{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 20.0,
  "eval_steps": 500,
  "global_step": 283380,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.07057661091114405,
      "grad_norm": 0.8544681072235107,
      "learning_rate": 9.375e-06,
      "loss": 5.8075,
      "step": 1000
    },
    {
      "epoch": 0.1411532218222881,
      "grad_norm": 0.9348427057266235,
      "learning_rate": 1.875e-05,
      "loss": 3.9507,
      "step": 2000
    },
    {
      "epoch": 0.21172983273343213,
      "grad_norm": 0.9987612366676331,
      "learning_rate": 2.8125e-05,
      "loss": 3.6154,
      "step": 3000
    },
    {
      "epoch": 0.2823064436445762,
      "grad_norm": 0.9734600782394409,
      "learning_rate": 3.75e-05,
      "loss": 3.444,
      "step": 4000
    },
    {
      "epoch": 0.3528830545557202,
      "grad_norm": 0.9963154196739197,
      "learning_rate": 4.6874999999999994e-05,
      "loss": 3.2983,
      "step": 5000
    },
    {
      "epoch": 0.42345966546686425,
      "grad_norm": 0.8833392858505249,
      "learning_rate": 5.625e-05,
      "loss": 3.1806,
      "step": 6000
    },
    {
      "epoch": 0.49403627637800834,
      "grad_norm": 0.8395134806632996,
      "learning_rate": 6.5625e-05,
      "loss": 3.0762,
      "step": 7000
    },
    {
      "epoch": 0.5646128872891524,
      "grad_norm": 0.9056613445281982,
      "learning_rate": 7.5e-05,
      "loss": 3.0103,
      "step": 8000
    },
    {
      "epoch": 0.6351894982002965,
      "grad_norm": 0.8697331547737122,
      "learning_rate": 8.437499999999999e-05,
      "loss": 2.9146,
      "step": 9000
    },
    {
      "epoch": 0.7057661091114404,
      "grad_norm": 0.7743774056434631,
      "learning_rate": 9.374999999999999e-05,
      "loss": 2.8472,
      "step": 10000
    },
    {
      "epoch": 0.7763427200225845,
      "grad_norm": 0.769959568977356,
      "learning_rate": 0.00010312499999999999,
      "loss": 2.8093,
      "step": 11000
    },
    {
      "epoch": 0.8469193309337285,
      "grad_norm": 0.7926977872848511,
      "learning_rate": 0.0001125,
      "loss": 2.7498,
      "step": 12000
    },
    {
      "epoch": 0.9174959418448726,
      "grad_norm": 0.7256486415863037,
      "learning_rate": 0.000121865625,
      "loss": 2.714,
      "step": 13000
    },
    {
      "epoch": 0.9880725527560167,
      "grad_norm": 0.7287374138832092,
      "learning_rate": 0.000131240625,
      "loss": 2.6575,
      "step": 14000
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.4749594473317534,
      "eval_loss": 2.8570854663848877,
      "eval_runtime": 122.7592,
      "eval_samples_per_second": 382.464,
      "eval_steps_per_second": 5.979,
      "step": 14169
    },
    {
      "epoch": 1.0586491636671607,
      "grad_norm": 0.7397704124450684,
      "learning_rate": 0.00014060625,
      "loss": 2.6238,
      "step": 15000
    },
    {
      "epoch": 1.1292257745783048,
      "grad_norm": 0.7099004983901978,
      "learning_rate": 0.000149971875,
      "loss": 2.5914,
      "step": 16000
    },
    {
      "epoch": 1.1998023854894488,
      "grad_norm": 0.6785891056060791,
      "learning_rate": 0.000159346875,
      "loss": 2.5765,
      "step": 17000
    },
    {
      "epoch": 1.2703789964005927,
      "grad_norm": 0.6459276080131531,
      "learning_rate": 0.000168703125,
      "loss": 2.5484,
      "step": 18000
    },
    {
      "epoch": 1.340955607311737,
      "grad_norm": 0.6296180486679077,
      "learning_rate": 0.000178078125,
      "loss": 2.5329,
      "step": 19000
    },
    {
      "epoch": 1.4115322182228809,
      "grad_norm": 0.648757815361023,
      "learning_rate": 0.00018745312499999998,
      "loss": 2.5078,
      "step": 20000
    },
    {
      "epoch": 1.482108829134025,
      "grad_norm": 0.6126940250396729,
      "learning_rate": 0.00019681874999999998,
      "loss": 2.5066,
      "step": 21000
    },
    {
      "epoch": 1.552685440045169,
      "grad_norm": 0.5499350428581238,
      "learning_rate": 0.00020618437499999995,
      "loss": 2.4882,
      "step": 22000
    },
    {
      "epoch": 1.623262050956313,
      "grad_norm": 0.7012745141983032,
      "learning_rate": 0.00021555937499999998,
      "loss": 2.4746,
      "step": 23000
    },
    {
      "epoch": 1.6938386618674572,
      "grad_norm": 0.563831627368927,
      "learning_rate": 0.00022493437499999998,
      "loss": 2.4607,
      "step": 24000
    },
    {
      "epoch": 1.764415272778601,
      "grad_norm": 0.4928041696548462,
      "learning_rate": 0.00023430937499999997,
      "loss": 2.4454,
      "step": 25000
    },
    {
      "epoch": 1.8349918836897452,
      "grad_norm": 0.5389479398727417,
      "learning_rate": 0.00024367499999999997,
      "loss": 2.4429,
      "step": 26000
    },
    {
      "epoch": 1.9055684946008893,
      "grad_norm": 0.549089252948761,
      "learning_rate": 0.000253040625,
      "loss": 2.4239,
      "step": 27000
    },
    {
      "epoch": 1.9761451055120332,
      "grad_norm": 0.5027530193328857,
      "learning_rate": 0.000262415625,
      "loss": 2.4179,
      "step": 28000
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.4989820084802377,
      "eval_loss": 2.6257801055908203,
      "eval_runtime": 123.8319,
      "eval_samples_per_second": 379.151,
      "eval_steps_per_second": 5.927,
      "step": 28338
    },
    {
      "epoch": 2.0467217164231775,
      "grad_norm": 0.4511743485927582,
      "learning_rate": 0.000271790625,
      "loss": 2.3885,
      "step": 29000
    },
    {
      "epoch": 2.1172983273343213,
      "grad_norm": 0.4871021807193756,
      "learning_rate": 0.00028115624999999994,
      "loss": 2.378,
      "step": 30000
    },
    {
      "epoch": 2.1878749382454656,
      "grad_norm": 0.4743562638759613,
      "learning_rate": 0.00029053124999999994,
      "loss": 2.3721,
      "step": 31000
    },
    {
      "epoch": 2.2584515491566095,
      "grad_norm": 0.4668987989425659,
      "learning_rate": 0.00029990624999999993,
      "loss": 2.3584,
      "step": 32000
    },
    {
      "epoch": 2.3290281600677534,
      "grad_norm": 0.4746561050415039,
      "learning_rate": 0.00029881971517224914,
      "loss": 2.3594,
      "step": 33000
    },
    {
      "epoch": 2.3996047709788977,
      "grad_norm": 0.43889379501342773,
      "learning_rate": 0.0002976263028084971,
      "loss": 2.3468,
      "step": 34000
    },
    {
      "epoch": 2.4701813818900415,
      "grad_norm": 0.4571814239025116,
      "learning_rate": 0.00029643289044474497,
      "loss": 2.3515,
