{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.901639344262295,
  "eval_steps": 500,
  "global_step": 120,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.032520325203252036,
      "grad_norm": 40.474364127833894,
      "learning_rate": 4e-07,
      "loss": 2.8277,
      "step": 1
    },
    {
      "epoch": 0.06504065040650407,
      "grad_norm": 39.80986919634387,
      "learning_rate": 8e-07,
      "loss": 2.7894,
      "step": 2
    },
    {
      "epoch": 0.0975609756097561,
      "grad_norm": 37.613931575844546,
      "learning_rate": 1.2e-06,
      "loss": 2.7705,
      "step": 3
    },
    {
      "epoch": 0.13008130081300814,
      "grad_norm": 34.058960023767305,
      "learning_rate": 1.6e-06,
      "loss": 2.6445,
      "step": 4
    },
    {
      "epoch": 0.16260162601626016,
      "grad_norm": 27.17276921169534,
      "learning_rate": 2e-06,
      "loss": 2.6669,
      "step": 5
    },
    {
      "epoch": 0.1951219512195122,
      "grad_norm": 15.411738818229358,
      "learning_rate": 1.9996268812619105e-06,
      "loss": 2.449,
      "step": 6
    },
    {
      "epoch": 0.22764227642276422,
      "grad_norm": 9.722583741958005,
      "learning_rate": 1.998507803482828e-06,
      "loss": 2.492,
      "step": 7
    },
    {
      "epoch": 0.2601626016260163,
      "grad_norm": 5.002874583844784,
      "learning_rate": 1.9966436017605294e-06,
      "loss": 2.3573,
      "step": 8
    },
    {
      "epoch": 0.2926829268292683,
      "grad_norm": 4.38783321163898,
      "learning_rate": 1.9940356672322033e-06,
      "loss": 2.2742,
      "step": 9
    },
    {
      "epoch": 0.3252032520325203,
      "grad_norm": 4.050625126869575,
      "learning_rate": 1.9906859460363304e-06,
      "loss": 2.3301,
      "step": 10
    },
    {
      "epoch": 0.35772357723577236,
      "grad_norm": 3.6789133162460774,
      "learning_rate": 1.986596937860402e-06,
      "loss": 2.2729,
      "step": 11
    },
    {
      "epoch": 0.3902439024390244,
      "grad_norm": 2.613575263087869,
      "learning_rate": 1.9817716940755586e-06,
      "loss": 2.3178,
      "step": 12
    },
    {
      "epoch": 0.42276422764227645,
      "grad_norm": 2.83098515959453,
      "learning_rate": 1.9762138154595447e-06,
      "loss": 2.3486,
      "step": 13
    },
    {
      "epoch": 0.45528455284552843,
      "grad_norm": 2.5132095083333734,
      "learning_rate": 1.969927449509671e-06,
      "loss": 2.273,
      "step": 14
    },
    {
      "epoch": 0.4878048780487805,
      "grad_norm": 2.137375987210218,
      "learning_rate": 1.9629172873477994e-06,
      "loss": 2.224,
      "step": 15
    },
    {
      "epoch": 0.5203252032520326,
      "grad_norm": 1.7542766611994582,
      "learning_rate": 1.955188560219648e-06,
      "loss": 2.3732,
      "step": 16
    },
    {
      "epoch": 0.5528455284552846,
      "grad_norm": 1.490783000173608,
      "learning_rate": 1.9467470355910435e-06,
      "loss": 2.2965,
      "step": 17
    },
    {
      "epoch": 0.5853658536585366,
      "grad_norm": 1.6159290325764584,
      "learning_rate": 1.93759901284402e-06,
      "loss": 2.2234,
      "step": 18
    },
    {
      "epoch": 0.6178861788617886,
      "grad_norm": 1.805845446790275,
      "learning_rate": 1.9277513185759843e-06,
      "loss": 2.2912,
      "step": 19
    },
    {
      "epoch": 0.6504065040650406,
      "grad_norm": 1.4604974196739857,
      "learning_rate": 1.9172113015054528e-06,
      "loss": 2.2642,
