": 314
    },
    {
      "epoch": 3.088560885608856,
      "grad_norm": 0.06271676867820344,
      "learning_rate": 4.065948911182015e-06,
      "loss": 0.018,
      "num_tokens": 140853306.0,
      "step": 315
    },
    {
      "epoch": 3.09840098400984,
      "grad_norm": 0.06529992912597996,
      "learning_rate": 4.038852388578925e-06,
      "loss": 0.0102,
      "num_tokens": 141293974.0,
      "step": 316
    },
    {
      "epoch": 3.108241082410824,
      "grad_norm": 0.0613594667302306,
      "learning_rate": 4.011814959307533e-06,
      "loss": 0.0101,
      "num_tokens": 141739396.0,
      "step": 317
    },
    {
      "epoch": 3.1180811808118083,
      "grad_norm": 0.06143281774280475,
      "learning_rate": 3.984837716845157e-06,
      "loss": 0.0098,
      "num_tokens": 142181417.0,
      "step": 318
    },
    {
      "epoch": 3.127921279212792,
      "grad_norm": 0.06065540767441434,
      "learning_rate": 3.957921752234982e-06,
      "loss": 0.0095,
      "num_tokens": 142615273.0,
      "step": 319
    },
    {
      "epoch": 3.137761377613776,
      "grad_norm": 0.0565367496699821,
      "learning_rate": 3.931068154041919e-06,
      "loss": 0.0156,
      "num_tokens": 143066695.0,
      "step": 320
    },
    {
      "epoch": 3.14760147601476,
      "grad_norm": 0.0928817994214938,
      "learning_rate": 3.904278008308589e-06,
      "loss": 0.0093,
      "num_tokens": 143543314.0,
      "step": 321
    },
    {
      "epoch": 3.1574415744157442,
      "grad_norm": 0.05348206917431186,
      "learning_rate": 3.877552398511409e-06,
      "loss": 0.0102,
      "num_tokens": 143978640.0,
      "step": 322
    },
    {
      "epoch": 3.167281672816728,
      "grad_norm": 0.05744861837720995,
      "learning_rate": 3.85089240551675e-06,
      "loss": 0.0096,
      "num_tokens": 144437143.0,
      "step": 323
    },
    {
      "epoch": 3.177121771217712,
      "grad_norm": 0.05917730480215664,
      "learning_rate": 3.8242991075372436e-06,
      "loss": 0.0103,
      "num_tokens": 144882614.0,
      "step": 324
    },
    {
      "epoch": 3.1869618696186963,
      "grad_norm": 0.06138753989215512,
      "learning_rate": 3.7977735800881687e-06,
      "loss": 0.01,
      "num_tokens": 145336615.0,
      "step": 325
    },
    {
      "epoch": 3.19680196801968,
      "grad_norm": 0.057934477141044834,
      "learning_rate": 3.7713168959439515e-06,
      "loss": 0.0097,
      "num_tokens": 145791703.0,
      "step": 326
    },
    {
      "epoch": 3.206642066420664,
      "grad_norm": 0.062311400511582536,
      "learning_rate": 3.74493012509478e-06,
      "loss": 0.0163,
      "num_tokens": 146256588.0,
      "step": 327
    },
    {
      "epoch": 3.2164821648216484,
      "grad_norm": 0.11046706497961999,
      "learning_rate": 3.718614334703339e-06,
      "loss": 0.0096,
      "num_tokens": 146704790.0,
      "step": 328
    },
    {
      "epoch": 3.2263222632226323,
      "grad_norm": 0.06040935915809342,
      "learning_rate": 3.692370589061639e-06,
      "loss": 0.0161,
      "num_tokens": 147150851.0,
      "step": 329
    },
    {
      "epoch": 3.236162361623616,
      "grad_norm": 0.06309596528426079,
      "learning_rate": 3.6661999495479772e-06,
      "loss": 0.0116,
      "num_tokens": 147586533.0,
      "step": 330
    },
    {
      "epoch": 3.2460024600246005,
      "grad_norm": 0.0775947611650109,
      "learning_rate": 3.640103474584016e-06,
      "loss": 0.0102,
      "num_tokens": 148012817.0,
      "step": 331
    },
    {
      "epoch": 3.2558425584255843,
      "grad_norm": 0.060442066581616015,
      "learning_rate": 3.614082219591972e-06,
      "loss": 0.0094,
      "num_tokens": 148454349.0,
      "step": 332
    },
    {
      "epoch": 3.265682656826568,
      "grad_norm": 0.0599277899760194,
      "learning_rate": 3.588137236951934e-06,
      "loss": 0.0096,
      "num_tokens": 148908837.0,
      "step": 333
    },
    {
      "epoch": 3.275522755227552,
      "grad_norm": 0.06389649266611047,
      "learning_rate": 3.5622695759592996e-06,
      "loss": 0.0091,
      "num_tokens": 149387409.0,
      "step": 334
    },
    {
      "epoch": 3.2853628536285364,
      "grad_norm": 0.059031876557593344,
      "learning_rate": 3.5364802827823397e-06,
      "loss": 0.0124,
      "num_tokens": 149842184.0,
      "step": 335
    },
    {
      "epoch": 3.2952029520295203,
      "grad_norm": 0.06425762134540147,
      "learning_rate": 3.5107704004198904e-06,
      "loss": 0.0096,
      "num_tokens": 150294624.0,
      "step": 336
    },
    {
      "epoch": 3.305043050430504,
      "grad_norm": 0.060359900802863305,
      "learning_rate": 3.485140968659166e-06,
      "loss": 0.0156,
      "num_tokens": 150757952.0,
      "step": 337
    },
    {
      "epoch": 3.3148831488314885,
      "grad_norm": 0.06451910432321761,
      "learning_rate": 3.4595930240337115e-06,
      "loss": 0.0093,
      "num_tokens": 151210941.0,
      "step": 338
    },
    {
      "epoch": 3.3247232472324724,
      "grad_norm": 0.05771756769585445,
      "learning_rate": 3.4341275997814795e-06,
      "loss": 0.0311,
      "num_tokens": 151659703.0,
      "step": 339
    },
    {