      "step": 35000
    },
    {
      "epoch": 2.5407579928011854,
      "grad_norm": 0.4081374704837799,
      "learning_rate": 0.0002952394780809929,
      "loss": 2.3321,
      "step": 36000
    },
    {
      "epoch": 2.6113346037123297,
      "grad_norm": 0.4098535180091858,
      "learning_rate": 0.0002940484525419683,
      "loss": 2.3259,
      "step": 37000
    },
    {
      "epoch": 2.681911214623474,
      "grad_norm": 0.4309552311897278,
      "learning_rate": 0.00029285504017821624,
      "loss": 2.3265,
      "step": 38000
    },
    {
      "epoch": 2.752487825534618,
      "grad_norm": 0.39243602752685547,
      "learning_rate": 0.0002916616278144641,
      "loss": 2.3119,
      "step": 39000
    },
    {
      "epoch": 2.8230644364457618,
      "grad_norm": 0.40959957242012024,
      "learning_rate": 0.0002904694088630758,
      "loss": 2.2985,
      "step": 40000
    },
    {
      "epoch": 2.893641047356906,
      "grad_norm": 0.35818567872047424,
      "learning_rate": 0.00028927599649932373,
      "loss": 2.3088,
      "step": 41000
    },
    {
      "epoch": 2.96421765826805,
      "grad_norm": 0.35112839937210083,
      "learning_rate": 0.0002880837775479354,
      "loss": 2.2903,
      "step": 42000
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5118994411509573,
      "eval_loss": 2.5130186080932617,
      "eval_runtime": 124.085,
      "eval_samples_per_second": 378.378,
      "eval_steps_per_second": 5.915,
      "step": 42507
    },
    {
      "epoch": 3.034794269179194,
      "grad_norm": 0.36554428935050964,
      "learning_rate": 0.0002868903651841833,
      "loss": 2.258,
      "step": 43000
    },
    {
      "epoch": 3.105370880090338,
      "grad_norm": 0.387246698141098,
      "learning_rate": 0.00028569814623279494,
      "loss": 2.2311,
      "step": 44000
    },
    {
      "epoch": 3.175947491001482,
      "grad_norm": 0.3703334927558899,
      "learning_rate": 0.00028450473386904283,
      "loss": 2.2455,
      "step": 45000
    },
    {
      "epoch": 3.2465241019126263,
      "grad_norm": 0.3958361744880676,
      "learning_rate": 0.00028331132150529077,
      "loss": 2.2292,
      "step": 46000
    },
    {
      "epoch": 3.31710071282377,
      "grad_norm": 0.4188041687011719,
      "learning_rate": 0.00028211910255390244,
      "loss": 2.2358,
      "step": 47000
    },
    {
      "epoch": 3.3876773237349145,
      "grad_norm": 0.39394712448120117,
      "learning_rate": 0.0002809256901901503,
      "loss": 2.2236,
      "step": 48000
    },
    {
      "epoch": 3.4582539346460583,
      "grad_norm": 0.36653855443000793,
      "learning_rate": 0.00027973227782639826,
      "loss": 2.2286,
      "step": 49000
    },
    {
      "epoch": 3.528830545557202,
      "grad_norm": 0.35694271326065063,
      "learning_rate": 0.00027853886546264615,
      "loss": 2.23,
      "step": 50000
    },
    {
      "epoch": 3.5994071564683465,
      "grad_norm": 0.331755131483078,
      "learning_rate": 0.0002773466465112578,
      "loss": 2.2242,
      "step": 51000
    },
    {
      "epoch": 3.6699837673794904,
      "grad_norm": 0.33919110894203186,
      "learning_rate": 0.0002761544275598695,
      "loss": 2.2175,
      "step": 52000
    },
    {
      "epoch": 3.7405603782906347,
      "grad_norm": 0.3551011383533478,
      "learning_rate": 0.0002749610151961174,
      "loss": 2.2124,
      "step": 53000
    },
    {
      "epoch": 3.8111369892017786,
      "grad_norm": 0.3324119746685028,
      "learning_rate": 0.0002737676028323653,
      "loss": 2.2112,
      "step": 54000
    },
    {
      "epoch": 3.8817136001129224,
      "grad_norm": 0.3544479012489319,
      "learning_rate": 0.00027257538388097697,
      "loss": 2.2114,
      "step": 55000
    },
    {
      "epoch": 3.9522902110240667,
      "grad_norm": 0.3497498333454132,
      "learning_rate": 0.0002713819715172249,
      "loss": 2.2222,
      "step": 56000
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5193685245326135,
      "eval_loss": 2.4525041580200195,
      "eval_runtime": 123.8971,
      "eval_samples_per_second": 378.952,
      "eval_steps_per_second": 5.924,
      "step": 56676
    },
    {
      "epoch": 4.022866821935211,
      "grad_norm": 0.3681366741657257,
      "learning_rate": 0.0002701885591534728,
      "loss": 2.1946,
      "step": 57000
    },
    {
      "epoch": 4.093443432846355,
      "grad_norm": 0.3502100110054016,
      "learning_rate": 0.0002689951467897207,
      "loss": 2.1505,
      "step": 58000
    },
    {
      "epoch": 4.164020043757499,
      "grad_norm": 0.34044522047042847,
      "learning_rate": 0.00026780292783833235,
      "loss": 2.1568,
      "step": 59000
    },
    {
      "epoch": 4.234596654668643,
      "grad_norm": 0.31450390815734863,
      "learning_rate": 0.0002666095154745803,
      "loss": 2.1578,
      "step": 60000
    },
    {
      "epoch": 4.3051732655797865,
      "grad_norm": 0.3611661493778229,
      "learning_rate": 0.00026541729652319196,
      "loss": 2.1543,
      "step": 61000
    },
    {
      "epoch": 4.375749876490931,
      "grad_norm": 0.3503958582878113,
      "learning_rate": 0.00026422388415943984,
      "loss": 2.1588,
      "step": 62000
    },
    {
      "epoch": 4.446326487402075,
      "grad_norm": 0.33794161677360535,
      "learning_rate": 0.0002630304717956878,
      "loss": 2.1623,
      "step": 63000
    },
    {
      "epoch": 4.516903098313219,
      "grad_norm": 0.38356584310531616,
      "learning_rate": 0.0002618394462566632,
      "loss": 2.1508,
      "step": 64000
    },
    {
      "epoch": 4.587479709224363,
      "grad_norm": 0.3673034906387329,
      "learning_rate": 0.0002606460338929111,
      "loss": 2.1532,
      "step": 65000
    },
    {
      "epoch": 4.658056320135507,
      "grad_norm": 0.3584080934524536,
      "learning_rate": 0.00025945262152915905,
      "loss": 2.1498,
      "step": 66000
    },
    {
      "epoch": 4.7286329310466515,
      "grad_norm": 0.35382455587387085,
      "learning_rate": 0.00025825920916540694,
      "loss": 2.1587,
      "step": 67000
    },
    {
      "epoch": 4.799209541957795,
      "grad_norm": 0.3333965539932251,
      "learning_rate": 0.0002570669902140186,
      "loss": 2.137,
      "step": 68000
    },
    {
      "epoch": 4.869786152868939,
      "grad_norm": 0.3372107446193695,
      "learning_rate": 0.00025587477126263027,
      "loss": 2.1408,
      "step": 69000
    },
    {
      "epoch": 4.940362763780083,
      "grad_norm": 0.33577457070350647,