      "step": 20
    },
    {
      "epoch": 0.6829268292682927,
      "grad_norm": 1.324880967104045,
      "learning_rate": 1.9059868269881636e-06,
      "loss": 2.2243,
      "step": 21
    },
    {
      "epoch": 0.7154471544715447,
      "grad_norm": 1.1866315224263582,
      "learning_rate": 1.894086271147651e-06,
      "loss": 2.1637,
      "step": 22
    },
    {
      "epoch": 0.7479674796747967,
      "grad_norm": 1.2828999422781817,
      "learning_rate": 1.8815185146246715e-06,
      "loss": 2.3203,
      "step": 23
    },
    {
      "epoch": 0.7804878048780488,
      "grad_norm": 1.1737064917540987,
      "learning_rate": 1.8682929359501337e-06,
      "loss": 2.2612,
      "step": 24
    },
    {
      "epoch": 0.8130081300813008,
      "grad_norm": 1.2633649535555074,
      "learning_rate": 1.8544194045464886e-06,
      "loss": 2.2641,
      "step": 25
    },
    {
      "epoch": 0.8455284552845529,
      "grad_norm": 1.214313162130321,
      "learning_rate": 1.8399082733627965e-06,
      "loss": 2.1903,
      "step": 26
    },
    {
      "epoch": 0.8780487804878049,
      "grad_norm": 1.1791856518491939,
      "learning_rate": 1.8247703711489684e-06,
      "loss": 2.2214,
      "step": 27
    },
    {
      "epoch": 0.9105691056910569,
      "grad_norm": 1.0774382245713374,
      "learning_rate": 1.8090169943749474e-06,
      "loss": 2.2613,
      "step": 28
    },
    {
      "epoch": 0.943089430894309,
      "grad_norm": 1.5312368454111793,
      "learning_rate": 1.792659898800858e-06,
      "loss": 2.2658,
      "step": 29
    },
    {
      "epoch": 0.975609756097561,
      "grad_norm": 1.1478826076751814,
      "learning_rate": 1.7757112907044198e-06,
      "loss": 2.3472,
      "step": 30
    },
    {
      "epoch": 1.008130081300813,
      "grad_norm": 1.109331171095074,
      "learning_rate": 1.7581838177721627e-06,
      "loss": 2.1404,
      "step": 31
    },
    {
      "epoch": 1.040650406504065,
      "grad_norm": 1.1772417429696853,
      "learning_rate": 1.7400905596612518e-06,
      "loss": 2.2428,
      "step": 32
    },
    {
      "epoch": 1.0731707317073171,
      "grad_norm": 1.3879748914777439,
      "learning_rate": 1.7214450182389558e-06,
      "loss": 2.2335,
      "step": 33
    },
    {
      "epoch": 1.1056910569105691,
      "grad_norm": 1.1514419535961085,
      "learning_rate": 1.7022611075070473e-06,
      "loss": 2.1321,
      "step": 34
    },
    {
      "epoch": 1.1382113821138211,
      "grad_norm": 1.034924143668093,
      "learning_rate": 1.682553143218654e-06,
      "loss": 2.1857,
      "step": 35
    },
    {
      "epoch": 1.170731707317073,
      "grad_norm": 1.097512127774361,
      "learning_rate": 1.6623358321953075e-06,
      "loss": 2.2363,
      "step": 36
    },
    {
      "epoch": 1.203252032520325,
      "grad_norm": 1.134434422186745,
      "learning_rate": 1.6416242613521608e-06,
      "loss": 2.2097,
      "step": 37
    },
    {
      "epoch": 1.2357723577235773,
      "grad_norm": 1.0344909435314726,
      "learning_rate": 1.620433886439568e-06,
      "loss": 2.1696,
      "step": 38
    },
    {
      "epoch": 1.2682926829268293,
      "grad_norm": 1.0032777259118921,
      "learning_rate": 1.5987805205094225e-06,
      "loss": 2.1489,
      "step": 39
    },
    {
      "epoch": 1.3008130081300813,
      "grad_norm": 1.0241113986774006,
      "learning_rate": 1.5766803221148673e-06,
      "loss": 2.2635,
      "step": 40