      "epoch": 3.3345633456334562,
      "grad_norm": 0.2709101034464869,
      "learning_rate": 3.408745725803042e-06,
      "loss": 0.0198,
      "num_tokens": 152096656.0,
      "step": 340
    },
    {
      "epoch": 3.34440344403444,
      "grad_norm": 0.2165805542100797,
      "learning_rate": 3.383448428619941e-06,
      "loss": 0.0109,
      "num_tokens": 152535937.0,
      "step": 341
    },
    {
      "epoch": 3.3542435424354244,
      "grad_norm": 0.06249104678860667,
      "learning_rate": 3.3582367313331692e-06,
      "loss": 0.0241,
      "num_tokens": 153012481.0,
      "step": 342
    },
    {
      "epoch": 3.3640836408364083,
      "grad_norm": 0.07444091538512662,
      "learning_rate": 3.3331116535817974e-06,
      "loss": 0.0096,
      "num_tokens": 153457239.0,
      "step": 343
    },
    {
      "epoch": 3.373923739237392,
      "grad_norm": 0.05744783875540723,
      "learning_rate": 3.308074211501732e-06,
      "loss": 0.0112,
      "num_tokens": 153885310.0,
      "step": 344
    },
    {
      "epoch": 3.3837638376383765,
      "grad_norm": 0.062108203142145886,
      "learning_rate": 3.2831254176846205e-06,
      "loss": 0.0102,
      "num_tokens": 154315565.0,
      "step": 345
    },
    {
      "epoch": 3.3936039360393604,
      "grad_norm": 0.06493988486024563,
      "learning_rate": 3.258266281136905e-06,
      "loss": 0.0154,
      "num_tokens": 154761237.0,
      "step": 346
    },
    {
      "epoch": 3.4034440344403443,
      "grad_norm": 0.07703452506780802,
      "learning_rate": 3.233497807239008e-06,
      "loss": 0.0149,
      "num_tokens": 155219079.0,
      "step": 347
    },
    {
      "epoch": 3.4132841328413286,
      "grad_norm": 0.07716474025857703,
      "learning_rate": 3.2088209977046657e-06,
      "loss": 0.0099,
      "num_tokens": 155672847.0,
      "step": 348
    },
    {
      "epoch": 3.4231242312423125,
      "grad_norm": 0.0598011605849924,
      "learning_rate": 3.1842368505404388e-06,
      "loss": 0.0097,
      "num_tokens": 156097592.0,
      "step": 349
    },
    {
      "epoch": 3.4329643296432963,
      "grad_norm": 0.06067024127693304,
      "learning_rate": 3.1597463600053258e-06,
      "loss": 0.0097,
      "num_tokens": 156543931.0,
      "step": 350
    },
    {
      "epoch": 3.4428044280442807,
      "grad_norm": 0.06276348610439125,
      "learning_rate": 3.135350516570559e-06,
      "loss": 0.0115,
      "num_tokens": 156993093.0,
      "step": 351
    },
    {
      "epoch": 3.4526445264452645,
      "grad_norm": 0.07056305058653452,
      "learning_rate": 3.111050306879556e-06,
      "loss": 0.0161,
      "num_tokens": 157435895.0,
      "step": 352
    },
    {
      "epoch": 3.4624846248462484,
      "grad_norm": 0.0692853066303934,
      "learning_rate": 3.0868467137080075e-06,
      "loss": 0.0124,
      "num_tokens": 157859703.0,
      "step": 353
    },
    {
      "epoch": 3.4723247232472323,
      "grad_norm": 0.06622059827297899,
      "learning_rate": 3.0627407159241273e-06,
      "loss": 0.0098,
      "num_tokens": 158319159.0,
      "step": 354
    },
    {
      "epoch": 3.4821648216482166,
      "grad_norm": 0.06424105970441871,
      "learning_rate": 3.0387332884490806e-06,
      "loss": 0.0105,
      "num_tokens": 158768974.0,
      "step": 355
    },
    {
      "epoch": 3.4920049200492005,
      "grad_norm": 0.06970655480927966,
      "learning_rate": 3.014825402217533e-06,
      "loss": 0.0099,
      "num_tokens": 159221319.0,
      "step": 356
    },
    {
      "epoch": 3.5018450184501844,
      "grad_norm": 0.06231852234082556,
      "learning_rate": 2.9910180241384014e-06,
      "loss": 0.0099,
      "num_tokens": 159657431.0,
      "step": 357
    },
    {
      "epoch": 3.5116851168511687,
      "grad_norm": 0.06403174372575768,
      "learning_rate": 2.9673121170557396e-06,
      "loss": 0.0099,
      "num_tokens": 160091184.0,
      "step": 358
    },
    {
      "epoch": 3.5215252152521526,
      "grad_norm": 0.06050506427522611,
      "learning_rate": 2.9437086397097996e-06,
      "loss": 0.0095,
      "num_tokens": 160538104.0,
      "step": 359
    },
    {
      "epoch": 3.5313653136531364,
      "grad_norm": 0.05914580967848918,
      "learning_rate": 2.92020854669826e-06,
      "loss": 0.0151,
      "num_tokens": 160984800.0,
      "step": 360
    },
    {
      "epoch": 3.5412054120541203,
      "grad_norm": 0.06615551474859403,
      "learning_rate": 2.896812788437615e-06,
      "loss": 0.0102,
      "num_tokens": 161437908.0,
      "step": 361
    },
    {
      "epoch": 3.5510455104551046,
      "grad_norm": 0.05688142632929498,
      "learning_rate": 2.8735223111247402e-06,
      "loss": 0.0094,
      "num_tokens": 161900209.0,
      "step": 362
    },
    {
      "epoch": 3.5608856088560885,
      "grad_norm": 0.05805719882416427,
      "learning_rate": 2.850338056698621e-06,
      "loss": 0.0094,
      "num_tokens": 162381378.0,
      "step": 363
    },
    {
      "epoch": 3.570725707257073,
      "grad_norm": 0.05665394777981862,
      "learning_rate": 2.827260962802263e-06,
      "loss": 0.0089,
      "num_tokens": 162818401.0,
      "step": 364
    },
    {