      "learning_rate": 0.00025468135889887816,
      "loss": 2.1477,
      "step": 70000
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5237092533013349,
      "eval_loss": 2.4206960201263428,
      "eval_runtime": 123.8441,
      "eval_samples_per_second": 379.114,
      "eval_steps_per_second": 5.927,
      "step": 70845
    },
    {
      "epoch": 5.010939374691227,
      "grad_norm": 0.34049326181411743,
      "learning_rate": 0.0002534879465351261,
      "loss": 2.1438,
      "step": 71000
    },
    {
      "epoch": 5.081515985602372,
      "grad_norm": 0.3590817451477051,
      "learning_rate": 0.000252294534171374,
      "loss": 2.0823,
      "step": 72000
    },
    {
      "epoch": 5.152092596513516,
      "grad_norm": 0.3182990550994873,
      "learning_rate": 0.00025110112180762187,
      "loss": 2.095,
      "step": 73000
    },
    {
      "epoch": 5.2226692074246595,
      "grad_norm": 0.32062873244285583,
      "learning_rate": 0.00024990890285623354,
      "loss": 2.0931,
      "step": 74000
    },
    {
      "epoch": 5.293245818335803,
      "grad_norm": 0.3651541471481323,
      "learning_rate": 0.0002487154904924815,
      "loss": 2.1021,
      "step": 75000
    },
    {
      "epoch": 5.363822429246947,
      "grad_norm": 0.3623022735118866,
      "learning_rate": 0.00024752327154109314,
      "loss": 2.1134,
      "step": 76000
    },
    {
      "epoch": 5.434399040158092,
      "grad_norm": 0.3717166483402252,
      "learning_rate": 0.0002463298591773411,
      "loss": 2.0979,
      "step": 77000
    },
    {
      "epoch": 5.504975651069236,
      "grad_norm": 0.34112003445625305,
      "learning_rate": 0.00024513764022595275,
      "loss": 2.1,
      "step": 78000
    },
    {
      "epoch": 5.57555226198038,
      "grad_norm": 0.3315085172653198,
      "learning_rate": 0.00024394422786220063,
      "loss": 2.1084,
      "step": 79000
    },
    {
      "epoch": 5.6461288728915235,
      "grad_norm": 0.31678348779678345,
      "learning_rate": 0.0002427520089108123,
      "loss": 2.102,
      "step": 80000
    },
    {
      "epoch": 5.716705483802667,
      "grad_norm": 0.3307001292705536,
      "learning_rate": 0.00024155978995942396,
      "loss": 2.1013,
      "step": 81000
    },
    {
      "epoch": 5.787282094713812,
      "grad_norm": 0.31418758630752563,
      "learning_rate": 0.00024036637759567185,
      "loss": 2.0996,
      "step": 82000
    },
    {
      "epoch": 5.857858705624956,
      "grad_norm": 0.34750932455062866,
      "learning_rate": 0.0002391729652319198,
      "loss": 2.102,
      "step": 83000
    },
    {
      "epoch": 5.9284353165361,
      "grad_norm": 0.3376181423664093,
      "learning_rate": 0.00023797955286816768,
      "loss": 2.1014,
      "step": 84000
    },
    {
      "epoch": 5.999011927447244,
      "grad_norm": 0.35281479358673096,
      "learning_rate": 0.0002367861405044156,
      "loss": 2.0967,
      "step": 85000
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.5265582372615145,
      "eval_loss": 2.399027109146118,
      "eval_runtime": 124.2581,
      "eval_samples_per_second": 377.851,
      "eval_steps_per_second": 5.907,
      "step": 85014
    },
    {
      "epoch": 6.069588538358388,
      "grad_norm": 0.3424261212348938,
      "learning_rate": 0.00023559511496539103,
      "loss": 2.0376,
      "step": 86000
    },
    {
      "epoch": 6.140165149269532,
      "grad_norm": 0.3347076177597046,
      "learning_rate": 0.00023440170260163892,
      "loss": 2.0425,
      "step": 87000
    },
    {
      "epoch": 6.210741760180676,
      "grad_norm": 0.33315178751945496,
      "learning_rate": 0.00023320829023788686,
      "loss": 2.0483,
      "step": 88000
    },
    {
      "epoch": 6.28131837109182,
      "grad_norm": 0.3443340063095093,
      "learning_rate": 0.00023201487787413475,
      "loss": 2.0431,
      "step": 89000
    },
    {
      "epoch": 6.351894982002964,
      "grad_norm": 0.35432785749435425,
      "learning_rate": 0.0002308226589227464,
      "loss": 2.0623,
      "step": 90000
    },
    {
      "epoch": 6.422471592914108,
      "grad_norm": 0.32286491990089417,
      "learning_rate": 0.00022962924655899432,
      "loss": 2.0542,
      "step": 91000
    },
    {
      "epoch": 6.493048203825253,
      "grad_norm": 0.3599454164505005,
      "learning_rate": 0.000228437027607606,
      "loss": 2.0582,
      "step": 92000
    },
    {
      "epoch": 6.5636248147363965,
      "grad_norm": 0.3492770791053772,
      "learning_rate": 0.0002272436152438539,
      "loss": 2.0642,
      "step": 93000
    },
    {
      "epoch": 6.63420142564754,
      "grad_norm": 0.37773609161376953,
      "learning_rate": 0.0002260513962924656,
      "loss": 2.0637,
      "step": 94000
    },
    {
      "epoch": 6.704778036558684,
      "grad_norm": 0.3557668924331665,
      "learning_rate": 0.00022485798392871348,
      "loss": 2.0659,
      "step": 95000
    },
    {
      "epoch": 6.775354647469829,
      "grad_norm": 0.3349735736846924,
      "learning_rate": 0.0002236645715649614,
      "loss": 2.0635,
      "step": 96000
    },
    {
      "epoch": 6.845931258380973,
      "grad_norm": 0.33342695236206055,
      "learning_rate": 0.00022247235261357306,
      "loss": 2.0614,
      "step": 97000
    },
    {
      "epoch": 6.916507869292117,
      "grad_norm": 0.3801983892917633,
      "learning_rate": 0.00022127894024982097,
      "loss": 2.0642,
      "step": 98000
    },
    {
      "epoch": 6.987084480203261,
      "grad_norm": 0.332112193107605,
      "learning_rate": 0.0002200855278860689,
      "loss": 2.0608,
      "step": 99000
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.5284439684489343,
      "eval_loss": 2.38849139213562,
      "eval_runtime": 124.26,
      "eval_samples_per_second": 377.845,
      "eval_steps_per_second": 5.907,
      "step": 99183
    },
    {
      "epoch": 7.057661091114404,
      "grad_norm": 0.3257732391357422,
      "learning_rate": 0.00021889330893468055,
      "loss": 2.0136,
      "step": 100000
    },
    {
      "epoch": 7.128237702025549,
      "grad_norm": 0.3609803020954132,
      "learning_rate": 0.00021769989657092844,
      "loss": 2.0096,
      "step": 101000
    },
    {
      "epoch": 7.198814312936693,
      "grad_norm": 0.35413938760757446,
      "learning_rate": 0.00021650648420717638,
      "loss": 2.0028,
      "step": 102000
    },
    {
      "epoch": 7.269390923847837,
      "grad_norm": 0.3548286557197571,
      "learning_rate": 0.00021531426525578804,
      "loss": 2.0097,
