{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.9932659932659933,
  "eval_steps": 500,
  "global_step": 296,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.006734006734006734,
      "grad_norm": 10.587559734000028,
      "learning_rate": 3.3333333333333335e-07,
      "loss": 0.3649,
      "step": 1
    },
    {
      "epoch": 0.013468013468013467,
      "grad_norm": 9.693900007266853,
      "learning_rate": 6.666666666666667e-07,
      "loss": 0.3686,
      "step": 2
    },
    {
      "epoch": 0.020202020202020204,
      "grad_norm": 10.061524524104307,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 0.3574,
      "step": 3
    },
    {
      "epoch": 0.026936026936026935,
      "grad_norm": 8.726573115555482,
      "learning_rate": 1.3333333333333334e-06,
      "loss": 0.3702,
      "step": 4
    },
    {
      "epoch": 0.03367003367003367,
      "grad_norm": 7.3549146274711585,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 0.309,
      "step": 5
    },
    {
      "epoch": 0.04040404040404041,
      "grad_norm": 8.093966841100372,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.3129,
      "step": 6
    },
    {
      "epoch": 0.04713804713804714,
      "grad_norm": 4.842714262161294,
      "learning_rate": 2.3333333333333336e-06,
      "loss": 0.2599,
      "step": 7
    },
    {
      "epoch": 0.05387205387205387,
      "grad_norm": 3.6588226035329976,
      "learning_rate": 2.666666666666667e-06,
      "loss": 0.2195,
      "step": 8
    },
    {
      "epoch": 0.06060606060606061,
      "grad_norm": 4.446495106085817,
      "learning_rate": 3e-06,
      "loss": 0.2421,
      "step": 9
    },
    {
      "epoch": 0.06734006734006734,
      "grad_norm": 2.8148929686162267,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.1974,
      "step": 10
    },
    {
      "epoch": 0.07407407407407407,
      "grad_norm": 3.100744939706759,
      "learning_rate": 3.6666666666666666e-06,
      "loss": 0.226,
      "step": 11
    },
    {
      "epoch": 0.08080808080808081,
      "grad_norm": 5.783874426714631,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.244,
      "step": 12
    },
    {
      "epoch": 0.08754208754208755,
      "grad_norm": 3.931311548892614,
      "learning_rate": 4.333333333333334e-06,
      "loss": 0.2085,
      "step": 13
    },
    {
      "epoch": 0.09427609427609428,
      "grad_norm": 3.160745880374999,
      "learning_rate": 4.666666666666667e-06,
      "loss": 0.193,
      "step": 14
    },
    {
      "epoch": 0.10101010101010101,
      "grad_norm": 2.593859671610231,
      "learning_rate": 5e-06,
      "loss": 0.1918,
      "step": 15
    },
    {
      "epoch": 0.10774410774410774,
      "grad_norm": 2.3411274601291923,
      "learning_rate": 5.333333333333334e-06,
      "loss": 0.1658,
      "step": 16
    },
    {
      "epoch": 0.11447811447811448,
      "grad_norm": 2.4713577315394915,
      "learning_rate": 5.666666666666667e-06,
      "loss": 0.1631,
      "step": 17
    },
    {
      "epoch": 0.12121212121212122,
      "grad_norm": 1.7868674672717264,
      "learning_rate": 6e-06,
      "loss": 0.1447,
      "step": 18
    },
    {
      "epoch": 0.12794612794612795,
      "grad_norm": 2.691562015290166,
      "learning_rate": 6.333333333333333e-06,
      "loss": 0.1642,
      "step": 19
    },
    {
      "epoch": 0.13468013468013468,
      "grad_norm": 2.656849943102207,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.1679,
      "step": 20
    },
    {
      "epoch": 0.1414141414141414,
      "grad_norm": 1.8157646932459017,
      "learning_rate": 7e-06,
      "loss": 0.148,
      "step": 21
    },
    {
      "epoch": 0.14814814814814814,
      "grad_norm": 1.8536112319855758,
      "learning_rate": 7.333333333333333e-06,
      "loss": 0.171,
      "step": 22
    },
    {
      "epoch": 0.15488215488215487,
      "grad_norm": 1.7970217082935407,
      "learning_rate": 7.666666666666667e-06,
      "loss": 0.1543,
      "step": 23
    },
    {
      "epoch": 0.16161616161616163,
      "grad_norm": 1.6194461511446632,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.1331,
      "step": 24
    },
    {
      "epoch": 0.16835016835016836,
      "grad_norm": 1.6671876155213552,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.1997,
      "step": 25
    },
    {
      "epoch": 0.1750841750841751,
      "grad_norm": 2.8083673280494987,
      "learning_rate": 8.666666666666668e-06,
      "loss": 0.149,
      "step": 26
    },
    {
      "epoch": 0.18181818181818182,
      "grad_norm": 2.3945625699544983,
      "learning_rate": 9e-06,
      "loss": 0.2017,
      "step": 27
    },
    {
      "epoch": 0.18855218855218855,
      "grad_norm": 2.547483046750297,
      "learning_rate": 9.333333333333334e-06,
      "loss": 0.1948,
      "step": 28
    },
    {
      "epoch": 0.19528619528619529,
      "grad_norm": 2.1650784865978103,
      "learning_rate": 9.666666666666667e-06,
      "loss": 0.1371,
      "step": 29
    },
    {
      "epoch": 0.20202020202020202,
      "grad_norm": 3.341335706430969,
      "learning_rate": 1e-05,
      "loss": 0.1475,
      "step": 30
    },
    {
      "epoch": 0.20875420875420875,
      "grad_norm": 3.4978479659224306,
      "learning_rate": 9.999651284354774e-06,
      "loss": 0.1742,
      "step": 31
    },
    {
      "epoch": 0.21548821548821548,
      "grad_norm": 2.8495339317165573,
      "learning_rate": 9.998605186060138e-06,
      "loss": 0.1318,
      "step": 32
    },
    {
      "epoch": 0.2222222222222222,
      "grad_norm": 2.4440237950440045,
      "learning_rate": 9.996861851032426e-06,
      "loss": 0.1636,
      "step": 33
    },
    {
      "epoch": 0.22895622895622897,
      "grad_norm": 2.606759351085706,
      "learning_rate": 9.99442152244292e-06,
      "loss": 0.1655,
      "step": 34
    },
    {
      "epoch": 0.2356902356902357,
      "grad_norm": 2.4143459809986276,
      "learning_rate": 9.991284540683922e-06,
      "loss": 0.1492,
      "step": 35
    },
    {
      "epoch": 0.24242424242424243,
      "grad_norm": 2.2585659926066226,
      "learning_rate": 9.98745134332128e-06,
      "loss": 0.1426,
      "step": 36
    },
    {
      "epoch": 0.24915824915824916,
      "grad_norm": 2.2201124689491434,
      "learning_rate": 9.98292246503335e-06,
      "loss": 0.1692,
      "step": 37
    },
    {
      "epoch": 0.2558922558922559,
      "grad_norm": 1.8870018630441274,
      "learning_rate": 9.97769853753642e-06,