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 1.0602051964794608,
      "learning_rate": 1.5541497832521748e-06,
      "loss": 2.1886,
      "step": 41
    },
    {
      "epoch": 1.3658536585365852,
      "grad_norm": 1.0062361310746801,
      "learning_rate": 1.5312057170538033e-06,
      "loss": 2.2285,
      "step": 42
    },
    {
      "epoch": 1.3983739837398375,
      "grad_norm": 0.9571915548795515,
      "learning_rate": 1.5078652452418061e-06,
      "loss": 2.1125,
      "step": 43
    },
    {
      "epoch": 1.4308943089430894,
      "grad_norm": 1.5795255196962399,
      "learning_rate": 1.4841457853509606e-06,
      "loss": 2.0965,
      "step": 44
    },
    {
      "epoch": 1.4634146341463414,
      "grad_norm": 1.0856704945936266,
      "learning_rate": 1.460065037731152e-06,
      "loss": 2.1168,
      "step": 45
    },
    {
      "epoch": 1.4959349593495934,
      "grad_norm": 1.0631441837149809,
      "learning_rate": 1.435640972338709e-06,
      "loss": 2.2444,
      "step": 46
    },
    {
      "epoch": 1.5284552845528454,
      "grad_norm": 0.988946477349452,
      "learning_rate": 1.4108918153265483e-06,
      "loss": 2.2796,
      "step": 47
    },
    {
      "epoch": 1.5609756097560976,
      "grad_norm": 0.9621462005907288,
      "learning_rate": 1.3858360354431353e-06,
      "loss": 2.1542,
      "step": 48
    },
    {
      "epoch": 1.5934959349593496,
      "grad_norm": 0.9604091441772962,
      "learning_rate": 1.3604923302504146e-06,
      "loss": 2.1484,
      "step": 49
    },
    {
      "epoch": 1.6260162601626016,
      "grad_norm": 1.0567173306723074,
      "learning_rate": 1.334879612170986e-06,
      "loss": 2.1706,
      "step": 50
    },
    {
      "epoch": 1.6585365853658538,
      "grad_norm": 0.9732333696956916,
      "learning_rate": 1.3090169943749473e-06,
      "loss": 2.1718,
      "step": 51
    },
    {
      "epoch": 1.6910569105691056,
      "grad_norm": 1.5486543005243605,
      "learning_rate": 1.28292377651693e-06,
      "loss": 2.1275,
      "step": 52
    },
    {
      "epoch": 1.7235772357723578,
      "grad_norm": 0.996222387540233,
      "learning_rate": 1.2566194303339736e-06,
      "loss": 2.1765,
      "step": 53
    },
    {
      "epoch": 1.7560975609756098,
      "grad_norm": 0.9297566584067041,
      "learning_rate": 1.2301235851149864e-06,
      "loss": 2.0952,
      "step": 54
    },
    {
      "epoch": 1.7886178861788617,
      "grad_norm": 0.949053529830878,
      "learning_rate": 1.203456013052634e-06,
      "loss": 2.1198,
      "step": 55
    },
    {
      "epoch": 1.821138211382114,
      "grad_norm": 0.9585717902192592,
      "learning_rate": 1.1766366144885876e-06,
      "loss": 2.1618,
      "step": 56
    },
    {
      "epoch": 1.8536585365853657,
      "grad_norm": 0.9467147998720515,
      "learning_rate": 1.1496854030631443e-06,
      "loss": 2.2308,
      "step": 57
    },
    {
      "epoch": 1.886178861788618,
      "grad_norm": 1.038472595778763,
      "learning_rate": 1.1226224907802983e-06,
      "loss": 2.2072,
      "step": 58
    },
    {
      "epoch": 1.91869918699187,
      "grad_norm": 1.0155796062739353,
      "learning_rate": 1.09546807299941e-06,
      "loss": 2.1518,
      "step": 59
    },
    {
      "epoch": 1.951219512195122,
      "grad_norm": 1.0214487524954055,
      "learning_rate": 1.068242413364671e-06,
      "loss": 2.1338,
      "step": 60
    },
    {
      "epoch": 1.9837398373983741,
      "grad_norm": 0.9403008925478624,