      "step": 103000
    },
    {
      "epoch": 7.339967534758981,
      "grad_norm": 0.3535342812538147,
      "learning_rate": 0.00021412085289203593,
      "loss": 2.018,
      "step": 104000
    },
    {
      "epoch": 7.410544145670125,
      "grad_norm": 0.3639167547225952,
      "learning_rate": 0.00021292744052828384,
      "loss": 2.0153,
      "step": 105000
    },
    {
      "epoch": 7.481120756581269,
      "grad_norm": 0.3649103343486786,
      "learning_rate": 0.00021173402816453179,
      "loss": 2.0228,
      "step": 106000
    },
    {
      "epoch": 7.551697367492413,
      "grad_norm": 0.374751478433609,
      "learning_rate": 0.00021054180921314345,
      "loss": 2.0248,
      "step": 107000
    },
    {
      "epoch": 7.622273978403557,
      "grad_norm": 0.39019080996513367,
      "learning_rate": 0.00020934959026175511,
      "loss": 2.0166,
      "step": 108000
    },
    {
      "epoch": 7.692850589314701,
      "grad_norm": 0.36250707507133484,
      "learning_rate": 0.000208156177898003,
      "loss": 2.0291,
      "step": 109000
    },
    {
      "epoch": 7.763427200225845,
      "grad_norm": 0.3835802972316742,
      "learning_rate": 0.00020696276553425091,
      "loss": 2.0311,
      "step": 110000
    },
    {
      "epoch": 7.83400381113699,
      "grad_norm": 0.3630106449127197,
      "learning_rate": 0.00020577054658286258,
      "loss": 2.0233,
      "step": 111000
    },
    {
      "epoch": 7.9045804220481335,
      "grad_norm": 0.3512779474258423,
      "learning_rate": 0.0002045771342191105,
      "loss": 2.0433,
      "step": 112000
    },
    {
      "epoch": 7.975157032959277,
      "grad_norm": 0.3067089915275574,
      "learning_rate": 0.00020338491526772218,
      "loss": 2.0231,
      "step": 113000
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.5295622798070709,
      "eval_loss": 2.3815085887908936,
      "eval_runtime": 125.1879,
      "eval_samples_per_second": 375.044,
      "eval_steps_per_second": 5.863,
      "step": 113352
    },
    {
      "epoch": 8.045733643870422,
      "grad_norm": 0.35720863938331604,
      "learning_rate": 0.00020219150290397007,
      "loss": 1.9788,
      "step": 114000
    },
    {
      "epoch": 8.116310254781565,
      "grad_norm": 0.37242060899734497,
      "learning_rate": 0.00020099809054021796,
      "loss": 1.9589,
      "step": 115000
    },
    {
      "epoch": 8.18688686569271,
      "grad_norm": 0.3982705771923065,
      "learning_rate": 0.0001998046781764659,
      "loss": 1.9783,
      "step": 116000
    },
    {
      "epoch": 8.257463476603853,
      "grad_norm": 0.3767191767692566,
      "learning_rate": 0.00019861245922507756,
      "loss": 1.9766,
      "step": 117000
    },
    {
      "epoch": 8.328040087514998,
      "grad_norm": 0.3721235394477844,
      "learning_rate": 0.00019742024027368923,
      "loss": 1.9867,
      "step": 118000
    },
    {
      "epoch": 8.398616698426142,
      "grad_norm": 0.4111528694629669,
      "learning_rate": 0.00019622682790993714,
      "loss": 1.9928,
      "step": 119000
    },
    {
      "epoch": 8.469193309337285,
      "grad_norm": 0.3644442558288574,
      "learning_rate": 0.00019503341554618503,
      "loss": 1.9861,
      "step": 120000
    },
    {
      "epoch": 8.53976992024843,
      "grad_norm": 0.37768858671188354,
      "learning_rate": 0.00019384000318243297,
      "loss": 1.9878,
      "step": 121000
    },
    {
      "epoch": 8.610346531159573,
      "grad_norm": 0.3622952997684479,
      "learning_rate": 0.00019264659081868086,
      "loss": 1.9932,
      "step": 122000
    },
    {
      "epoch": 8.680923142070718,
      "grad_norm": 0.3601091802120209,
      "learning_rate": 0.00019145437186729252,
      "loss": 1.9977,
      "step": 123000
    },
    {
      "epoch": 8.751499752981863,
      "grad_norm": 0.3668394386768341,
      "learning_rate": 0.00019026095950354043,
      "loss": 1.9926,
      "step": 124000
    },
    {
      "epoch": 8.822076363893006,
      "grad_norm": 0.4132552146911621,
      "learning_rate": 0.0001890687405521521,
      "loss": 2.0003,
      "step": 125000
    },
    {
      "epoch": 8.89265297480415,
      "grad_norm": 0.34833863377571106,
      "learning_rate": 0.00018787532818840004,
      "loss": 2.0056,
      "step": 126000
    },
    {
      "epoch": 8.963229585715293,
      "grad_norm": 0.33041203022003174,
      "learning_rate": 0.00018668310923701168,
      "loss": 1.9966,
      "step": 127000
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.5303685678760878,
      "eval_loss": 2.3803319931030273,
      "eval_runtime": 124.6238,
      "eval_samples_per_second": 376.742,
      "eval_steps_per_second": 5.89,
      "step": 127521
    },
    {
      "epoch": 9.033806196626438,
      "grad_norm": 0.3751468360424042,
      "learning_rate": 0.0001854896968732596,
      "loss": 1.9627,
      "step": 128000
    },
    {
      "epoch": 9.104382807537583,
      "grad_norm": 0.4026428461074829,
      "learning_rate": 0.00018429747792187125,
      "loss": 1.9282,
      "step": 129000
    },
    {
      "epoch": 9.174959418448726,
      "grad_norm": 0.3918458819389343,
      "learning_rate": 0.00018310406555811917,
      "loss": 1.9479,
      "step": 130000
    },
    {
      "epoch": 9.24553602935987,
      "grad_norm": 0.3758170008659363,
      "learning_rate": 0.00018191184660673083,
      "loss": 1.946,
      "step": 131000
    },
    {
      "epoch": 9.316112640271013,
      "grad_norm": 0.38886573910713196,
      "learning_rate": 0.00018071843424297872,
      "loss": 1.9459,
      "step": 132000
    },
    {
      "epoch": 9.386689251182158,
      "grad_norm": 0.3674320876598358,
      "learning_rate": 0.00017952621529159038,
      "loss": 1.958,
      "step": 133000
    },
    {
      "epoch": 9.457265862093303,
      "grad_norm": 0.38517245650291443,
      "learning_rate": 0.00017833280292783833,
      "loss": 1.9553,
      "step": 134000
    },
    {
      "epoch": 9.527842473004446,
      "grad_norm": 0.3472743630409241,
      "learning_rate": 0.0001771393905640862,
      "loss": 1.9591,
      "step": 135000
    },
    {
      "epoch": 9.59841908391559,
      "grad_norm": 0.38845178484916687,
      "learning_rate": 0.0001759471716126979,
      "loss": 1.961,
      "step": 136000
    },
    {
      "epoch": 9.668995694826734,
      "grad_norm": 0.3705516755580902,
      "learning_rate": 0.0001747537592489458,
      "loss": 1.9711,
      "step": 137000
    },
    {
      "epoch": 9.739572305737878,