      "loss": 0.1139,
      "step": 38
    },
    {
      "epoch": 0.26262626262626265,
      "grad_norm": 1.571794239015075,
      "learning_rate": 9.971780289496585e-06,
      "loss": 0.1301,
      "step": 39
    },
    {
      "epoch": 0.26936026936026936,
      "grad_norm": 1.8448276378849247,
      "learning_rate": 9.965168546428122e-06,
      "loss": 0.1707,
      "step": 40
    },
    {
      "epoch": 0.2760942760942761,
      "grad_norm": 1.739556757210725,
      "learning_rate": 9.95786423057833e-06,
      "loss": 0.1229,
      "step": 41
    },
    {
      "epoch": 0.2828282828282828,
      "grad_norm": 2.3063563017251667,
      "learning_rate": 9.949868360798893e-06,
      "loss": 0.1528,
      "step": 42
    },
    {
      "epoch": 0.2895622895622896,
      "grad_norm": 1.6568054539708137,
      "learning_rate": 9.941182052403768e-06,
      "loss": 0.1402,
      "step": 43
    },
    {
      "epoch": 0.2962962962962963,
      "grad_norm": 2.4925535277590254,
      "learning_rate": 9.931806517013612e-06,
      "loss": 0.1805,
      "step": 44
    },
    {
      "epoch": 0.30303030303030304,
      "grad_norm": 1.4549013854977153,
      "learning_rate": 9.921743062386773e-06,
      "loss": 0.1167,
      "step": 45
    },
    {
      "epoch": 0.30976430976430974,
      "grad_norm": 1.3094244266752977,
      "learning_rate": 9.910993092236878e-06,
      "loss": 0.125,
      "step": 46
    },
    {
      "epoch": 0.3164983164983165,
      "grad_norm": 1.7107110369252345,
      "learning_rate": 9.899558106037039e-06,
      "loss": 0.1509,
      "step": 47
    },
    {
      "epoch": 0.32323232323232326,
      "grad_norm": 1.2447980159357512,
      "learning_rate": 9.887439698810694e-06,
      "loss": 0.1101,
      "step": 48
    },
    {
      "epoch": 0.32996632996632996,
      "grad_norm": 1.4004468828469865,
      "learning_rate": 9.874639560909118e-06,
      "loss": 0.1526,
      "step": 49
    },
    {
      "epoch": 0.3367003367003367,
      "grad_norm": 1.7542103007123673,
      "learning_rate": 9.861159477775653e-06,
      "loss": 0.1446,
      "step": 50
    },
    {
      "epoch": 0.3434343434343434,
      "grad_norm": 1.2222510349896305,
      "learning_rate": 9.847001329696653e-06,
      "loss": 0.118,
      "step": 51
    },
    {
      "epoch": 0.3501683501683502,
      "grad_norm": 1.8427984079271709,
      "learning_rate": 9.832167091539215e-06,
      "loss": 0.142,
      "step": 52
    },
    {
      "epoch": 0.3569023569023569,
      "grad_norm": 2.031779034578671,
      "learning_rate": 9.816658832475709e-06,
      "loss": 0.1314,
      "step": 53
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 1.6115001478932334,
      "learning_rate": 9.800478715695165e-06,
      "loss": 0.1389,
      "step": 54
    },
    {
      "epoch": 0.37037037037037035,
      "grad_norm": 1.3977600344091157,
      "learning_rate": 9.783628998101525e-06,
      "loss": 0.1223,
      "step": 55
    },
    {
      "epoch": 0.3771043771043771,
      "grad_norm": 1.2541929470324433,
      "learning_rate": 9.766112029998847e-06,
      "loss": 0.1188,
      "step": 56
    },
    {
      "epoch": 0.3838383838383838,
      "grad_norm": 1.1839769901258788,
      "learning_rate": 9.747930254763467e-06,
      "loss": 0.1445,
      "step": 57
    },
    {
      "epoch": 0.39057239057239057,
      "grad_norm": 2.1758765383002476,
      "learning_rate": 9.729086208503174e-06,
      "loss": 0.1361,
      "step": 58
    },
    {
      "epoch": 0.39730639730639733,
      "grad_norm": 1.308302111463449,
      "learning_rate": 9.70958251970347e-06,
      "loss": 0.1206,
      "step": 59
    },
    {
      "epoch": 0.40404040404040403,
      "grad_norm": 1.2548920096786198,
      "learning_rate": 9.689421908860928e-06,
      "loss": 0.1547,
      "step": 60
    },
    {
      "epoch": 0.4107744107744108,
      "grad_norm": 1.8106730974081224,
      "learning_rate": 9.668607188103708e-06,
      "loss": 0.1739,
      "step": 61
    },
    {
      "epoch": 0.4175084175084175,
      "grad_norm": 1.7503742895673335,
      "learning_rate": 9.64714126079933e-06,
      "loss": 0.126,
      "step": 62
    },
    {
      "epoch": 0.42424242424242425,
      "grad_norm": 1.9002464919585416,
      "learning_rate": 9.625027121149665e-06,
      "loss": 0.1209,
      "step": 63
    },
    {
      "epoch": 0.43097643097643096,
      "grad_norm": 1.3005333196675646,
      "learning_rate": 9.602267853773301e-06,
      "loss": 0.1198,
      "step": 64
    },
    {
      "epoch": 0.4377104377104377,
      "grad_norm": 1.534465904527491,
      "learning_rate": 9.578866633275289e-06,
      "loss": 0.1118,
      "step": 65
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 1.2985721012757516,
      "learning_rate": 9.554826723804304e-06,
      "loss": 0.108,
      "step": 66
    },
    {
      "epoch": 0.4511784511784512,
      "grad_norm": 1.7026939587233383,
      "learning_rate": 9.530151478597366e-06,
      "loss": 0.1229,
      "step": 67
    },
    {
      "epoch": 0.45791245791245794,
      "grad_norm": 1.1948241076690054,
      "learning_rate": 9.504844339512096e-06,
      "loss": 0.1221,
      "step": 68
    },
    {
      "epoch": 0.46464646464646464,
      "grad_norm": 1.4429708012011027,
      "learning_rate": 9.478908836546629e-06,
      "loss": 0.1154,
      "step": 69
    },
    {
      "epoch": 0.4713804713804714,
      "grad_norm": 1.3983865141656082,
      "learning_rate": 9.452348587347224e-06,
      "loss": 0.1024,
      "step": 70
    },
    {
      "epoch": 0.4781144781144781,
      "grad_norm": 1.316559372373498,
      "learning_rate": 9.425167296703655e-06,
      "loss": 0.1299,
      "step": 71
    },
    {
      "epoch": 0.48484848484848486,
      "grad_norm": 0.8721397993491847,
      "learning_rate": 9.397368756032445e-06,
      "loss": 0.0838,
      "step": 72
    },
    {
      "epoch": 0.49158249158249157,
      "grad_norm": 2.3999305030312006,
      "learning_rate": 9.368956842848014e-06,
      "loss": 0.1407,
      "step": 73
    },
    {
      "epoch": 0.4983164983164983,
      "grad_norm": 1.2702381897041746,
      "learning_rate": 9.339935520221816e-06,
      "loss": 0.1185,
      "step": 74
    },
    {
      "epoch": 0.5050505050505051,
      "grad_norm": 1.2224610902674924,
      "learning_rate": 9.310308836229548e-06,
      "loss": 0.1044,
      "step": 75
    },
    {
      "epoch": 0.5117845117845118,
      "grad_norm": 1.0145648591712295,