      "learning_rate": 1.0409658286836142e-06,
      "loss": 2.1361,
      "step": 61
    },
    {
      "epoch": 2.0081967213114753,
      "grad_norm": 1.1388669949384538,
      "learning_rate": 1.013658673765951e-06,
      "loss": 2.171,
      "step": 62
    },
    {
      "epoch": 2.040983606557377,
      "grad_norm": 0.9997896017746226,
      "learning_rate": 9.863413262340491e-07,
      "loss": 2.1322,
      "step": 63
    },
    {
      "epoch": 2.0737704918032787,
      "grad_norm": 0.9423037971088007,
      "learning_rate": 9.590341713163857e-07,
      "loss": 2.1031,
      "step": 64
    },
    {
      "epoch": 2.1065573770491803,
      "grad_norm": 1.0065342001300284,
      "learning_rate": 9.317575866353291e-07,
      "loss": 2.0782,
      "step": 65
    },
    {
      "epoch": 2.139344262295082,
      "grad_norm": 1.0640356979867651,
      "learning_rate": 9.045319270005899e-07,
      "loss": 2.151,
      "step": 66
    },
    {
      "epoch": 2.1721311475409837,
      "grad_norm": 0.96356393117048,
      "learning_rate": 8.773775092197017e-07,
      "loss": 2.1346,
      "step": 67
    },
    {
      "epoch": 2.2049180327868854,
      "grad_norm": 1.0424768882394724,
      "learning_rate": 8.503145969368561e-07,
      "loss": 2.1752,
      "step": 68
    },
    {
      "epoch": 2.237704918032787,
      "grad_norm": 1.0140764180527542,
      "learning_rate": 8.233633855114126e-07,
      "loss": 2.154,
      "step": 69
    },
    {
      "epoch": 2.2704918032786887,
      "grad_norm": 1.018418216015876,
      "learning_rate": 7.965439869473663e-07,
      "loss": 2.1966,
      "step": 70
    },
    {
      "epoch": 2.30327868852459,
      "grad_norm": 1.0128623906395264,
      "learning_rate": 7.698764148850136e-07,
      "loss": 2.1806,
      "step": 71
    },
    {
      "epoch": 2.3360655737704916,
      "grad_norm": 0.9667113834663045,
      "learning_rate": 7.433805696660265e-07,
      "loss": 2.1525,
      "step": 72
    },
    {
      "epoch": 2.3688524590163933,
      "grad_norm": 0.9344509916886253,
      "learning_rate": 7.170762234830698e-07,
      "loss": 2.1361,
      "step": 73
    },
    {
      "epoch": 2.401639344262295,
      "grad_norm": 0.9042845175098417,
      "learning_rate": 6.909830056250526e-07,
      "loss": 2.1161,
      "step": 74
    },
    {
      "epoch": 2.4344262295081966,
      "grad_norm": 0.9514949190478957,
      "learning_rate": 6.651203878290138e-07,
      "loss": 2.0655,
      "step": 75
    },
    {
      "epoch": 2.4672131147540983,
      "grad_norm": 0.8962344593391931,
      "learning_rate": 6.395076697495854e-07,
      "loss": 2.0969,
      "step": 76
    },
    {
      "epoch": 2.5,
      "grad_norm": 1.220856487692698,
      "learning_rate": 6.141639645568645e-07,
      "loss": 2.0064,
      "step": 77
    },
    {
      "epoch": 2.5327868852459017,
      "grad_norm": 1.0699273674702574,
      "learning_rate": 5.891081846734518e-07,
      "loss": 2.2242,
      "step": 78
    },
    {
      "epoch": 2.5655737704918034,
      "grad_norm": 0.9133956662300596,
      "learning_rate": 5.643590276612909e-07,
      "loss": 2.1035,
      "step": 79
    },
    {
      "epoch": 2.598360655737705,
      "grad_norm": 1.2397651626885557,
      "learning_rate": 5.399349622688478e-07,
      "loss": 2.0706,
      "step": 80
    },
    {
      "epoch": 2.6311475409836067,
      "grad_norm": 1.0961123294212762,
      "learning_rate": 5.158542146490399e-07,
      "loss": 2.1598,