      "grad_norm": 0.3988391160964966,
      "learning_rate": 0.00017356154029755745,
      "loss": 1.9713,
      "step": 138000
    },
    {
      "epoch": 9.810148916649023,
      "grad_norm": 0.39870232343673706,
      "learning_rate": 0.0001723681279338054,
      "loss": 1.9709,
      "step": 139000
    },
    {
      "epoch": 9.880725527560166,
      "grad_norm": 0.36726516485214233,
      "learning_rate": 0.00017117590898241706,
      "loss": 1.9721,
      "step": 140000
    },
    {
      "epoch": 9.951302138471311,
      "grad_norm": 0.36698198318481445,
      "learning_rate": 0.00016998249661866495,
      "loss": 1.974,
      "step": 141000
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.5310687763142748,
      "eval_loss": 2.380207061767578,
      "eval_runtime": 125.3944,
      "eval_samples_per_second": 374.427,
      "eval_steps_per_second": 5.854,
      "step": 141690
    },
    {
      "epoch": 10.021878749382454,
      "grad_norm": 0.4232293367385864,
      "learning_rate": 0.0001687902776672766,
      "loss": 1.9502,
      "step": 142000
    },
    {
      "epoch": 10.092455360293599,
      "grad_norm": 0.3572341799736023,
      "learning_rate": 0.00016759686530352452,
      "loss": 1.9131,
      "step": 143000
    },
    {
      "epoch": 10.163031971204743,
      "grad_norm": 0.4433085322380066,
      "learning_rate": 0.00016640345293977247,
      "loss": 1.9064,
      "step": 144000
    },
    {
      "epoch": 10.233608582115886,
      "grad_norm": 0.4200994074344635,
      "learning_rate": 0.00016521123398838413,
      "loss": 1.9154,
      "step": 145000
    },
    {
      "epoch": 10.304185193027031,
      "grad_norm": 0.3958752751350403,
      "learning_rate": 0.00016401782162463202,
      "loss": 1.9273,
      "step": 146000
    },
    {
      "epoch": 10.374761803938174,
      "grad_norm": 0.4018735885620117,
      "learning_rate": 0.00016282560267324368,
      "loss": 1.9299,
      "step": 147000
    },
    {
      "epoch": 10.445338414849319,
      "grad_norm": 0.3771366477012634,
      "learning_rate": 0.00016163338372185535,
      "loss": 1.9281,
      "step": 148000
    },
    {
      "epoch": 10.515915025760464,
      "grad_norm": 0.3960859179496765,
      "learning_rate": 0.00016043997135810326,
      "loss": 1.9319,
      "step": 149000
    },
    {
      "epoch": 10.586491636671607,
      "grad_norm": 0.42455655336380005,
      "learning_rate": 0.00015924655899435117,
      "loss": 1.9339,
      "step": 150000
    },
    {
      "epoch": 10.657068247582751,
      "grad_norm": 0.3864282965660095,
      "learning_rate": 0.0001580531466305991,
      "loss": 1.931,
      "step": 151000
    },
    {
      "epoch": 10.727644858493894,
      "grad_norm": 0.42341336607933044,
      "learning_rate": 0.00015686092767921075,
      "loss": 1.9476,
      "step": 152000
    },
    {
      "epoch": 10.79822146940504,
      "grad_norm": 0.3670462369918823,
      "learning_rate": 0.00015566870872782242,
      "loss": 1.9418,
      "step": 153000
    },
    {
      "epoch": 10.868798080316184,
      "grad_norm": 0.39386090636253357,
      "learning_rate": 0.0001544752963640703,
      "loss": 1.9427,
      "step": 154000
    },
    {
      "epoch": 10.939374691227327,
      "grad_norm": 0.3562389314174652,
      "learning_rate": 0.00015328188400031824,
      "loss": 1.9469,
      "step": 155000
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.531509172531641,
      "eval_loss": 2.3798038959503174,
      "eval_runtime": 124.5053,
      "eval_samples_per_second": 377.1,
      "eval_steps_per_second": 5.895,
      "step": 155859
    },
    {
      "epoch": 11.009951302138472,
      "grad_norm": 0.4261242151260376,
      "learning_rate": 0.00015208847163656616,
      "loss": 1.9379,
      "step": 156000
    },
    {
      "epoch": 11.080527913049615,
      "grad_norm": 0.4348481595516205,
      "learning_rate": 0.00015089625268517782,
      "loss": 1.8776,
      "step": 157000
    },
    {
      "epoch": 11.15110452396076,
      "grad_norm": 0.41202861070632935,
      "learning_rate": 0.0001497028403214257,
      "loss": 1.8838,
      "step": 158000
    },
    {
      "epoch": 11.221681134871904,
      "grad_norm": 0.40852221846580505,
      "learning_rate": 0.00014851062137003737,
      "loss": 1.8916,
      "step": 159000
    },
    {
      "epoch": 11.292257745783047,
      "grad_norm": 0.4663460850715637,
      "learning_rate": 0.0001473172090062853,
      "loss": 1.8901,
      "step": 160000
    },
    {
      "epoch": 11.362834356694192,
      "grad_norm": 0.4087861180305481,
      "learning_rate": 0.0001461237966425332,
      "loss": 1.9009,
      "step": 161000
    },
    {
      "epoch": 11.433410967605335,
      "grad_norm": 0.38091129064559937,
      "learning_rate": 0.00014493038427878111,
      "loss": 1.909,
      "step": 162000
    },
    {
      "epoch": 11.50398757851648,
      "grad_norm": 0.3991438150405884,
      "learning_rate": 0.00014373816532739278,
      "loss": 1.903,
      "step": 163000
    },
    {
      "epoch": 11.574564189427624,
      "grad_norm": 0.3866070806980133,
      "learning_rate": 0.0001425447529636407,
      "loss": 1.9105,
      "step": 164000
    },
    {
      "epoch": 11.645140800338767,
      "grad_norm": 0.3476414084434509,
      "learning_rate": 0.0001413513405998886,
      "loss": 1.9135,
      "step": 165000
    },
    {
      "epoch": 11.715717411249912,
      "grad_norm": 0.3765086233615875,
      "learning_rate": 0.00014015912164850027,
      "loss": 1.9097,
      "step": 166000
    },
    {
      "epoch": 11.786294022161055,
      "grad_norm": 0.37178364396095276,
      "learning_rate": 0.00013896570928474819,
      "loss": 1.9263,
      "step": 167000
    },
    {
      "epoch": 11.8568706330722,
      "grad_norm": 0.42377331852912903,
      "learning_rate": 0.0001377722969209961,
      "loss": 1.9175,
      "step": 168000
    },
    {
      "epoch": 11.927447243983345,
      "grad_norm": 0.37589961290359497,
      "learning_rate": 0.00013658007796960776,
      "loss": 1.9259,
      "step": 169000
    },
    {
      "epoch": 11.998023854894488,
      "grad_norm": 0.4170770049095154,
      "learning_rate": 0.00013538666560585568,
      "loss": 1.9241,
      "step": 170000
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.5317882945602614,
      "eval_loss": 2.385725975036621,
      "eval_runtime": 125.4957,
      "eval_samples_per_second": 374.124,
      "eval_steps_per_second": 5.849,
      "step": 170028
    },
    {
      "epoch": 12.068600465805632,