      "learning_rate": 9.280080923386501e-06,
      "loss": 0.0793,
      "step": 76
    },
    {
      "epoch": 0.5185185185185185,
      "grad_norm": 1.8590770931022098,
      "learning_rate": 9.249255998071127e-06,
      "loss": 0.1218,
      "step": 77
    },
    {
      "epoch": 0.5252525252525253,
      "grad_norm": 1.1046753339013544,
      "learning_rate": 9.217838359936914e-06,
      "loss": 0.0982,
      "step": 78
    },
    {
      "epoch": 0.531986531986532,
      "grad_norm": 1.5353842647876887,
      "learning_rate": 9.185832391312644e-06,
      "loss": 0.0976,
      "step": 79
    },
    {
      "epoch": 0.5387205387205387,
      "grad_norm": 1.6322780255697593,
      "learning_rate": 9.153242556591115e-06,
      "loss": 0.1073,
      "step": 80
    },
    {
      "epoch": 0.5454545454545454,
      "grad_norm": 1.4318238106918786,
      "learning_rate": 9.120073401606427e-06,
      "loss": 0.1535,
      "step": 81
    },
    {
      "epoch": 0.5521885521885522,
      "grad_norm": 1.2692358298751287,
      "learning_rate": 9.08632955299989e-06,
      "loss": 0.1162,
      "step": 82
    },
    {
      "epoch": 0.5589225589225589,
      "grad_norm": 0.9386668388367472,
      "learning_rate": 9.052015717574683e-06,
      "loss": 0.0819,
      "step": 83
    },
    {
      "epoch": 0.5656565656565656,
      "grad_norm": 1.578041471845064,
      "learning_rate": 9.017136681639307e-06,
      "loss": 0.1438,
      "step": 84
    },
    {
      "epoch": 0.5723905723905723,
      "grad_norm": 1.3747097561675499,
      "learning_rate": 8.981697310339972e-06,
      "loss": 0.1129,
      "step": 85
    },
    {
      "epoch": 0.5791245791245792,
      "grad_norm": 1.2857978501576468,
      "learning_rate": 8.94570254698197e-06,
      "loss": 0.1153,
      "step": 86
    },
    {
      "epoch": 0.5858585858585859,
      "grad_norm": 1.1071476461695378,
      "learning_rate": 8.90915741234015e-06,
      "loss": 0.1018,
      "step": 87
    },
    {
      "epoch": 0.5925925925925926,
      "grad_norm": 1.1979772322060398,
      "learning_rate": 8.872067003958597e-06,
      "loss": 0.0939,
      "step": 88
    },
    {
      "epoch": 0.5993265993265994,
      "grad_norm": 1.4043079567849108,
      "learning_rate": 8.834436495439588e-06,
      "loss": 0.1022,
      "step": 89
    },
    {
      "epoch": 0.6060606060606061,
      "grad_norm": 1.182762910272131,
      "learning_rate": 8.796271135721944e-06,
      "loss": 0.1099,
      "step": 90
    },
    {
      "epoch": 0.6127946127946128,
      "grad_norm": 0.9919737781684954,
      "learning_rate": 8.757576248348883e-06,
      "loss": 0.101,
      "step": 91
    },
    {
      "epoch": 0.6195286195286195,
      "grad_norm": 1.0613946248613535,
      "learning_rate": 8.71835723072545e-06,
      "loss": 0.1074,
      "step": 92
    },
    {
      "epoch": 0.6262626262626263,
      "grad_norm": 1.1119387505650915,
      "learning_rate": 8.67861955336566e-06,
      "loss": 0.0999,
      "step": 93
    },
    {
      "epoch": 0.632996632996633,
      "grad_norm": 1.0663092046509486,
      "learning_rate": 8.638368759129433e-06,
      "loss": 0.1044,
      "step": 94
    },
    {
      "epoch": 0.6397306397306397,
      "grad_norm": 1.5025873764978885,
      "learning_rate": 8.597610462449441e-06,
      "loss": 0.092,
      "step": 95
    },
    {
      "epoch": 0.6464646464646465,
      "grad_norm": 1.1482831508449691,
      "learning_rate": 8.556350348547978e-06,
      "loss": 0.099,
      "step": 96
    },
    {
      "epoch": 0.6531986531986532,
      "grad_norm": 1.1531128286621257,
      "learning_rate": 8.514594172643934e-06,
      "loss": 0.1122,
      "step": 97
    },
    {
      "epoch": 0.6599326599326599,
      "grad_norm": 0.92890945229994,
      "learning_rate": 8.472347759150044e-06,
      "loss": 0.0934,
      "step": 98
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.8710829747699658,
      "learning_rate": 8.429617000860441e-06,
      "loss": 0.0963,
      "step": 99
    },
    {
      "epoch": 0.6734006734006734,
      "grad_norm": 1.5913618273955719,
      "learning_rate": 8.386407858128707e-06,
      "loss": 0.1017,
      "step": 100
    },
    {
      "epoch": 0.6801346801346801,
      "grad_norm": 2.0477106016376423,
      "learning_rate": 8.342726358036473e-06,
      "loss": 0.1643,
      "step": 101
    },
    {
      "epoch": 0.6868686868686869,
      "grad_norm": 1.310766070822457,
      "learning_rate": 8.298578593552737e-06,
      "loss": 0.1129,
      "step": 102
    },
    {
      "epoch": 0.6936026936026936,
      "grad_norm": 1.0689180927391349,
      "learning_rate": 8.253970722683968e-06,
      "loss": 0.0805,
      "step": 103
    },
    {
      "epoch": 0.7003367003367004,
      "grad_norm": 1.021500128946588,
      "learning_rate": 8.208908967615159e-06,
      "loss": 0.0854,
      "step": 104
    },
    {
      "epoch": 0.7070707070707071,
      "grad_norm": 1.6146318951430105,
      "learning_rate": 8.163399613841903e-06,
      "loss": 0.1026,
      "step": 105
    },
    {
      "epoch": 0.7138047138047138,
      "grad_norm": 1.0430932372821804,
      "learning_rate": 8.117449009293668e-06,
      "loss": 0.0911,
      "step": 106
    },
    {
      "epoch": 0.7205387205387206,
      "grad_norm": 0.9627471656747238,
      "learning_rate": 8.071063563448341e-06,
      "loss": 0.0864,
      "step": 107
    },
    {
      "epoch": 0.7272727272727273,
      "grad_norm": 0.978325652238678,
      "learning_rate": 8.024249746438189e-06,
      "loss": 0.0767,
      "step": 108
    },
    {
      "epoch": 0.734006734006734,
      "grad_norm": 1.5716697095342131,
      "learning_rate": 7.977014088147375e-06,
      "loss": 0.117,
      "step": 109
    },
    {
      "epoch": 0.7407407407407407,
      "grad_norm": 0.9226258206362579,
      "learning_rate": 7.929363177301124e-06,
      "loss": 0.0898,
      "step": 110
    },
    {
      "epoch": 0.7474747474747475,
      "grad_norm": 0.979018465840073,
      "learning_rate": 7.881303660546684e-06,
      "loss": 0.096,
      "step": 111
    },
    {
      "epoch": 0.7542087542087542,
      "grad_norm": 1.405839836354203,
      "learning_rate": 7.832842241526212e-06,
      "loss": 0.0969,
      "step": 112
    },
    {
      "epoch": 0.7609427609427609,
      "grad_norm": 1.7228913938880963,
      "learning_rate": 7.78398567994171e-06,
      "loss": 0.1229,
      "step": 113
    },
    {
      "epoch": 0.7676767676767676,