      "step": 81
    },
    {
      "epoch": 2.663934426229508,
      "grad_norm": 0.9165116327925802,
      "learning_rate": 4.921347547581939e-07,
      "loss": 2.1695,
      "step": 82
    },
    {
      "epoch": 2.69672131147541,
      "grad_norm": 0.9090092775656153,
      "learning_rate": 4.687942829461968e-07,
      "loss": 2.0438,
      "step": 83
    },
    {
      "epoch": 2.7295081967213113,
      "grad_norm": 0.9896033695086067,
      "learning_rate": 4.458502167478253e-07,
      "loss": 2.1833,
      "step": 84
    },
    {
      "epoch": 2.762295081967213,
      "grad_norm": 0.881034061647054,
      "learning_rate": 4.233196778851329e-07,
      "loss": 2.0399,
      "step": 85
    },
    {
      "epoch": 2.7950819672131146,
      "grad_norm": 0.9375689797050751,
      "learning_rate": 4.0121947949057745e-07,
      "loss": 2.1272,
      "step": 86
    },
    {
      "epoch": 2.8278688524590163,
      "grad_norm": 0.8821398618057048,
      "learning_rate": 3.7956611356043187e-07,
      "loss": 2.0202,
      "step": 87
    },
    {
      "epoch": 2.860655737704918,
      "grad_norm": 0.9748643169208698,
      "learning_rate": 3.5837573864783886e-07,
      "loss": 2.1083,
      "step": 88
    },
    {
      "epoch": 2.8934426229508197,
      "grad_norm": 1.0731341218225272,
      "learning_rate": 3.376641678046925e-07,
      "loss": 2.2301,
      "step": 89
    },
    {
      "epoch": 2.9262295081967213,
      "grad_norm": 0.9215961788843178,
      "learning_rate": 3.174468567813461e-07,
      "loss": 2.0992,
      "step": 90
    },
    {
      "epoch": 2.959016393442623,
      "grad_norm": 2.0795366148561176,
      "learning_rate": 2.977388924929529e-07,
      "loss": 2.0236,
      "step": 91
    },
    {
      "epoch": 2.9918032786885247,
      "grad_norm": 1.1809086784745415,
      "learning_rate": 2.785549817610443e-07,
      "loss": 2.1925,
      "step": 92
    },
    {
      "epoch": 3.0163934426229506,
      "grad_norm": 0.8746993725144835,
      "learning_rate": 2.5990944033874805e-07,
      "loss": 2.0646,
      "step": 93
    },
    {
      "epoch": 3.0491803278688523,
      "grad_norm": 1.03932864954254,
      "learning_rate": 2.418161822278374e-07,
      "loss": 2.1336,
      "step": 94
    },
    {
      "epoch": 3.081967213114754,
      "grad_norm": 1.0972246699016714,
      "learning_rate": 2.2428870929558007e-07,
      "loss": 2.0616,
      "step": 95
    },
    {
      "epoch": 3.1147540983606556,
      "grad_norm": 0.9958626927367774,
      "learning_rate": 2.0734010119914192e-07,
      "loss": 2.1695,
      "step": 96
    },
    {
      "epoch": 3.1475409836065573,
      "grad_norm": 0.9296483485841406,
      "learning_rate": 1.9098300562505264e-07,
      "loss": 2.1026,
      "step": 97
    },
    {
      "epoch": 3.180327868852459,
      "grad_norm": 0.9960968483767407,
      "learning_rate": 1.7522962885103143e-07,
      "loss": 2.0965,
      "step": 98
    },
    {
      "epoch": 3.2131147540983607,
      "grad_norm": 0.9123613741490157,
      "learning_rate": 1.600917266372035e-07,
      "loss": 2.1119,
      "step": 99
    },
    {
      "epoch": 3.2459016393442623,
      "grad_norm": 0.9060636102949841,
      "learning_rate": 1.4558059545351142e-07,
      "loss": 2.0868,
      "step": 100
    },
    {
      "epoch": 3.278688524590164,
      "grad_norm": 0.8985000818503242,
      "learning_rate": 1.3170706404986641e-07,
      "loss": 2.0528,
      "step": 101
    },
    {