      "grad_norm": 0.45514151453971863,
      "learning_rate": 0.00013419325324210356,
      "loss": 1.8578,
      "step": 171000
    },
    {
      "epoch": 12.139177076716775,
      "grad_norm": 0.4281666576862335,
      "learning_rate": 0.00013299984087835148,
      "loss": 1.8558,
      "step": 172000
    },
    {
      "epoch": 12.20975368762792,
      "grad_norm": 0.41793227195739746,
      "learning_rate": 0.0001318088153393269,
      "loss": 1.8685,
      "step": 173000
    },
    {
      "epoch": 12.280330298539065,
      "grad_norm": 0.48373010754585266,
      "learning_rate": 0.0001306154029755748,
      "loss": 1.8829,
      "step": 174000
    },
    {
      "epoch": 12.350906909450208,
      "grad_norm": 0.486074835062027,
      "learning_rate": 0.00012942199061182275,
      "loss": 1.8712,
      "step": 175000
    },
    {
      "epoch": 12.421483520361353,
      "grad_norm": 0.3793754279613495,
      "learning_rate": 0.00012822857824807063,
      "loss": 1.8786,
      "step": 176000
    },
    {
      "epoch": 12.492060131272495,
      "grad_norm": 0.4211224317550659,
      "learning_rate": 0.00012703755270904605,
      "loss": 1.8852,
      "step": 177000
    },
    {
      "epoch": 12.56263674218364,
      "grad_norm": 0.42993125319480896,
      "learning_rate": 0.00012584414034529396,
      "loss": 1.8953,
      "step": 178000
    },
    {
      "epoch": 12.633213353094785,
      "grad_norm": 0.39096274971961975,
      "learning_rate": 0.00012465072798154188,
      "loss": 1.8927,
      "step": 179000
    },
    {
      "epoch": 12.703789964005928,
      "grad_norm": 0.4215902090072632,
      "learning_rate": 0.0001234573156177898,
      "loss": 1.8854,
      "step": 180000
    },
    {
      "epoch": 12.774366574917073,
      "grad_norm": 0.37804627418518066,
      "learning_rate": 0.00012226509666640146,
      "loss": 1.8858,
      "step": 181000
    },
    {
      "epoch": 12.844943185828216,
      "grad_norm": 0.3929870128631592,
      "learning_rate": 0.00012107168430264937,
      "loss": 1.8907,
      "step": 182000
    },
    {
      "epoch": 12.91551979673936,
      "grad_norm": 0.3814956545829773,
      "learning_rate": 0.00011987946535126103,
      "loss": 1.8936,
      "step": 183000
    },
    {
      "epoch": 12.986096407650505,
      "grad_norm": 0.43260428309440613,
      "learning_rate": 0.00011868605298750895,
      "loss": 1.8975,
      "step": 184000
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.5319071368966752,
      "eval_loss": 2.390778064727783,
      "eval_runtime": 124.2689,
      "eval_samples_per_second": 377.818,
      "eval_steps_per_second": 5.907,
      "step": 184197
    },
    {
      "epoch": 13.056673018561648,
      "grad_norm": 0.4306144714355469,
      "learning_rate": 0.00011749383403612061,
      "loss": 1.8527,
      "step": 185000
    },
    {
      "epoch": 13.127249629472793,
      "grad_norm": 0.4380647838115692,
      "learning_rate": 0.00011630042167236851,
      "loss": 1.8438,
      "step": 186000
    },
    {
      "epoch": 13.197826240383936,
      "grad_norm": 0.39878422021865845,
      "learning_rate": 0.00011510820272098018,
      "loss": 1.8457,
      "step": 187000
    },
    {
      "epoch": 13.26840285129508,
      "grad_norm": 0.41546088457107544,
      "learning_rate": 0.00011391479035722809,
      "loss": 1.8498,
      "step": 188000
    },
    {
      "epoch": 13.338979462206225,
      "grad_norm": 0.4789869785308838,
      "learning_rate": 0.000112721377993476,
      "loss": 1.8485,
      "step": 189000
    },
    {
      "epoch": 13.409556073117368,
      "grad_norm": 0.45348218083381653,
      "learning_rate": 0.00011153035245445142,
      "loss": 1.8562,
      "step": 190000
    },
    {
      "epoch": 13.480132684028513,
      "grad_norm": 0.422708123922348,
      "learning_rate": 0.00011033694009069932,
      "loss": 1.8608,
      "step": 191000
    },
    {
      "epoch": 13.550709294939658,
      "grad_norm": 0.4499293267726898,
      "learning_rate": 0.00010914352772694723,
      "loss": 1.8608,
      "step": 192000
    },
    {
      "epoch": 13.621285905850801,
      "grad_norm": 0.45754432678222656,
      "learning_rate": 0.00010795011536319516,
      "loss": 1.8639,
      "step": 193000
    },
    {
      "epoch": 13.691862516761946,
      "grad_norm": 0.468192458152771,
      "learning_rate": 0.00010675789641180682,
      "loss": 1.8718,
      "step": 194000
    },
    {
      "epoch": 13.762439127673089,
      "grad_norm": 0.5351200103759766,
      "learning_rate": 0.00010556448404805473,
      "loss": 1.8667,
      "step": 195000
    },
    {
      "epoch": 13.833015738584233,
      "grad_norm": 0.40231311321258545,
      "learning_rate": 0.00010437226509666639,
      "loss": 1.8673,
      "step": 196000
    },
    {
      "epoch": 13.903592349495376,
      "grad_norm": 0.45186904072761536,
      "learning_rate": 0.0001031788527329143,
      "loss": 1.8697,
      "step": 197000
    },
    {
      "epoch": 13.974168960406521,
      "grad_norm": 0.4116153120994568,
      "learning_rate": 0.00010198663378152597,
      "loss": 1.8763,
      "step": 198000
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.5319930779725895,
      "eval_loss": 2.394319772720337,
      "eval_runtime": 124.9604,
      "eval_samples_per_second": 375.727,
      "eval_steps_per_second": 5.874,
      "step": 198366
    },
    {
      "epoch": 14.044745571317666,
      "grad_norm": 0.47304603457450867,
      "learning_rate": 0.00010079322141777388,
      "loss": 1.8384,
      "step": 199000
    },
    {
      "epoch": 14.115322182228809,
      "grad_norm": 0.5089701414108276,
      "learning_rate": 9.960100246638555e-05,
      "loss": 1.8174,
      "step": 200000
    },
    {
      "epoch": 14.185898793139954,
      "grad_norm": 0.4620361924171448,
      "learning_rate": 9.840759010263345e-05,
      "loss": 1.8259,
      "step": 201000
    },
    {
      "epoch": 14.256475404051098,
      "grad_norm": 0.42748963832855225,
      "learning_rate": 9.721417773888136e-05,
      "loss": 1.8259,
      "step": 202000
    },
    {
      "epoch": 14.327052014962241,
      "grad_norm": 0.48661568760871887,
      "learning_rate": 9.602076537512929e-05,
      "loss": 1.8358,
      "step": 203000
    },
    {
      "epoch": 14.397628625873386,
      "grad_norm": 0.417021781206131,
      "learning_rate": 9.482854642374095e-05,
      "loss": 1.8365,
      "step": 204000
    },
    {
      "epoch": 14.468205236784529,
      "grad_norm": 0.42631858587265015,