      "grad_norm": 1.7211262216477865,
      "learning_rate": 7.734740790612137e-06,
      "loss": 0.1156,
      "step": 114
    },
    {
      "epoch": 0.7744107744107744,
      "grad_norm": 1.1471367685176275,
      "learning_rate": 7.685114442522831e-06,
      "loss": 0.0953,
      "step": 115
    },
    {
      "epoch": 0.7811447811447811,
      "grad_norm": 1.0484067219863682,
      "learning_rate": 7.635113557867395e-06,
      "loss": 0.0957,
      "step": 116
    },
    {
      "epoch": 0.7878787878787878,
      "grad_norm": 1.3643749264979563,
      "learning_rate": 7.584745111082128e-06,
      "loss": 0.0985,
      "step": 117
    },
    {
      "epoch": 0.7946127946127947,
      "grad_norm": 1.1409003431687448,
      "learning_rate": 7.5340161278732e-06,
      "loss": 0.0948,
      "step": 118
    },
    {
      "epoch": 0.8013468013468014,
      "grad_norm": 1.1845755031970973,
      "learning_rate": 7.482933684236654e-06,
      "loss": 0.1278,
      "step": 119
    },
    {
      "epoch": 0.8080808080808081,
      "grad_norm": 1.2048614884456728,
      "learning_rate": 7.431504905471407e-06,
      "loss": 0.0866,
      "step": 120
    },
    {
      "epoch": 0.8148148148148148,
      "grad_norm": 1.1024340198116762,
      "learning_rate": 7.379736965185369e-06,
      "loss": 0.0886,
      "step": 121
    },
    {
      "epoch": 0.8215488215488216,
      "grad_norm": 1.1875877764472678,
      "learning_rate": 7.327637084294818e-06,
      "loss": 0.0876,
      "step": 122
    },
    {
      "epoch": 0.8282828282828283,
      "grad_norm": 0.8892335379224572,
      "learning_rate": 7.2752125300171835e-06,
      "loss": 0.0829,
      "step": 123
    },
    {
      "epoch": 0.835016835016835,
      "grad_norm": 1.1503853071251364,
      "learning_rate": 7.22247061485738e-06,
      "loss": 0.0945,
      "step": 124
    },
    {
      "epoch": 0.8417508417508418,
      "grad_norm": 0.9116461855370438,
      "learning_rate": 7.169418695587791e-06,
      "loss": 0.0642,
      "step": 125
    },
    {
      "epoch": 0.8484848484848485,
      "grad_norm": 0.9169291992757941,
      "learning_rate": 7.1160641722221255e-06,
      "loss": 0.0762,
      "step": 126
    },
    {
      "epoch": 0.8552188552188552,
      "grad_norm": 0.8458786678974077,
      "learning_rate": 7.062414486983197e-06,
      "loss": 0.0782,
      "step": 127
    },
    {
      "epoch": 0.8619528619528619,
      "grad_norm": 2.0440916025334688,
      "learning_rate": 7.008477123264849e-06,
      "loss": 0.1654,
      "step": 128
    },
    {
      "epoch": 0.8686868686868687,
      "grad_norm": 1.2692910029856195,
      "learning_rate": 6.954259604588114e-06,
      "loss": 0.0993,
      "step": 129
    },
    {
      "epoch": 0.8754208754208754,
      "grad_norm": 1.0371836145890874,
      "learning_rate": 6.8997694935518e-06,
      "loss": 0.0792,
      "step": 130
    },
    {
      "epoch": 0.8821548821548821,
      "grad_norm": 0.8129848414149057,
      "learning_rate": 6.845014390777595e-06,
      "loss": 0.0631,
      "step": 131
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 1.0818279823466028,
      "learning_rate": 6.7900019338499005e-06,
      "loss": 0.084,
      "step": 132
    },
    {
      "epoch": 0.8956228956228957,
      "grad_norm": 1.2323415448858157,
      "learning_rate": 6.734739796250477e-06,
      "loss": 0.1037,
      "step": 133
    },
    {
      "epoch": 0.9023569023569024,
      "grad_norm": 0.9072503259074285,
      "learning_rate": 6.6792356862881144e-06,
      "loss": 0.1022,
      "step": 134
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 1.0444248401571283,
      "learning_rate": 6.6234973460234184e-06,
      "loss": 0.0952,
      "step": 135
    },
    {
      "epoch": 0.9158249158249159,
      "grad_norm": 0.7905929983463482,
      "learning_rate": 6.567532550188908e-06,
      "loss": 0.0717,
      "step": 136
    },
    {
      "epoch": 0.9225589225589226,
      "grad_norm": 0.7434134587240314,
      "learning_rate": 6.511349105104534e-06,
      "loss": 0.0944,
      "step": 137
    },
    {
      "epoch": 0.9292929292929293,
      "grad_norm": 0.768332895445934,
      "learning_rate": 6.454954847588824e-06,
      "loss": 0.0678,
      "step": 138
    },
    {
      "epoch": 0.936026936026936,
      "grad_norm": 1.0841144826502787,
      "learning_rate": 6.398357643865731e-06,
      "loss": 0.0845,
      "step": 139
    },
    {
      "epoch": 0.9427609427609428,
      "grad_norm": 1.0657965630616948,
      "learning_rate": 6.341565388467425e-06,
      "loss": 0.0845,
      "step": 140
    },
    {
      "epoch": 0.9494949494949495,
      "grad_norm": 0.7896126816989001,
      "learning_rate": 6.284586003133096e-06,
      "loss": 0.0922,
      "step": 141
    },
    {
      "epoch": 0.9562289562289562,
      "grad_norm": 1.4442123017819188,
      "learning_rate": 6.227427435703997e-06,
      "loss": 0.1182,
      "step": 142
    },
    {
      "epoch": 0.9629629629629629,
      "grad_norm": 0.9662733291916084,
      "learning_rate": 6.170097659014812e-06,
      "loss": 0.0915,
      "step": 143
    },
    {
      "epoch": 0.9696969696969697,
      "grad_norm": 0.9920505746388618,
      "learning_rate": 6.112604669781572e-06,
      "loss": 0.093,
      "step": 144
    },
    {
      "epoch": 0.9764309764309764,
      "grad_norm": 1.09207653298068,
      "learning_rate": 6.054956487486212e-06,
      "loss": 0.0919,
      "step": 145
    },
    {
      "epoch": 0.9831649831649831,
      "grad_norm": 0.7024022249718929,
      "learning_rate": 5.997161153257963e-06,
      "loss": 0.0786,
      "step": 146
    },
    {
      "epoch": 0.98989898989899,
      "grad_norm": 0.8553768429510301,
      "learning_rate": 5.939226728751733e-06,
      "loss": 0.091,
      "step": 147
    },
    {
      "epoch": 0.9966329966329966,
      "grad_norm": 0.7368165994452094,
      "learning_rate": 5.88116129502361e-06,
      "loss": 0.0841,
      "step": 148
    },
    {
      "epoch": 0.9966329966329966,
      "eval_loss": 0.0851789191365242,
      "eval_runtime": 74.3023,
      "eval_samples_per_second": 26.89,
      "eval_steps_per_second": 0.848,
      "step": 148
    },
    {
      "epoch": 1.0033670033670035,
      "grad_norm": 1.3624837722625542,
      "learning_rate": 5.82297295140367e-06,
      "loss": 0.0942,
      "step": 149
    },
    {
      "epoch": 1.0101010101010102,
      "grad_norm": 0.8091703688887699,
      "learning_rate": 5.764669814366231e-06,