      "epoch": 3.3114754098360657,
      "grad_norm": 0.933241393667304,
      "learning_rate": 1.1848148537532842e-07,
      "loss": 2.0586,
      "step": 102
    },
    {
      "epoch": 3.3442622950819674,
      "grad_norm": 0.9051287170275933,
      "learning_rate": 1.0591372885234884e-07,
      "loss": 2.1613,
      "step": 103
    },
    {
      "epoch": 3.3770491803278686,
      "grad_norm": 0.9060065419813347,
      "learning_rate": 9.401317301183653e-08,
      "loss": 2.1464,
      "step": 104
    },
    {
      "epoch": 3.4098360655737707,
      "grad_norm": 0.8843915324359818,
      "learning_rate": 8.278869849454717e-08,
      "loss": 2.1046,
      "step": 105
    },
    {
      "epoch": 3.442622950819672,
      "grad_norm": 0.8898196332033881,
      "learning_rate": 7.224868142401541e-08,
      "loss": 2.0694,
      "step": 106
    },
    {
      "epoch": 3.4754098360655736,
      "grad_norm": 0.9233349730875204,
      "learning_rate": 6.240098715597975e-08,
      "loss": 2.128,
      "step": 107
    },
    {
      "epoch": 3.5081967213114753,
      "grad_norm": 0.9371484306417918,
      "learning_rate": 5.325296440895621e-08,
      "loss": 2.1957,
      "step": 108
    },
    {
      "epoch": 3.540983606557377,
      "grad_norm": 0.9179416563771874,
      "learning_rate": 4.4811439780351954e-08,
      "loss": 2.145,
      "step": 109
    },
    {
      "epoch": 3.5737704918032787,
      "grad_norm": 0.9376004729460732,
      "learning_rate": 3.708271265220087e-08,
      "loss": 2.1227,
      "step": 110
    },
    {
      "epoch": 3.6065573770491803,
      "grad_norm": 0.9511371344993869,
      "learning_rate": 3.0072550490328754e-08,
      "loss": 2.0773,
      "step": 111
    },
    {
      "epoch": 3.639344262295082,
      "grad_norm": 1.0479288992380316,
      "learning_rate": 2.3786184540455445e-08,
      "loss": 2.0928,
      "step": 112
    },
    {
      "epoch": 3.6721311475409837,
      "grad_norm": 1.0663190435639258,
      "learning_rate": 1.8228305924441466e-08,
      "loss": 2.0365,
      "step": 113
    },
    {
      "epoch": 3.7049180327868854,
      "grad_norm": 0.9713642517206076,
      "learning_rate": 1.3403062139598077e-08,
      "loss": 2.1737,
      "step": 114
    },
    {
      "epoch": 3.737704918032787,
      "grad_norm": 0.940012546600253,
      "learning_rate": 9.314053963669244e-09,
      "loss": 2.1013,
      "step": 115
    },
    {
      "epoch": 3.7704918032786887,
      "grad_norm": 0.8831670505877759,
      "learning_rate": 5.964332767796399e-09,
      "loss": 2.0455,
      "step": 116
    },
    {
      "epoch": 3.80327868852459,
      "grad_norm": 0.8719892537418172,
      "learning_rate": 3.356398239470426e-09,
      "loss": 2.0159,
      "step": 117
    },
    {
      "epoch": 3.836065573770492,
      "grad_norm": 0.9053087729946562,
      "learning_rate": 1.4921965171720286e-09,
      "loss": 2.0851,
      "step": 118
    },
    {
      "epoch": 3.8688524590163933,
      "grad_norm": 1.0766975040034732,
      "learning_rate": 3.731187380893175e-10,
      "loss": 2.1016,
      "step": 119
    },
    {
      "epoch": 3.901639344262295,
      "grad_norm": 1.0996423970334042,
      "learning_rate": 0.0,
      "loss": 2.0144,
      "step": 120
    }
  ],
  "logging_steps": 1,
  "max_steps": 120,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 30,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 7.441561537216512e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}