      "epoch": 3.5805658056580567,
      "grad_norm": 0.058540688861597474,
      "learning_rate": 2.804291962744768e-06,
      "loss": 0.0102,
      "num_tokens": 163261663.0,
      "step": 365
    },
    {
      "epoch": 3.5904059040590406,
      "grad_norm": 0.06068364561780823,
      "learning_rate": 2.7814319854635875e-06,
      "loss": 0.0096,
      "num_tokens": 163706510.0,
      "step": 366
    },
    {
      "epoch": 3.6002460024600245,
      "grad_norm": 0.0593859542792967,
      "learning_rate": 2.758681955486955e-06,
      "loss": 0.0097,
      "num_tokens": 164145145.0,
      "step": 367
    },
    {
      "epoch": 3.6100861008610083,
      "grad_norm": 0.059439587082302694,
      "learning_rate": 2.736042792896495e-06,
      "loss": 0.0104,
      "num_tokens": 164588218.0,
      "step": 368
    },
    {
      "epoch": 3.6199261992619927,
      "grad_norm": 0.06426940128348262,
      "learning_rate": 2.7135154132900133e-06,
      "loss": 0.0203,
      "num_tokens": 165039642.0,
      "step": 369
    },
    {
      "epoch": 3.6297662976629765,
      "grad_norm": 0.059031373381084176,
      "learning_rate": 2.691100727744458e-06,
      "loss": 0.0091,
      "num_tokens": 165502439.0,
      "step": 370
    },
    {
      "epoch": 3.639606396063961,
      "grad_norm": 0.05706397506461239,
      "learning_rate": 2.668799642779093e-06,
      "loss": 0.0106,
      "num_tokens": 165957611.0,
      "step": 371
    },
    {
      "epoch": 3.6494464944649447,
      "grad_norm": 0.06337690848780857,
      "learning_rate": 2.6466130603188157e-06,
      "loss": 0.01,
      "num_tokens": 166404741.0,
      "step": 372
    },
    {
      "epoch": 3.6592865928659286,
      "grad_norm": 0.057865704503962175,
      "learning_rate": 2.624541877657685e-06,
      "loss": 0.1951,
      "num_tokens": 166908892.0,
      "step": 373
    },
    {
      "epoch": 3.6691266912669125,
      "grad_norm": 0.6748913551790232,
      "learning_rate": 2.602586987422643e-06,
      "loss": 0.0094,
      "num_tokens": 167346017.0,
      "step": 374
    },
    {
      "epoch": 3.678966789667897,
      "grad_norm": 0.06271310429727074,
      "learning_rate": 2.580749277537399e-06,
      "loss": 0.0093,
      "num_tokens": 167795779.0,
      "step": 375
    },
    {
      "epoch": 3.6888068880688807,
      "grad_norm": 0.05728241738284472,
      "learning_rate": 2.5590296311865294e-06,
      "loss": 0.0092,
      "num_tokens": 168246613.0,
      "step": 376
    },
    {
      "epoch": 3.6986469864698646,
      "grad_norm": 0.05730319671770116,
      "learning_rate": 2.537428926779758e-06,
      "loss": 0.0104,
      "num_tokens": 168703193.0,
      "step": 377
    },
    {
      "epoch": 3.708487084870849,
      "grad_norm": 0.061789009881383514,
      "learning_rate": 2.515948037916423e-06,
      "loss": 0.0104,
      "num_tokens": 169166239.0,
      "step": 378
    },
    {
      "epoch": 3.7183271832718328,
      "grad_norm": 0.05958784070544453,
      "learning_rate": 2.494587833350153e-06,
      "loss": 0.0564,
      "num_tokens": 169618415.0,
      "step": 379
    },
    {
      "epoch": 3.7281672816728166,
      "grad_norm": 0.22039415728368103,
      "learning_rate": 2.473349176953736e-06,
      "loss": 0.0094,
      "num_tokens": 170079318.0,
      "step": 380
    },
    {
      "epoch": 3.7380073800738005,
      "grad_norm": 0.05930397129828618,
      "learning_rate": 2.4522329276841664e-06,
      "loss": 0.0198,
      "num_tokens": 170524571.0,
      "step": 381
    },
    {
      "epoch": 3.747847478474785,
      "grad_norm": 0.06047568038440854,
      "learning_rate": 2.431239939547921e-06,
      "loss": 0.0094,
      "num_tokens": 170983016.0,
      "step": 382
    },
    {
      "epoch": 3.7576875768757687,
      "grad_norm": 0.061680315681806853,
      "learning_rate": 2.4103710615664145e-06,
      "loss": 0.0089,
      "num_tokens": 171426486.0,
      "step": 383
    },
    {
      "epoch": 3.767527675276753,
      "grad_norm": 0.05588539351574886,
      "learning_rate": 2.389627137741662e-06,
      "loss": 0.0094,
      "num_tokens": 171871834.0,
      "step": 384
    },
    {
      "epoch": 3.777367773677737,
      "grad_norm": 0.061780123368904795,
      "learning_rate": 2.369009007022146e-06,
      "loss": 0.0093,
      "num_tokens": 172337523.0,
      "step": 385
    },
    {
      "epoch": 3.787207872078721,
      "grad_norm": 0.05632561272908436,
      "learning_rate": 2.3485175032688865e-06,
      "loss": 0.0088,
      "num_tokens": 172775826.0,
      "step": 386
    },
    {
      "epoch": 3.7970479704797047,
      "grad_norm": 0.058782272770165275,
      "learning_rate": 2.328153455221717e-06,
      "loss": 0.0095,
      "num_tokens": 173234709.0,
      "step": 387
    },
    {
      "epoch": 3.8068880688806885,
      "grad_norm": 0.057526356469471435,
      "learning_rate": 2.3079176864657673e-06,
      "loss": 0.0097,
      "num_tokens": 173700055.0,
      "step": 388
    },
    {
      "epoch": 3.816728167281673,
      "grad_norm": 0.06609619441495819,
      "learning_rate": 2.2878110153981565e-06,
      "loss": 0.0111,
      "num_tokens": 174147961.0,
      "step": 389
    },
    {