      "loss": 0.0673,
      "step": 150
    },
    {
      "epoch": 1.0168350168350169,
      "grad_norm": 1.1689964332508505,
      "learning_rate": 5.70626001639771e-06,
      "loss": 0.0904,
      "step": 151
    },
    {
      "epoch": 1.0235690235690236,
      "grad_norm": 0.9720282544537069,
      "learning_rate": 5.647751704862263e-06,
      "loss": 0.0764,
      "step": 152
    },
    {
      "epoch": 1.0303030303030303,
      "grad_norm": 0.7814529677090204,
      "learning_rate": 5.589153040865333e-06,
      "loss": 0.0766,
      "step": 153
    },
    {
      "epoch": 1.037037037037037,
      "grad_norm": 0.6208562049624894,
      "learning_rate": 5.530472198115291e-06,
      "loss": 0.0546,
      "step": 154
    },
    {
      "epoch": 1.0437710437710437,
      "grad_norm": 1.0609611170144315,
      "learning_rate": 5.471717361783312e-06,
      "loss": 0.0907,
      "step": 155
    },
    {
      "epoch": 1.0505050505050506,
      "grad_norm": 0.9895862720799716,
      "learning_rate": 5.412896727361663e-06,
      "loss": 0.0756,
      "step": 156
    },
    {
      "epoch": 1.0572390572390573,
      "grad_norm": 0.8283511990419397,
      "learning_rate": 5.354018499520536e-06,
      "loss": 0.0733,
      "step": 157
    },
    {
      "epoch": 1.063973063973064,
      "grad_norm": 0.7064918637395274,
      "learning_rate": 5.2950908909636144e-06,
      "loss": 0.0669,
      "step": 158
    },
    {
      "epoch": 1.0707070707070707,
      "grad_norm": 0.7673084442281245,
      "learning_rate": 5.2361221212825175e-06,
      "loss": 0.063,
      "step": 159
    },
    {
      "epoch": 1.0774410774410774,
      "grad_norm": 0.7418675838307957,
      "learning_rate": 5.177120415810271e-06,
      "loss": 0.0556,
      "step": 160
    },
    {
      "epoch": 1.0841750841750841,
      "grad_norm": 0.9302869463074369,
      "learning_rate": 5.11809400447399e-06,
      "loss": 0.0937,
      "step": 161
    },
    {
      "epoch": 1.0909090909090908,
      "grad_norm": 1.2470452298913122,
      "learning_rate": 5.059051120646924e-06,
      "loss": 0.0827,
      "step": 162
    },
    {
      "epoch": 1.0976430976430978,
      "grad_norm": 0.5572520947809881,
      "learning_rate": 5e-06,
      "loss": 0.0537,
      "step": 163
    },
    {
      "epoch": 1.1043771043771045,
      "grad_norm": 0.9002485043341543,
      "learning_rate": 4.940948879353078e-06,
      "loss": 0.088,
      "step": 164
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 0.7257311476037439,
      "learning_rate": 4.8819059955260105e-06,
      "loss": 0.0531,
      "step": 165
    },
    {
      "epoch": 1.1178451178451179,
      "grad_norm": 0.7312827459622674,
      "learning_rate": 4.822879584189732e-06,
      "loss": 0.0734,
      "step": 166
    },
    {
      "epoch": 1.1245791245791246,
      "grad_norm": 0.6288896372866642,
      "learning_rate": 4.763877878717484e-06,
      "loss": 0.0538,
      "step": 167
    },
    {
      "epoch": 1.1313131313131313,
      "grad_norm": 1.002048278793203,
      "learning_rate": 4.704909109036387e-06,
      "loss": 0.0843,
      "step": 168
    },
    {
      "epoch": 1.138047138047138,
      "grad_norm": 0.7883344068499383,
      "learning_rate": 4.645981500479466e-06,
      "loss": 0.0656,
      "step": 169
    },
    {
      "epoch": 1.144781144781145,
      "grad_norm": 0.8382010718049546,
      "learning_rate": 4.587103272638339e-06,
      "loss": 0.0811,
      "step": 170
    },
    {
      "epoch": 1.1515151515151516,
      "grad_norm": 0.7760482065843696,
      "learning_rate": 4.528282638216689e-06,
      "loss": 0.0504,
      "step": 171
    },
    {
      "epoch": 1.1582491582491583,
      "grad_norm": 0.7830043876217352,
      "learning_rate": 4.46952780188471e-06,
      "loss": 0.0715,
      "step": 172
    },
    {
      "epoch": 1.164983164983165,
      "grad_norm": 1.0630254048827579,
      "learning_rate": 4.410846959134667e-06,
      "loss": 0.0947,
      "step": 173
    },
    {
      "epoch": 1.1717171717171717,
      "grad_norm": 1.0925071520075709,
      "learning_rate": 4.352248295137739e-06,
      "loss": 0.0879,
      "step": 174
    },
    {
      "epoch": 1.1784511784511784,
      "grad_norm": 0.6169774600371488,
      "learning_rate": 4.293739983602292e-06,
      "loss": 0.0611,
      "step": 175
    },
    {
      "epoch": 1.1851851851851851,
      "grad_norm": 0.9939160789185896,
      "learning_rate": 4.23533018563377e-06,
      "loss": 0.0883,
      "step": 176
    },
    {
      "epoch": 1.1919191919191918,
      "grad_norm": 1.1279865240839675,
      "learning_rate": 4.17702704859633e-06,
      "loss": 0.0913,
      "step": 177
    },
    {
      "epoch": 1.1986531986531987,
      "grad_norm": 0.9240576378263561,
      "learning_rate": 4.118838704976392e-06,
      "loss": 0.0766,
      "step": 178
    },
    {
      "epoch": 1.2053872053872055,
      "grad_norm": 1.2001616277657225,
      "learning_rate": 4.06077327124827e-06,
      "loss": 0.1161,
      "step": 179
    },
    {
      "epoch": 1.2121212121212122,
      "grad_norm": 0.8147954747129799,
      "learning_rate": 4.002838846742039e-06,
      "loss": 0.0735,
      "step": 180
    },
    {
      "epoch": 1.2188552188552189,
      "grad_norm": 0.7054084380327903,
      "learning_rate": 3.94504351251379e-06,
      "loss": 0.0594,
      "step": 181
    },
    {
      "epoch": 1.2255892255892256,
      "grad_norm": 1.1137616381347046,
      "learning_rate": 3.887395330218429e-06,
      "loss": 0.1081,
      "step": 182
    },
    {
      "epoch": 1.2323232323232323,
      "grad_norm": 0.671871366714128,
      "learning_rate": 3.829902340985189e-06,
      "loss": 0.0697,
      "step": 183
    },
    {
      "epoch": 1.239057239057239,
      "grad_norm": 0.7473723466728038,
      "learning_rate": 3.7725725642960047e-06,
      "loss": 0.0621,
      "step": 184
    },
    {
      "epoch": 1.2457912457912457,
      "grad_norm": 0.592505978208727,
      "learning_rate": 3.7154139968669043e-06,
      "loss": 0.058,
      "step": 185
    },
    {
      "epoch": 1.2525252525252526,
      "grad_norm": 0.8357383325436504,
      "learning_rate": 3.658434611532578e-06,
      "loss": 0.0637,
      "step": 186
    },
    {
      "epoch": 1.2592592592592593,
      "grad_norm": 0.7866708646652821,
      "learning_rate": 3.6016423561342707e-06,
      "loss": 0.0627,
      "step": 187
    },
    {
      "epoch": 1.265993265993266,