      "learning_rate": 9.363632747235262e-05,
      "loss": 1.8396,
      "step": 205000
    },
    {
      "epoch": 14.538781847695674,
      "grad_norm": 0.4228135049343109,
      "learning_rate": 9.244291510860052e-05,
      "loss": 1.8435,
      "step": 206000
    },
    {
      "epoch": 14.609358458606817,
      "grad_norm": 0.4370954632759094,
      "learning_rate": 9.124950274484843e-05,
      "loss": 1.8386,
      "step": 207000
    },
    {
      "epoch": 14.679935069517962,
      "grad_norm": 0.4384244680404663,
      "learning_rate": 9.005609038109634e-05,
      "loss": 1.8463,
      "step": 208000
    },
    {
      "epoch": 14.750511680429106,
      "grad_norm": 0.48162171244621277,
      "learning_rate": 8.886387142970801e-05,
      "loss": 1.8399,
      "step": 209000
    },
    {
      "epoch": 14.82108829134025,
      "grad_norm": 0.44134023785591125,
      "learning_rate": 8.767165247831966e-05,
      "loss": 1.8454,
      "step": 210000
    },
    {
      "epoch": 14.891664902251394,
      "grad_norm": 0.40469464659690857,
      "learning_rate": 8.647824011456757e-05,
      "loss": 1.8504,
      "step": 211000
    },
    {
      "epoch": 14.962241513162539,
      "grad_norm": 0.45577868819236755,
      "learning_rate": 8.528482775081549e-05,
      "loss": 1.8519,
      "step": 212000
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.5317183483036831,
      "eval_loss": 2.4029204845428467,
      "eval_runtime": 124.5491,
      "eval_samples_per_second": 376.968,
      "eval_steps_per_second": 5.893,
      "step": 212535
    },
    {
      "epoch": 15.032818124073682,
      "grad_norm": 0.4118717610836029,
      "learning_rate": 8.409260879942715e-05,
      "loss": 1.823,
      "step": 213000
    },
    {
      "epoch": 15.103394734984827,
      "grad_norm": 0.4281434118747711,
      "learning_rate": 8.289919643567505e-05,
      "loss": 1.7981,
      "step": 214000
    },
    {
      "epoch": 15.17397134589597,
      "grad_norm": 0.48909133672714233,
      "learning_rate": 8.170697748428673e-05,
      "loss": 1.8034,
      "step": 215000
    },
    {
      "epoch": 15.244547956807114,
      "grad_norm": 0.43051743507385254,
      "learning_rate": 8.051356512053464e-05,
      "loss": 1.8084,
      "step": 216000
    },
    {
      "epoch": 15.315124567718257,
      "grad_norm": 0.442118376493454,
      "learning_rate": 7.932015275678256e-05,
      "loss": 1.815,
      "step": 217000
    },
    {
      "epoch": 15.385701178629402,
      "grad_norm": 0.4444899559020996,
      "learning_rate": 7.812674039303046e-05,
      "loss": 1.8117,
      "step": 218000
    },
    {
      "epoch": 15.456277789540547,
      "grad_norm": 0.45674997568130493,
      "learning_rate": 7.693452144164212e-05,
      "loss": 1.8143,
      "step": 219000
    },
    {
      "epoch": 15.52685440045169,
      "grad_norm": 0.5022088289260864,
      "learning_rate": 7.574230249025379e-05,
      "loss": 1.8165,
      "step": 220000
    },
    {
      "epoch": 15.597431011362834,
      "grad_norm": 0.5027050971984863,
      "learning_rate": 7.45488901265017e-05,
      "loss": 1.8236,
      "step": 221000
    },
    {
      "epoch": 15.66800762227398,
      "grad_norm": 0.49043428897857666,
      "learning_rate": 7.335547776274961e-05,
      "loss": 1.8207,
      "step": 222000
    },
    {
      "epoch": 15.738584233185122,
      "grad_norm": 0.4271089732646942,
      "learning_rate": 7.216325881136128e-05,
      "loss": 1.8182,
      "step": 223000
    },
    {
      "epoch": 15.809160844096267,
      "grad_norm": 0.38394105434417725,
      "learning_rate": 7.096984644760919e-05,
      "loss": 1.8252,
      "step": 224000
    },
    {
      "epoch": 15.87973745500741,
      "grad_norm": 0.4386895000934601,
      "learning_rate": 6.97764340838571e-05,
      "loss": 1.8353,
      "step": 225000
    },
    {
      "epoch": 15.950314065918555,
      "grad_norm": 0.44788655638694763,
      "learning_rate": 6.858421513246877e-05,
      "loss": 1.8306,
      "step": 226000
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.5316992871200185,
      "eval_loss": 2.409809112548828,
      "eval_runtime": 124.7218,
      "eval_samples_per_second": 376.446,
      "eval_steps_per_second": 5.885,
      "step": 226704
    },
    {
      "epoch": 16.020890676829698,
      "grad_norm": 0.4792301654815674,
      "learning_rate": 6.739080276871667e-05,
      "loss": 1.8194,
      "step": 227000
    },
    {
      "epoch": 16.091467287740844,
      "grad_norm": 0.48394349217414856,
      "learning_rate": 6.61973904049646e-05,
      "loss": 1.7817,
      "step": 228000
    },
    {
      "epoch": 16.162043898651987,
      "grad_norm": 0.48044681549072266,
      "learning_rate": 6.50039780412125e-05,
      "loss": 1.785,
      "step": 229000
    },
    {
      "epoch": 16.23262050956313,
      "grad_norm": 0.46069419384002686,
      "learning_rate": 6.381175908982416e-05,
      "loss": 1.7904,
      "step": 230000
    },
    {
      "epoch": 16.303197120474273,
      "grad_norm": 0.4775301516056061,
      "learning_rate": 6.261834672607208e-05,
      "loss": 1.7848,
      "step": 231000
    },
    {
      "epoch": 16.37377373138542,
      "grad_norm": 0.51164311170578,
      "learning_rate": 6.142612777468374e-05,
      "loss": 1.7991,
      "step": 232000
    },
    {
      "epoch": 16.444350342296563,
      "grad_norm": 0.46276217699050903,
      "learning_rate": 6.023271541093165e-05,
      "loss": 1.7928,
      "step": 233000
    },
    {
      "epoch": 16.514926953207706,
      "grad_norm": 0.48281413316726685,
      "learning_rate": 5.904049645954331e-05,
      "loss": 1.8001,
      "step": 234000
    },
    {
      "epoch": 16.585503564118852,
      "grad_norm": 0.4853989779949188,
      "learning_rate": 5.784708409579123e-05,
      "loss": 1.801,
      "step": 235000
    },
    {
      "epoch": 16.656080175029995,
      "grad_norm": 0.448294073343277,
      "learning_rate": 5.665486514440289e-05,
      "loss": 1.8012,
      "step": 236000
    },
    {
      "epoch": 16.726656785941138,
      "grad_norm": 0.44111698865890503,
      "learning_rate": 5.546264619301455e-05,
      "loss": 1.8043,
      "step": 237000
    },
    {
      "epoch": 16.797233396852285,
      "grad_norm": 0.472347617149353,
      "learning_rate": 5.426923382926247e-05,
      "loss": 1.8108,
      "step": 238000
    },
    {
      "epoch": 16.867810007763428,
      "grad_norm": 0.5000669360160828,
      "learning_rate": 5.307582146551038e-05,
      "loss": 1.8092,
      "step": 239000
    },