      "epoch": 3.8265682656826567,
      "grad_norm": 0.06703233332357492,
      "learning_rate": 2.267834255194894e-06,
      "loss": 0.0116,
      "num_tokens": 174586991.0,
      "step": 390
    },
    {
      "epoch": 3.836408364083641,
      "grad_norm": 0.06522848493729735,
      "learning_rate": 2.2479882137779903e-06,
      "loss": 0.0106,
      "num_tokens": 175006875.0,
      "step": 391
    },
    {
      "epoch": 3.846248462484625,
      "grad_norm": 0.06306752932488521,
      "learning_rate": 2.228273693782784e-06,
      "loss": 0.0094,
      "num_tokens": 175451007.0,
      "step": 392
    },
    {
      "epoch": 3.856088560885609,
      "grad_norm": 0.062263756072231294,
      "learning_rate": 2.208691492525481e-06,
      "loss": 0.0135,
      "num_tokens": 175896902.0,
      "step": 393
    },
    {
      "epoch": 3.8659286592865927,
      "grad_norm": 0.06835430681220003,
      "learning_rate": 2.189242401970908e-06,
      "loss": 0.0092,
      "num_tokens": 176346616.0,
      "step": 394
    },
    {
      "epoch": 3.875768757687577,
      "grad_norm": 0.05728313379563115,
      "learning_rate": 2.169927208700482e-06,
      "loss": 0.0098,
      "num_tokens": 176802124.0,
      "step": 395
    },
    {
      "epoch": 3.885608856088561,
      "grad_norm": 0.06299115193931754,
      "learning_rate": 2.1507466938804013e-06,
      "loss": 0.0089,
      "num_tokens": 177233961.0,
      "step": 396
    },
    {
      "epoch": 3.8954489544895448,
      "grad_norm": 0.060076198285498296,
      "learning_rate": 2.131701633230045e-06,
      "loss": 0.0098,
      "num_tokens": 177684662.0,
      "step": 397
    },
    {
      "epoch": 3.905289052890529,
      "grad_norm": 0.06517531508961912,
      "learning_rate": 2.112792796990616e-06,
      "loss": 0.0095,
      "num_tokens": 178123825.0,
      "step": 398
    },
    {
      "epoch": 3.915129151291513,
      "grad_norm": 0.05863263973572925,
      "learning_rate": 2.0940209498939732e-06,
      "loss": 0.009,
      "num_tokens": 178562641.0,
      "step": 399
    },
    {
      "epoch": 3.924969249692497,
      "grad_norm": 0.05798991563312477,
      "learning_rate": 2.075386851131711e-06,
      "loss": 0.0094,
      "num_tokens": 179007017.0,
      "step": 400
    },
    {
      "epoch": 3.9348093480934807,
      "grad_norm": 0.06118488260559937,
      "learning_rate": 2.056891254324459e-06,
      "loss": 0.0095,
      "num_tokens": 179449125.0,
      "step": 401
    },
    {
      "epoch": 3.944649446494465,
      "grad_norm": 0.06403534407994695,
      "learning_rate": 2.038534907491396e-06,
      "loss": 0.009,
      "num_tokens": 179887646.0,
      "step": 402
    },
    {
      "epoch": 3.954489544895449,
      "grad_norm": 0.08058699039926022,
      "learning_rate": 2.0203185530199983e-06,
      "loss": 0.0138,
      "num_tokens": 180341944.0,
      "step": 403
    },
    {
      "epoch": 3.9643296432964332,
      "grad_norm": 0.056026267406971995,
      "learning_rate": 2.0022429276360256e-06,
      "loss": 0.0097,
      "num_tokens": 180787775.0,
      "step": 404
    },
    {
      "epoch": 3.974169741697417,
      "grad_norm": 0.058787256460149456,
      "learning_rate": 1.9843087623737097e-06,
      "loss": 0.0088,
      "num_tokens": 181276015.0,
      "step": 405
    },
    {
      "epoch": 3.984009840098401,
      "grad_norm": 0.054638072869340186,
      "learning_rate": 1.966516782546199e-06,
      "loss": 0.009,
      "num_tokens": 181724759.0,
      "step": 406
    },
    {
      "epoch": 3.993849938499385,
      "grad_norm": 0.05931097745374889,
      "learning_rate": 1.94886770771623e-06,
      "loss": 0.0098,
      "num_tokens": 182165821.0,
      "step": 407
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.06697953375930626,
      "learning_rate": 1.931362251667008e-06,
      "loss": 0.027,
      "num_tokens": 182364260.0,
      "step": 408
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.1028980016708374,
      "eval_num_tokens": 182364260.0,
      "eval_runtime": 53.8919,
      "eval_samples_per_second": 41.676,
      "eval_steps_per_second": 5.214,
      "step": 408
    },
    {
      "epoch": 4.009840098400984,
      "grad_norm": 0.07632643003764507,
      "learning_rate": 1.9140011223733576e-06,
      "loss": 0.0082,
      "num_tokens": 182806025.0,
      "step": 409
    },
    {
      "epoch": 4.019680196801968,
      "grad_norm": 0.05437436276939388,
      "learning_rate": 1.8967850219730799e-06,
      "loss": 0.0081,
      "num_tokens": 183278654.0,
      "step": 410
    },
    {
      "epoch": 4.029520295202952,
      "grad_norm": 0.05114318878211908,
      "learning_rate": 1.8797146467385604e-06,
      "loss": 0.0076,
      "num_tokens": 183720645.0,
      "step": 411
    },
    {
      "epoch": 4.039360393603936,
      "grad_norm": 0.053465044974803935,
      "learning_rate": 1.8627906870486063e-06,
      "loss": 0.0082,
      "num_tokens": 184191637.0,
      "step": 412
    },
    {
      "epoch": 4.04920049200492,
      "grad_norm": 0.054542981072468875,
      "learning_rate": 1.8460138273605265e-06,
      "loss": 0.008,
      "num_tokens": 184634141.0,
      "step": 413
    },
    {
      "epoch": 4.059040590405904,