      "grad_norm": 0.714908923134987,
      "learning_rate": 3.545045152411178e-06,
      "loss": 0.0632,
      "step": 188
    },
    {
      "epoch": 1.2727272727272727,
      "grad_norm": 0.5485873177126037,
      "learning_rate": 3.4886508948954656e-06,
      "loss": 0.0599,
      "step": 189
    },
    {
      "epoch": 1.2794612794612794,
      "grad_norm": 1.09587990802126,
      "learning_rate": 3.4324674498110956e-06,
      "loss": 0.0613,
      "step": 190
    },
    {
      "epoch": 1.2861952861952861,
      "grad_norm": 0.8460192690372663,
      "learning_rate": 3.3765026539765832e-06,
      "loss": 0.0847,
      "step": 191
    },
    {
      "epoch": 1.2929292929292928,
      "grad_norm": 0.7434506857298966,
      "learning_rate": 3.3207643137118872e-06,
      "loss": 0.0592,
      "step": 192
    },
    {
      "epoch": 1.2996632996632997,
      "grad_norm": 0.7460450427195592,
      "learning_rate": 3.2652602037495247e-06,
      "loss": 0.0579,
      "step": 193
    },
    {
      "epoch": 1.3063973063973064,
      "grad_norm": 1.1738509828924104,
      "learning_rate": 3.2099980661501016e-06,
      "loss": 0.0788,
      "step": 194
    },
    {
      "epoch": 1.3131313131313131,
      "grad_norm": 0.8215622674250468,
      "learning_rate": 3.154985609222405e-06,
      "loss": 0.0656,
      "step": 195
    },
    {
      "epoch": 1.3198653198653199,
      "grad_norm": 0.7211793878206931,
      "learning_rate": 3.1002305064482006e-06,
      "loss": 0.0643,
      "step": 196
    },
    {
      "epoch": 1.3265993265993266,
      "grad_norm": 0.975561078244494,
      "learning_rate": 3.045740395411886e-06,
      "loss": 0.0671,
      "step": 197
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.5847136475401894,
      "learning_rate": 2.991522876735154e-06,
      "loss": 0.0483,
      "step": 198
    },
    {
      "epoch": 1.34006734006734,
      "grad_norm": 0.7843079810152986,
      "learning_rate": 2.9375855130168046e-06,
      "loss": 0.0689,
      "step": 199
    },
    {
      "epoch": 1.3468013468013469,
      "grad_norm": 0.8005399461255156,
      "learning_rate": 2.8839358277778758e-06,
      "loss": 0.0664,
      "step": 200
    },
    {
      "epoch": 1.3535353535353536,
      "grad_norm": 0.6772138624158527,
      "learning_rate": 2.83058130441221e-06,
      "loss": 0.0615,
      "step": 201
    },
    {
      "epoch": 1.3602693602693603,
      "grad_norm": 0.7191213036719818,
      "learning_rate": 2.7775293851426233e-06,
      "loss": 0.0774,
      "step": 202
    },
    {
      "epoch": 1.367003367003367,
      "grad_norm": 0.7605152392156752,
      "learning_rate": 2.7247874699828186e-06,
      "loss": 0.0928,
      "step": 203
    },
    {
      "epoch": 1.3737373737373737,
      "grad_norm": 0.6195090121798972,
      "learning_rate": 2.6723629157051844e-06,
      "loss": 0.0596,
      "step": 204
    },
    {
      "epoch": 1.3804713804713804,
      "grad_norm": 0.6643501098994239,
      "learning_rate": 2.6202630348146323e-06,
      "loss": 0.0523,
      "step": 205
    },
    {
      "epoch": 1.387205387205387,
      "grad_norm": 0.8554198672675404,
      "learning_rate": 2.5684950945285937e-06,
      "loss": 0.0903,
      "step": 206
    },
    {
      "epoch": 1.393939393939394,
      "grad_norm": 0.8364523777121037,
      "learning_rate": 2.517066315763348e-06,
      "loss": 0.0758,
      "step": 207
    },
    {
      "epoch": 1.4006734006734007,
      "grad_norm": 0.7135150317434039,
      "learning_rate": 2.4659838721268005e-06,
      "loss": 0.0603,
      "step": 208
    },
    {
      "epoch": 1.4074074074074074,
      "grad_norm": 0.6321370814051162,
      "learning_rate": 2.4152548889178722e-06,
      "loss": 0.093,
      "step": 209
    },
    {
      "epoch": 1.4141414141414141,
      "grad_norm": 0.600961736861407,
      "learning_rate": 2.364886442132606e-06,
      "loss": 0.0443,
      "step": 210
    },
    {
      "epoch": 1.4208754208754208,
      "grad_norm": 0.9011898089044157,
      "learning_rate": 2.3148855574771706e-06,
      "loss": 0.0713,
      "step": 211
    },
    {
      "epoch": 1.4276094276094276,
      "grad_norm": 0.572568082620369,
      "learning_rate": 2.265259209387867e-06,
      "loss": 0.0567,
      "step": 212
    },
    {
      "epoch": 1.4343434343434343,
      "grad_norm": 0.5810025449791688,
      "learning_rate": 2.2160143200582906e-06,
      "loss": 0.0561,
      "step": 213
    },
    {
      "epoch": 1.4410774410774412,
      "grad_norm": 0.5830304960007473,
      "learning_rate": 2.16715775847379e-06,
      "loss": 0.048,
      "step": 214
    },
    {
      "epoch": 1.4478114478114479,
      "grad_norm": 0.6676748383092154,
      "learning_rate": 2.1186963394533165e-06,
      "loss": 0.0593,
      "step": 215
    },
    {
      "epoch": 1.4545454545454546,
      "grad_norm": 0.6977658852399721,
      "learning_rate": 2.0706368226988772e-06,
      "loss": 0.0594,
      "step": 216
    },
    {
      "epoch": 1.4612794612794613,
      "grad_norm": 0.506749158721991,
      "learning_rate": 2.0229859118526244e-06,
      "loss": 0.0556,
      "step": 217
    },
    {
      "epoch": 1.468013468013468,
      "grad_norm": 0.7142907436715384,
      "learning_rate": 1.9757502535618137e-06,
      "loss": 0.0524,
      "step": 218
    },
    {
      "epoch": 1.4747474747474747,
      "grad_norm": 0.6949958378386697,
      "learning_rate": 1.928936436551661e-06,
      "loss": 0.0666,
      "step": 219
    },
    {
      "epoch": 1.4814814814814814,
      "grad_norm": 0.7555665471689874,
      "learning_rate": 1.8825509907063328e-06,
      "loss": 0.0716,
      "step": 220
    },
    {
      "epoch": 1.4882154882154883,
      "grad_norm": 0.8498899722890332,
      "learning_rate": 1.8366003861580966e-06,
      "loss": 0.0521,
      "step": 221
    },
    {
      "epoch": 1.494949494949495,
      "grad_norm": 0.7857371564300432,
      "learning_rate": 1.7910910323848435e-06,
      "loss": 0.0523,
      "step": 222
    },
    {
      "epoch": 1.5016835016835017,
      "grad_norm": 0.6313660611784068,
      "learning_rate": 1.7460292773160315e-06,
      "loss": 0.0485,
      "step": 223
    },
    {
      "epoch": 1.5084175084175084,
      "grad_norm": 0.790847117411599,
      "learning_rate": 1.7014214064472646e-06,
      "loss": 0.0808,
      "step": 224
    },
    {
      "epoch": 1.5151515151515151,
      "grad_norm": 1.1225175609085682,
      "learning_rate": 1.6572736419635288e-06,