      "grad_norm": 0.052414283521576004,
      "learning_rate": 1.8293847461824538e-06,
      "loss": 0.0079,
      "num_tokens": 185081741.0,
      "step": 414
    },
    {
      "epoch": 4.068880688806888,
      "grad_norm": 0.05289967674124652,
      "learning_rate": 1.8129041160458966e-06,
      "loss": 0.008,
      "num_tokens": 185495440.0,
      "step": 415
    },
    {
      "epoch": 4.078720787207872,
      "grad_norm": 0.0584668942852983,
      "learning_rate": 1.7965726034785466e-06,
      "loss": 0.0081,
      "num_tokens": 185938291.0,
      "step": 416
    },
    {
      "epoch": 4.088560885608856,
      "grad_norm": 0.05897150659800833,
      "learning_rate": 1.780390868977318e-06,
      "loss": 0.0086,
      "num_tokens": 186409542.0,
      "step": 417
    },
    {
      "epoch": 4.0984009840098405,
      "grad_norm": 0.05118034680985974,
      "learning_rate": 1.7643595669816378e-06,
      "loss": 0.0077,
      "num_tokens": 186852482.0,
      "step": 418
    },
    {
      "epoch": 4.108241082410824,
      "grad_norm": 0.05911903344070817,
      "learning_rate": 1.7484793458469745e-06,
      "loss": 0.0081,
      "num_tokens": 187306570.0,
      "step": 419
    },
    {
      "epoch": 4.118081180811808,
      "grad_norm": 0.058617479568280846,
      "learning_rate": 1.7327508478186216e-06,
      "loss": 0.0075,
      "num_tokens": 187738802.0,
      "step": 420
    },
    {
      "epoch": 4.127921279212792,
      "grad_norm": 0.05743950460862962,
      "learning_rate": 1.7171747090057201e-06,
      "loss": 0.0081,
      "num_tokens": 188188275.0,
      "step": 421
    },
    {
      "epoch": 4.137761377613776,
      "grad_norm": 0.0578427653677817,
      "learning_rate": 1.7017515593555295e-06,
      "loss": 0.008,
      "num_tokens": 188626310.0,
      "step": 422
    },
    {
      "epoch": 4.14760147601476,
      "grad_norm": 0.055381917249045204,
      "learning_rate": 1.6864820226279607e-06,
      "loss": 0.0079,
      "num_tokens": 189058824.0,
      "step": 423
    },
    {
      "epoch": 4.157441574415744,
      "grad_norm": 0.0566904301682134,
      "learning_rate": 1.6713667163703348e-06,
      "loss": 0.008,
      "num_tokens": 189488025.0,
      "step": 424
    },
    {
      "epoch": 4.167281672816729,
      "grad_norm": 0.0591657691393218,
      "learning_rate": 1.6564062518924202e-06,
      "loss": 0.0093,
      "num_tokens": 189949176.0,
      "step": 425
    },
    {
      "epoch": 4.177121771217712,
      "grad_norm": 0.058609260537066755,
      "learning_rate": 1.6416012342417056e-06,
      "loss": 0.0075,
      "num_tokens": 190405187.0,
      "step": 426
    },
    {
      "epoch": 4.186961869618696,
      "grad_norm": 0.05376660491247955,
      "learning_rate": 1.6269522621789246e-06,
      "loss": 0.0094,
      "num_tokens": 190839466.0,
      "step": 427
    },
    {
      "epoch": 4.19680196801968,
      "grad_norm": 0.062048025442225076,
      "learning_rate": 1.6124599281538452e-06,
      "loss": 0.02,
      "num_tokens": 191280153.0,
      "step": 428
    },
    {
      "epoch": 4.206642066420664,
      "grad_norm": 0.06071173185238267,
      "learning_rate": 1.5981248182813136e-06,
      "loss": 0.0073,
      "num_tokens": 191734314.0,
      "step": 429
    },
    {
      "epoch": 4.216482164821648,
      "grad_norm": 0.05301725414979279,
      "learning_rate": 1.583947512317537e-06,
      "loss": 0.0117,
      "num_tokens": 192202492.0,
      "step": 430
    },
    {
      "epoch": 4.226322263222632,
      "grad_norm": 0.06832062526218917,
      "learning_rate": 1.5699285836366488e-06,
      "loss": 0.0093,
      "num_tokens": 192667915.0,
      "step": 431
    },
    {
      "epoch": 4.236162361623617,
      "grad_norm": 0.05748762603533909,
      "learning_rate": 1.5560685992075141e-06,
      "loss": 0.0078,
      "num_tokens": 193136794.0,
      "step": 432
    },
    {
      "epoch": 4.2460024600246005,
      "grad_norm": 0.0737572203685775,
      "learning_rate": 1.5423681195707997e-06,
      "loss": 0.0073,
      "num_tokens": 193598491.0,
      "step": 433
    },
    {
      "epoch": 4.255842558425584,
      "grad_norm": 0.05225082250599676,
      "learning_rate": 1.528827698816306e-06,
      "loss": 0.0077,
      "num_tokens": 194023980.0,
      "step": 434
    },
    {
      "epoch": 4.265682656826568,
      "grad_norm": 0.05296466266803098,
      "learning_rate": 1.515447884560556e-06,
      "loss": 0.0074,
      "num_tokens": 194481167.0,
      "step": 435
    },
    {
      "epoch": 4.275522755227552,
      "grad_norm": 0.05336380722303185,
      "learning_rate": 1.502229217924649e-06,
      "loss": 0.0075,
      "num_tokens": 194915312.0,
      "step": 436
    },
    {
      "epoch": 4.285362853628536,
      "grad_norm": 0.05458180686808586,
      "learning_rate": 1.489172233512376e-06,
      "loss": 0.0076,
      "num_tokens": 195368266.0,
      "step": 437
    },
    {
      "epoch": 4.29520295202952,
      "grad_norm": 0.05542603913086383,
      "learning_rate": 1.4762774593885986e-06,
      "loss": 0.0081,
      "num_tokens": 195810914.0,
      "step": 438
    },
    {
      "epoch": 4.305043050430505,
      "grad_norm": 0.054344537083576325,
      "learning_rate": 1.4635454170578917e-06,
      "loss": 0.0072,
      "num_tokens": 196263940.0,
      "step": 439
    },
    {
      "epoch": 4.3148831488314885,
      "grad_norm": 0.052701156778993646,
      "learning_rate": 1.4509766214434535e-06,
      "loss": 0.0077,
      "num_tokens": 196718774.0,
      "step": 440
    },
    {
      "epoch": 4.324723247232472,
      "grad_norm": 0.05423178707270067,
      "learning_rate": 1.4385715808662787e-06,
      "loss": 0.008,
      "num_tokens": 197161519.0,
      "step": 441
    },
    {
      "epoch": 4.334563345633456,
      "grad_norm": 0.055354896441224044,
      "learning_rate": 1.4263307970246027e-06,
      "loss": 0.008,
      "num_tokens": 197621081.0,
      "step": 442
    },
    {
      "epoch": 4.34440344403444,
      "grad_norm": 0.05816305513011695,
      "learning_rate": 1.41425476497361e-06,
      "loss": 0.0078,
      "num_tokens": 198087857.0,
      "step": 443
    },
    {
      "epoch": 4.354243542435424,
      "grad_norm": 0.05127845466920968,
      "learning_rate": 1.4023439731054112e-06,
      "loss": 0.0077,
      "num_tokens": 198533672.0,
      "step": 444
    },
    {
      "epoch": 4.364083640836409,
      "grad_norm": 0.07067731738580797,
      "learning_rate": 1.390598903129296e-06,
      "loss": 0.0322,
      "num_tokens": 199022227.0,
      "step": 445
    },
    {
      "epoch": 4.373923739237393,
      "grad_norm": 0.05511218194004341,
      "learning_rate": 1.3790200300522413e-06,
      "loss": 0.0077,
      "num_tokens": 199462215.0,
      "step": 446
    },
    {
      "epoch": 4.3837638376383765,
      "grad_norm": 0.05735730379081794,
      "learning_rate": 1.3676078221597157e-06,
      "loss": 0.0074,
      "num_tokens": 199907231.0,
      "step": 447
    },
    {
      "epoch": 4.39360393603936,
      "grad_norm": 0.05442936039834661,
      "learning_rate": 1.3563627409967257e-06,
      "loss": 0.1955,
      "num_tokens": 200376904.0,
      "step": 448
    },
    {
      "epoch": 4.403444034440344,
      "grad_norm": 0.5930661652942222,
      "learning_rate": 1.3452852413491563e-06,
      "loss": 0.0074,
      "num_tokens": 200853967.0,
      "step": 449
    },
    {
      "epoch": 4.413284132841328,
      "grad_norm": 0.05077867679984549,
      "learning_rate": 1.3343757712253804e-06,
      "loss": 0.0076,
      "num_tokens": 201323621.0,
      "step": 450
    },
    {
      "epoch": 4.423124231242312,
      "grad_norm": 0.058807424527887606,
      "learning_rate": 1.3236347718381338e-06,
      "loss": 0.0096,
      "num_tokens": 201753687.0,
      "step": 451
    },
    {
      "epoch": 4.432964329643297,
      "grad_norm": 0.06001374322910319,
      "learning_rate": 1.3130626775866743e-06,
      "loss": 0.0081,
      "num_tokens": 202203799.0,
      "step": 452
    },
    {
      "epoch": 4.442804428044281,
      "grad_norm": 0.06273437087252197,
      "learning_rate": 1.3026599160392173e-06,
      "loss": 0.0092,
      "num_tokens": 202627243.0,
      "step": 453
    },
    {
      "epoch": 4.4526445264452645,
      "grad_norm": 0.06372618537836224,
      "learning_rate": 1.292426907915634e-06,
      "loss": 0.0076,
      "num_tokens": 203077433.0,
      "step": 454
    },
    {
      "epoch": 4.462484624846248,
      "grad_norm": 0.057948321757535656,
      "learning_rate": 1.2823640670704443e-06,
      "loss": 0.0229,
      "num_tokens": 203532517.0,
      "step": 455
    },
    {
      "epoch": 4.472324723247232,
      "grad_norm": 0.06607138604150303,
      "learning_rate": 1.2724718004760794e-06,
      "loss": 0.0078,
      "num_tokens": 203967752.0,
      "step": 456
    },
    {
      "epoch": 4.482164821648216,
      "grad_norm": 0.05725783304801458,
      "learning_rate": 1.2627505082064144e-06,
      "loss": 0.0076,
      "num_tokens": 204424349.0,
      "step": 457
    },
    {
      "epoch": 4.492004920049201,
      "grad_norm": 0.055427831791831646,
      "learning_rate": 1.2532005834205976e-06,
      "loss": 0.0079,
      "num_tokens": 204846138.0,
      "step": 458
    },
    {
      "epoch": 4.501845018450185,
      "grad_norm": 0.05460191637217484,
      "learning_rate": 1.2438224123471442e-06,
      "loss": 0.0192,
      "num_tokens": 205306730.0,
      "step": 459
    },
    {
      "epoch": 4.511685116851169,
      "grad_norm": 0.06279438477449967,
      "learning_rate": 1.2346163742683185e-06,
      "loss": 0.0117,
      "num_tokens": 205759609.0,
      "step": 460
    },
    {
      "epoch": 4.521525215252153,
      "grad_norm": 0.05702285396092694,
      "learning_rate": 1.2255828415047932e-06,
      "loss": 0.0076,
      "num_tokens": 206171295.0,
      "step": 461
    },
    {
      "epoch": 4.531365313653136,
      "grad_norm": 0.054521558454890394,
      "learning_rate": 1.216722179400592e-06,
      "loss": 0.0076,
      "num_tokens": 206639148.0,
      "step": 462
    },
    {
      "epoch": 4.54120541205412,
      "grad_norm": 0.05168283263697403,
      "learning_rate": 1.208034746308315e-06,
      "loss": 0.0068,
      "num_tokens": 207094260.0,
      "step": 463
    },
    {
      "epoch": 4.551045510455104,