      "loss": 0.0875,
      "step": 225
    },
    {
      "epoch": 1.5218855218855218,
      "grad_norm": 0.8723669974198242,
      "learning_rate": 1.6135921418712959e-06,
      "loss": 0.0614,
      "step": 226
    },
    {
      "epoch": 1.5286195286195285,
      "grad_norm": 0.5803673000386498,
      "learning_rate": 1.5703829991395602e-06,
      "loss": 0.0521,
      "step": 227
    },
    {
      "epoch": 1.5353535353535355,
      "grad_norm": 0.5913655803974819,
      "learning_rate": 1.5276522408499567e-06,
      "loss": 0.0567,
      "step": 228
    },
    {
      "epoch": 1.542087542087542,
      "grad_norm": 0.5913682879212468,
      "learning_rate": 1.4854058273560667e-06,
      "loss": 0.0575,
      "step": 229
    },
    {
      "epoch": 1.5488215488215489,
      "grad_norm": 0.6925238047248119,
      "learning_rate": 1.4436496514520253e-06,
      "loss": 0.0582,
      "step": 230
    },
    {
      "epoch": 1.5555555555555556,
      "grad_norm": 0.6395725970290016,
      "learning_rate": 1.4023895375505608e-06,
      "loss": 0.0605,
      "step": 231
    },
    {
      "epoch": 1.5622895622895623,
      "grad_norm": 0.5799228625915099,
      "learning_rate": 1.361631240870569e-06,
      "loss": 0.0572,
      "step": 232
    },
    {
      "epoch": 1.569023569023569,
      "grad_norm": 1.1307976257157992,
      "learning_rate": 1.321380446634342e-06,
      "loss": 0.0683,
      "step": 233
    },
    {
      "epoch": 1.5757575757575757,
      "grad_norm": 0.53126448503362,
      "learning_rate": 1.281642769274552e-06,
      "loss": 0.044,
      "step": 234
    },
    {
      "epoch": 1.5824915824915826,
      "grad_norm": 0.5123950135517319,
      "learning_rate": 1.242423751651119e-06,
      "loss": 0.0804,
      "step": 235
    },
    {
      "epoch": 1.589225589225589,
      "grad_norm": 0.6099923775928446,
      "learning_rate": 1.2037288642780575e-06,
      "loss": 0.0408,
      "step": 236
    },
    {
      "epoch": 1.595959595959596,
      "grad_norm": 0.6248835012817582,
      "learning_rate": 1.165563504560413e-06,
      "loss": 0.0544,
      "step": 237
    },
    {
      "epoch": 1.6026936026936027,
      "grad_norm": 0.6510387000168023,
      "learning_rate": 1.1279329960414047e-06,
      "loss": 0.0815,
      "step": 238
    },
    {
      "epoch": 1.6094276094276094,
      "grad_norm": 0.6393553382001962,
      "learning_rate": 1.0908425876598512e-06,
      "loss": 0.0538,
      "step": 239
    },
    {
      "epoch": 1.6161616161616161,
      "grad_norm": 0.5759287160512089,
      "learning_rate": 1.0542974530180327e-06,
      "loss": 0.0574,
      "step": 240
    },
    {
      "epoch": 1.6228956228956228,
      "grad_norm": 0.7684356456332082,
      "learning_rate": 1.0183026896600284e-06,
      "loss": 0.0704,
      "step": 241
    },
    {
      "epoch": 1.6296296296296298,
      "grad_norm": 0.7092834204839467,
      "learning_rate": 9.82863318360695e-07,
      "loss": 0.0688,
      "step": 242
    },
    {
      "epoch": 1.6363636363636362,
      "grad_norm": 0.8525774187670475,
      "learning_rate": 9.479842824253182e-07,
      "loss": 0.058,
      "step": 243
    },
    {
      "epoch": 1.6430976430976432,
      "grad_norm": 0.494283557256197,
      "learning_rate": 9.136704470001101e-07,
      "loss": 0.0455,
      "step": 244
    },
    {
      "epoch": 1.6498316498316499,
      "grad_norm": 0.5072796987904575,
      "learning_rate": 8.799265983935734e-07,
      "loss": 0.0573,
      "step": 245
    },
    {
      "epoch": 1.6565656565656566,
      "grad_norm": 0.6544970574576625,
      "learning_rate": 8.46757443408886e-07,
      "loss": 0.0606,
      "step": 246
    },
    {
      "epoch": 1.6632996632996633,
      "grad_norm": 0.5325045241743026,
      "learning_rate": 8.141676086873574e-07,
      "loss": 0.0476,
      "step": 247
    },
    {
      "epoch": 1.67003367003367,
      "grad_norm": 0.571954858670152,
      "learning_rate": 7.821616400630866e-07,
      "loss": 0.0553,
      "step": 248
    },
    {
      "epoch": 1.676767676767677,
      "grad_norm": 0.7124362240685431,
      "learning_rate": 7.507440019288742e-07,
      "loss": 0.0648,
      "step": 249
    },
    {
      "epoch": 1.6835016835016834,
      "grad_norm": 0.8074026482316966,
      "learning_rate": 7.199190766135001e-07,
      "loss": 0.0677,
      "step": 250
    },
    {
      "epoch": 1.6902356902356903,
      "grad_norm": 0.49821428047955796,
      "learning_rate": 6.896911637704534e-07,
      "loss": 0.0414,
      "step": 251
    },
    {
      "epoch": 1.696969696969697,
      "grad_norm": 0.6996554021349439,
      "learning_rate": 6.600644797781847e-07,
      "loss": 0.0666,
      "step": 252
    },
    {
      "epoch": 1.7037037037037037,
      "grad_norm": 0.5640914695952624,
      "learning_rate": 6.310431571519865e-07,
      "loss": 0.076,
      "step": 253
    },
    {
      "epoch": 1.7104377104377104,
      "grad_norm": 0.5394553999404826,
      "learning_rate": 6.026312439675553e-07,
      "loss": 0.0498,
      "step": 254
    },
    {
      "epoch": 1.7171717171717171,
      "grad_norm": 0.47737348342117025,
      "learning_rate": 5.748327032963464e-07,
      "loss": 0.0446,
      "step": 255
    },
    {
      "epoch": 1.723905723905724,
      "grad_norm": 0.5870399819271779,
      "learning_rate": 5.476514126527771e-07,
      "loss": 0.085,
      "step": 256
    },
    {
      "epoch": 1.7306397306397305,
      "grad_norm": 0.6059552958362416,
      "learning_rate": 5.210911634533722e-07,
      "loss": 0.0628,
      "step": 257
    },
    {
      "epoch": 1.7373737373737375,
      "grad_norm": 0.4612493978879547,
      "learning_rate": 4.951556604879049e-07,
      "loss": 0.045,
      "step": 258
    },
    {
      "epoch": 1.7441077441077442,
      "grad_norm": 0.9732831834165256,
      "learning_rate": 4.698485214026349e-07,
      "loss": 0.0719,
      "step": 259
    },
    {
      "epoch": 1.7508417508417509,
      "grad_norm": 0.7084062235433236,
      "learning_rate": 4.4517327619569784e-07,
      "loss": 0.0451,
      "step": 260
    },
    {
      "epoch": 1.7575757575757576,
      "grad_norm": 0.5150893466864267,
      "learning_rate": 4.211333667247125e-07,
      "loss": 0.054,
      "step": 261
    },
    {
      "epoch": 1.7643097643097643,
      "grad_norm": 0.7197694152938333,
      "learning_rate": 3.9773214622669974e-07,
      "loss": 0.0528,
      "step": 262
    },
    {
      "epoch": 1.7710437710437712,