      "grad_norm": 0.05161429329359664,
      "learning_rate": 1.1995208935746437e-06,
      "loss": 0.0081,
      "num_tokens": 207533375.0,
      "step": 464
    },
    {
      "epoch": 4.560885608856088,
      "grad_norm": 0.058514508257411606,
      "learning_rate": 1.1911809655261333e-06,
      "loss": 0.0081,
      "num_tokens": 207969517.0,
      "step": 465
    },
    {
      "epoch": 4.570725707257073,
      "grad_norm": 0.056665893017668854,
      "learning_rate": 1.1830152994552866e-06,
      "loss": 0.0086,
      "num_tokens": 208408117.0,
      "step": 466
    },
    {
      "epoch": 4.580565805658057,
      "grad_norm": 0.056163462620316754,
      "learning_rate": 1.175024225606912e-06,
      "loss": 0.0074,
      "num_tokens": 208879227.0,
      "step": 467
    },
    {
      "epoch": 4.590405904059041,
      "grad_norm": 0.05409385523794747,
      "learning_rate": 1.1672080671647695e-06,
      "loss": 0.0078,
      "num_tokens": 209325103.0,
      "step": 468
    },
    {
      "epoch": 4.6002460024600245,
      "grad_norm": 0.05629255243399504,
      "learning_rate": 1.1595671402384966e-06,
      "loss": 0.0102,
      "num_tokens": 209791894.0,
      "step": 469
    },
    {
      "epoch": 4.610086100861008,
      "grad_norm": 0.051104203707396316,
      "learning_rate": 1.152101753850828e-06,
      "loss": 0.0072,
      "num_tokens": 210254182.0,
      "step": 470
    },
    {
      "epoch": 4.619926199261993,
      "grad_norm": 0.05229454749737629,
      "learning_rate": 1.1448122099250946e-06,
      "loss": 0.0104,
      "num_tokens": 210702900.0,
      "step": 471
    },
    {
      "epoch": 4.629766297662977,
      "grad_norm": 0.060177504722208404,
      "learning_rate": 1.1376988032730135e-06,
      "loss": 0.0079,
      "num_tokens": 211151465.0,
      "step": 472
    },
    {
      "epoch": 4.639606396063961,
      "grad_norm": 0.05182456184289124,
      "learning_rate": 1.130761821582766e-06,
      "loss": 0.0072,
      "num_tokens": 211619464.0,
      "step": 473
    },
    {
      "epoch": 4.649446494464945,
      "grad_norm": 0.05574225668849545,
      "learning_rate": 1.1240015454073622e-06,
      "loss": 0.0085,
      "num_tokens": 212064266.0,
      "step": 474
    },
    {
      "epoch": 4.659286592865929,
      "grad_norm": 0.06359820975154429,
      "learning_rate": 1.1174182481532943e-06,
      "loss": 0.0081,
      "num_tokens": 212499724.0,
      "step": 475
    },
    {
      "epoch": 4.6691266912669125,
      "grad_norm": 0.05622656000305094,
      "learning_rate": 1.1110121960694773e-06,
      "loss": 0.0079,
      "num_tokens": 212945879.0,
      "step": 476
    },
    {
      "epoch": 4.678966789667896,
      "grad_norm": 0.06093763072714235,
      "learning_rate": 1.104783648236486e-06,
      "loss": 0.0084,
      "num_tokens": 213379787.0,
      "step": 477
    },
    {
      "epoch": 4.68880688806888,
      "grad_norm": 0.0543614373855231,
      "learning_rate": 1.0987328565560711e-06,
      "loss": 0.0075,
      "num_tokens": 213824263.0,
      "step": 478
    },
    {
      "epoch": 4.698646986469865,
      "grad_norm": 0.056905167227697236,
      "learning_rate": 1.0928600657409751e-06,
      "loss": 0.0082,
      "num_tokens": 214265208.0,
      "step": 479
    },
    {
      "epoch": 4.708487084870849,
      "grad_norm": 0.057351833542733925,
      "learning_rate": 1.0871655133050372e-06,
      "loss": 0.0082,
      "num_tokens": 214744301.0,
      "step": 480
    },
    {
      "epoch": 4.718327183271833,
      "grad_norm": 0.29349816338215157,
      "learning_rate": 1.081649429553581e-06,
      "loss": 0.0553,
      "num_tokens": 215194355.0,
      "step": 481
    },
    {
      "epoch": 4.728167281672817,
      "grad_norm": 0.051057953015104116,
      "learning_rate": 1.076312037574106e-06,
      "loss": 0.0074,
      "num_tokens": 215632060.0,
      "step": 482
    },
    {
      "epoch": 4.7380073800738005,
      "grad_norm": 0.056594540815463674,
      "learning_rate": 1.0711535532272632e-06,
      "loss": 0.0235,
      "num_tokens": 216097276.0,
      "step": 483
    },
    {
      "epoch": 4.747847478474784,
      "grad_norm": 0.068871190152495,
      "learning_rate": 1.0661741851381256e-06,
      "loss": 0.0077,
      "num_tokens": 216544463.0,
      "step": 484
    },
    {
      "epoch": 4.757687576875769,
      "grad_norm": 0.05907548729697175,
      "learning_rate": 1.0613741346877498e-06,
      "loss": 0.0084,
      "num_tokens": 216972058.0,
      "step": 485
    },
    {
      "epoch": 4.767527675276753,
      "grad_norm": 0.055592377746762095,
      "learning_rate": 1.056753596005032e-06,
      "loss": 0.0074,
      "num_tokens": 217401900.0,
      "step": 486
    },
    {
      "epoch": 4.777367773677737,
      "grad_norm": 0.05562394957573223,
      "learning_rate": 1.0523127559588579e-06,
      "loss": 0.0075,
      "num_tokens": 217845453.0,
      "step": 487
    },
    {
      "epoch": 4.787207872078721,
      "grad_norm": 0.05258367575789477,
      "learning_rate": 1.0480517941505428e-06,
      "loss": 0.0073,
      "num_tokens": 218272871.0,
      "step": 488
    },
    {
      "epoch": 4.797047970479705,
      "grad_norm": 0.05390618674507445,
      "learning_rate": 1.0439708829065708e-06,
      "loss": 0.0078,
      "num_tokens": 218732597.0,
      "step": 489
    },
    {
      "epoch": 4.8068880688806885,
      "grad_norm": 0.06946151381547928,
      "learning_rate": 1.0400701872716227e-06,
      "loss": 0.0223,
      "num_tokens": 219194340.0,
      "step": 490
    },
    {
      "epoch": 4.816728167281672,
      "grad_norm": 0.05582170906207444,
      "learning_rate": 1.0363498650019023e-06,
      "loss": 0.0077,
      "num_tokens": 219673692.0,
      "step": 491
    },
    {
      "epoch": 4.826568265682657,
      "grad_norm": 0.05244987983803676,
      "learning_rate": 1.0328100665587573e-06,
      "loss": 0.0073,
      "num_tokens": 220118246.0,
      "step": 492
    },
    {
      "epoch": 4.836408364083641,
      "grad_norm": 0.055024340070040305,
      "learning_rate": 1.029450935102592e-06,
      "loss": 0.0077,
      "num_tokens": 220555806.0,
      "step": 493
    },
    {
      "epoch": 4.846248462484625,
      "grad_norm": 0.05338628090134423,
      "learning_rate": 1.0262726064870801e-06,
      "loss": 0.0073,
      "num_tokens": 220997187.0,
      "step": 494
    },
    {
      "epoch": 4.856088560885609,
      "grad_norm": 0.058254094197714025,
      "learning_rate": 1.0232752092536666e-06,
      "loss": 0.0074,
      "num_tokens": 221434681.0,
      "step": 495
    },
    {
      "epoch": 4.865928659286593,
      "grad_norm": 0.05261616134189719,
      "learning_rate": 1.0204588646263731e-06,
      "loss": 0.0074,
      "num_tokens": 221884850.0,
      "step": 496
    },
    {
      "epoch": 4.875768757687577,
      "grad_norm": 0.052167915998619634,
      "learning_rate": 1.0178236865068933e-06,
      "loss": 0.0072,
      "num_tokens": 222333225.0,
      "step": 497
    },
    {
      "epoch": 4.885608856088561,
      "grad_norm": 0.06187153122740552,
      "learning_rate": 1.0153697814699858e-06,
      "loss": 0.0106,
      "num_tokens": 222774591.0,
      "step": 498
    },
    {
      "epoch": 4.895448954489545,
      "grad_norm": 0.054905669170180534,
      "learning_rate": 1.0130972487591658e-06,
      "loss": 0.0112,
      "num_tokens": 223227943.0,
      "step": 499
    },
    {
      "epoch": 4.905289052890529,
      "grad_norm": 0.06206228565326619,
      "learning_rate": 1.0110061802826889e-06,
      "loss": 0.0076,
      "num_tokens": 223680989.0,
      "step": 500
    },
    {
      "epoch": 4.915129151291513,
      "grad_norm": 0.05437071230251554,
      "learning_rate": 1.009096660609837e-06,
      "loss": 0.1789,
      "num_tokens": 224171724.0,
      "step": 501
    },
    {
      "epoch": 4.924969249692497,
      "grad_norm": 0.12358300885271949,
      "learning_rate": 1.0073687669674949e-06,
      "loss": 0.0081,
      "num_tokens": 224621243.0,
      "step": 502
    },
    {
      "epoch": 4.934809348093481,
      "grad_norm": 0.05743551551374671,
      "learning_rate": 1.0058225692370299e-06,
      "loss": 0.0077,
      "num_tokens": 225053570.0,
      "step": 503
    },
    {
      "epoch": 4.944649446494465,
      "grad_norm": 0.05705289715957623,
      "learning_rate": 1.0044581299514638e-06,
      "loss": 0.0077,
      "num_tokens": 225475922.0,
      "step": 504
    },
    {
      "epoch": 4.9544895448954485,
      "grad_norm": 0.052608564457681,
      "learning_rate": 1.003275504292944e-06,
      "loss": 0.0072,
      "num_tokens": 225944888.0,
      "step": 505
    },
    {
      "epoch": 4.964329643296433,
      "grad_norm": 0.05546452983023311,
      "learning_rate": 1.0022747400905126e-06,
      "loss": 0.0079,
      "num_tokens": 226384045.0,
      "step": 506
    },
    {
      "epoch": 4.974169741697417,
      "grad_norm": 0.05754539826487939,
      "learning_rate": 1.0014558778181714e-06,
      "loss": 0.0073,
      "num_tokens": 226815343.0,
      "step": 507
    },
    {
      "epoch": 4.984009840098401,
      "grad_norm": 0.05456913560891108,
      "learning_rate": 1.0008189505932444e-06,
      "loss": 0.0084,
      "num_tokens": 227286168.0,
      "step": 508
    },
    {
      "epoch": 4.993849938499385,
      "grad_norm": 0.053799541560384294,
      "learning_rate": 1.0003639841750404e-06,
      "loss": 0.0076,
      "num_tokens": 227746824.0,
      "step": 509
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.07884368824115337,
      "learning_rate": 1.0000909969638097e-06,
      "loss": 0.0089,
      "num_tokens": 227957450.0,
      "step": 510
    },
    {
      "epoch": 5.0,
      "eval_loss": 0.11205815523862839,
      "eval_num_tokens": 227957450.0,
      "eval_runtime": 53.843,
      "eval_samples_per_second": 41.714,
      "eval_steps_per_second": 5.219,
      "step": 510
    },
    {
      "epoch": 5.0,
      "step": 510,
      "total_flos": 7.689061516716278e+17,
      "train_loss": 0.0504409685922677,
      "train_runtime": 7612.3259,
      "train_samples_per_second": 8.537,
      "train_steps_per_second": 0.067
    }
  ],
  "logging_steps": 1,
  "max_steps": 510,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "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": 7.689061516716278e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}