      "grad_norm": 0.7937673202702573,
      "learning_rate": 3.7497287885033763e-07,
      "loss": 0.0551,
      "step": 263
    },
    {
      "epoch": 1.7777777777777777,
      "grad_norm": 0.7807725839915187,
      "learning_rate": 3.528587392006716e-07,
      "loss": 0.0503,
      "step": 264
    },
    {
      "epoch": 1.7845117845117846,
      "grad_norm": 0.5505115937056173,
      "learning_rate": 3.313928118962906e-07,
      "loss": 0.0623,
      "step": 265
    },
    {
      "epoch": 1.791245791245791,
      "grad_norm": 0.6073147135377622,
      "learning_rate": 3.105780911390738e-07,
      "loss": 0.0674,
      "step": 266
    },
    {
      "epoch": 1.797979797979798,
      "grad_norm": 0.4386889000592362,
      "learning_rate": 2.904174802965293e-07,
      "loss": 0.0337,
      "step": 267
    },
    {
      "epoch": 1.8047138047138047,
      "grad_norm": 0.5371720109059138,
      "learning_rate": 2.7091379149682683e-07,
      "loss": 0.056,
      "step": 268
    },
    {
      "epoch": 1.8114478114478114,
      "grad_norm": 0.4897357275916543,
      "learning_rate": 2.520697452365345e-07,
      "loss": 0.0442,
      "step": 269
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 0.6143440216760881,
      "learning_rate": 2.3388797000115427e-07,
      "loss": 0.0644,
      "step": 270
    },
    {
      "epoch": 1.8249158249158248,
      "grad_norm": 0.6184051449052006,
      "learning_rate": 2.163710018984766e-07,
      "loss": 0.0494,
      "step": 271
    },
    {
      "epoch": 1.8316498316498318,
      "grad_norm": 0.5174263740286721,
      "learning_rate": 1.9952128430483718e-07,
      "loss": 0.0733,
      "step": 272
    },
    {
      "epoch": 1.8383838383838382,
      "grad_norm": 0.5711547890486989,
      "learning_rate": 1.8334116752429243e-07,
      "loss": 0.0546,
      "step": 273
    },
    {
      "epoch": 1.8451178451178452,
      "grad_norm": 0.543912361463854,
      "learning_rate": 1.6783290846078714e-07,
      "loss": 0.0467,
      "step": 274
    },
    {
      "epoch": 1.8518518518518519,
      "grad_norm": 0.6068680622136619,
      "learning_rate": 1.5299867030334815e-07,
      "loss": 0.0569,
      "step": 275
    },
    {
      "epoch": 1.8585858585858586,
      "grad_norm": 0.5927823999637648,
      "learning_rate": 1.388405222243472e-07,
      "loss": 0.0533,
      "step": 276
    },
    {
      "epoch": 1.8653198653198653,
      "grad_norm": 0.47694438499884656,
      "learning_rate": 1.253604390908819e-07,
      "loss": 0.0564,
      "step": 277
    },
    {
      "epoch": 1.872053872053872,
      "grad_norm": 0.5019664348313649,
      "learning_rate": 1.1256030118930727e-07,
      "loss": 0.0421,
      "step": 278
    },
    {
      "epoch": 1.878787878787879,
      "grad_norm": 0.5244072558220748,
      "learning_rate": 1.0044189396296144e-07,
      "loss": 0.0443,
      "step": 279
    },
    {
      "epoch": 1.8855218855218854,
      "grad_norm": 0.5694979771175117,
      "learning_rate": 8.900690776312282e-08,
      "loss": 0.0519,
      "step": 280
    },
    {
      "epoch": 1.8922558922558923,
      "grad_norm": 0.5383345898694687,
      "learning_rate": 7.825693761322861e-08,
      "loss": 0.0645,
      "step": 281
    },
    {
      "epoch": 1.898989898989899,
      "grad_norm": 0.58996301402648,
      "learning_rate": 6.819348298638839e-08,
      "loss": 0.0534,
      "step": 282
    },
    {
      "epoch": 1.9057239057239057,
      "grad_norm": 0.4519126591505613,
      "learning_rate": 5.881794759623194e-08,
      "loss": 0.0397,
      "step": 283
    },
    {
      "epoch": 1.9124579124579124,
      "grad_norm": 0.459397062343545,
      "learning_rate": 5.013163920110864e-08,
      "loss": 0.0562,
      "step": 284
    },
    {
      "epoch": 1.9191919191919191,
      "grad_norm": 0.6692683465122977,
      "learning_rate": 4.21357694216723e-08,
      "loss": 0.0536,
      "step": 285
    },
    {
      "epoch": 1.925925925925926,
      "grad_norm": 0.5515458851174179,
      "learning_rate": 3.483145357187967e-08,
      "loss": 0.0709,
      "step": 286
    },
    {
      "epoch": 1.9326599326599325,
      "grad_norm": 0.7465197849269988,
      "learning_rate": 2.8219710503416543e-08,
      "loss": 0.0656,
      "step": 287
    },
    {
      "epoch": 1.9393939393939394,
      "grad_norm": 0.5871951635424402,
      "learning_rate": 2.230146246358256e-08,
      "loss": 0.0667,
      "step": 288
    },
    {
      "epoch": 1.9461279461279462,
      "grad_norm": 0.4842964057678575,
      "learning_rate": 1.7077534966650767e-08,
      "loss": 0.0408,
      "step": 289
    },
    {
      "epoch": 1.9528619528619529,
      "grad_norm": 0.5849783564235355,
      "learning_rate": 1.2548656678721404e-08,
      "loss": 0.067,
      "step": 290
    },
    {
      "epoch": 1.9595959595959596,
      "grad_norm": 0.6264347265804303,
      "learning_rate": 8.715459316078756e-09,
      "loss": 0.0492,
      "step": 291
    },
    {
      "epoch": 1.9663299663299663,
      "grad_norm": 0.5290968473577989,
      "learning_rate": 5.578477557081074e-09,
      "loss": 0.0543,
      "step": 292
    },
    {
      "epoch": 1.9730639730639732,
      "grad_norm": 0.4771694715604212,
      "learning_rate": 3.1381489675746946e-09,
      "loss": 0.0419,
      "step": 293
    },
    {
      "epoch": 1.9797979797979797,
      "grad_norm": 0.7387620656079369,
      "learning_rate": 1.3948139398628492e-09,
      "loss": 0.0613,
      "step": 294
    },
    {
      "epoch": 1.9865319865319866,
      "grad_norm": 0.5259388355087269,
      "learning_rate": 3.487156452258722e-10,
      "loss": 0.0405,
      "step": 295
    },
    {
      "epoch": 1.9932659932659933,
      "grad_norm": 0.7981313513015809,
      "learning_rate": 0.0,
      "loss": 0.0774,
      "step": 296
    },
    {
      "epoch": 1.9932659932659933,
      "eval_loss": 0.06412597745656967,
      "eval_runtime": 70.2541,
      "eval_samples_per_second": 28.44,
      "eval_steps_per_second": 0.897,
      "step": 296
    },
    {
      "epoch": 1.9932659932659933,
      "step": 296,
      "total_flos": 1.5329682729109094e+17,
      "train_loss": 0.098479394486325,
      "train_runtime": 5160.5268,
      "train_samples_per_second": 7.355,
      "train_steps_per_second": 0.057
    }
  ],
  "logging_steps": 1,
  "max_steps": 296,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.5329682729109094e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}