    {
      "epoch": 16.93838661867457,
      "grad_norm": 0.46460166573524475,
      "learning_rate": 5.188240910175829e-05,
      "loss": 1.8045,
      "step": 240000
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.5315416594185826,
      "eval_loss": 2.4174108505249023,
      "eval_runtime": 125.0994,
      "eval_samples_per_second": 375.31,
      "eval_steps_per_second": 5.867,
      "step": 240873
    },
    {
      "epoch": 17.008963229585714,
      "grad_norm": 0.5037421584129333,
      "learning_rate": 5.0688996738006205e-05,
      "loss": 1.8006,
      "step": 241000
    },
    {
      "epoch": 17.07953984049686,
      "grad_norm": 0.450307160615921,
      "learning_rate": 4.949677778661787e-05,
      "loss": 1.7693,
      "step": 242000
    },
    {
      "epoch": 17.150116451408003,
      "grad_norm": 0.4222373962402344,
      "learning_rate": 4.830455883522953e-05,
      "loss": 1.7649,
      "step": 243000
    },
    {
      "epoch": 17.220693062319146,
      "grad_norm": 0.44716086983680725,
      "learning_rate": 4.711114647147744e-05,
      "loss": 1.7715,
      "step": 244000
    },
    {
      "epoch": 17.291269673230293,
      "grad_norm": 0.5217083096504211,
      "learning_rate": 4.591773410772535e-05,
      "loss": 1.7749,
      "step": 245000
    },
    {
      "epoch": 17.361846284141436,
      "grad_norm": 0.4576815068721771,
      "learning_rate": 4.472432174397327e-05,
      "loss": 1.7776,
      "step": 246000
    },
    {
      "epoch": 17.43242289505258,
      "grad_norm": 0.4578355848789215,
      "learning_rate": 4.353210279258493e-05,
      "loss": 1.7719,
      "step": 247000
    },
    {
      "epoch": 17.502999505963725,
      "grad_norm": 0.5052251219749451,
      "learning_rate": 4.233869042883284e-05,
      "loss": 1.7839,
      "step": 248000
    },
    {
      "epoch": 17.573576116874868,
      "grad_norm": 0.4527839720249176,
      "learning_rate": 4.1146471477444504e-05,
      "loss": 1.7797,
      "step": 249000
    },
    {
      "epoch": 17.64415272778601,
      "grad_norm": 0.5012311339378357,
      "learning_rate": 3.995305911369241e-05,
      "loss": 1.7834,
      "step": 250000
    },
    {
      "epoch": 17.714729338697154,
      "grad_norm": 0.4569910764694214,
      "learning_rate": 3.875964674994033e-05,
      "loss": 1.7897,
      "step": 251000
    },
    {
      "epoch": 17.7853059496083,
      "grad_norm": 0.4750159978866577,
      "learning_rate": 3.756742779855198e-05,
      "loss": 1.786,
      "step": 252000
    },
    {
      "epoch": 17.855882560519444,
      "grad_norm": 0.5255258083343506,
      "learning_rate": 3.63740154347999e-05,
      "loss": 1.7901,
      "step": 253000
    },
    {
      "epoch": 17.926459171430587,
      "grad_norm": 0.5186047554016113,
      "learning_rate": 3.518298989577532e-05,
      "loss": 1.7786,
      "step": 254000
    },
    {
      "epoch": 17.997035782341733,
      "grad_norm": 0.485440731048584,
      "learning_rate": 3.3989577532023225e-05,
      "loss": 1.7801,
      "step": 255000
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.5313575946841508,
      "eval_loss": 2.4237334728240967,
      "eval_runtime": 124.1981,
      "eval_samples_per_second": 378.033,
      "eval_steps_per_second": 5.91,
      "step": 255042
    },
    {
      "epoch": 18.067612393252876,
      "grad_norm": 0.486215204000473,
      "learning_rate": 3.279616516827114e-05,
      "loss": 1.7537,
      "step": 256000
    },
    {
      "epoch": 18.13818900416402,
      "grad_norm": 0.5382990837097168,
      "learning_rate": 3.160275280451905e-05,
      "loss": 1.754,
      "step": 257000
    },
    {
      "epoch": 18.208765615075166,
      "grad_norm": 0.48317405581474304,
      "learning_rate": 3.0410533853130714e-05,
      "loss": 1.7526,
      "step": 258000
    },
    {
      "epoch": 18.27934222598631,
      "grad_norm": 0.4930921494960785,
      "learning_rate": 2.9217121489378628e-05,
      "loss": 1.7619,
      "step": 259000
    },
    {
      "epoch": 18.34991883689745,
      "grad_norm": 0.5008710622787476,
      "learning_rate": 2.8023709125626538e-05,
      "loss": 1.7606,
      "step": 260000
    },
    {
      "epoch": 18.420495447808594,
      "grad_norm": 0.5138815641403198,
      "learning_rate": 2.6831490174238206e-05,
      "loss": 1.7632,
      "step": 261000
    },
    {
      "epoch": 18.49107205871974,
      "grad_norm": 0.4885752201080322,
      "learning_rate": 2.5638077810486113e-05,
      "loss": 1.7607,
      "step": 262000
    },
    {
      "epoch": 18.561648669630884,
      "grad_norm": 0.525771975517273,
      "learning_rate": 2.4444665446734027e-05,
      "loss": 1.7622,
      "step": 263000
    },
    {
      "epoch": 18.632225280542027,
      "grad_norm": 0.4830356240272522,
      "learning_rate": 2.3251253082981938e-05,
      "loss": 1.7667,
      "step": 264000
    },
    {
      "epoch": 18.702801891453174,
      "grad_norm": 0.41345518827438354,
      "learning_rate": 2.205784071922985e-05,
      "loss": 1.7629,
      "step": 265000
    },
    {
      "epoch": 18.773378502364316,
      "grad_norm": 0.5144524574279785,
      "learning_rate": 2.0866815180205263e-05,
      "loss": 1.762,
      "step": 266000
    },
    {
      "epoch": 18.84395511327546,
      "grad_norm": 0.4117266833782196,
      "learning_rate": 1.9673402816453177e-05,
      "loss": 1.7636,
      "step": 267000
    },
    {
      "epoch": 18.914531724186606,
      "grad_norm": 0.4791117310523987,
      "learning_rate": 1.847999045270109e-05,
      "loss": 1.7703,
      "step": 268000
    },
    {
      "epoch": 18.98510833509775,
      "grad_norm": 0.4984685778617859,
      "learning_rate": 1.7286578088949e-05,
      "loss": 1.7649,
      "step": 269000
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.5311354075823896,
      "eval_loss": 2.430837392807007,
      "eval_runtime": 124.7266,
      "eval_samples_per_second": 376.431,
      "eval_steps_per_second": 5.885,
      "step": 269211
    },
    {
      "epoch": 19.055684946008892,
      "grad_norm": 0.49468734860420227,
      "learning_rate": 1.6094359137560665e-05,
      "loss": 1.7503,
      "step": 270000
    },
    {
      "epoch": 19.126261556920035,
      "grad_norm": 0.5025900602340698,
      "learning_rate": 1.4902140186172328e-05,
      "loss": 1.7426,
      "step": 271000
    },
    {
      "epoch": 19.19683816783118,
      "grad_norm": 0.49257925152778625,
      "learning_rate": 1.3708727822420239e-05,
      "loss": 1.7478,
      "step": 272000
    },
    {
      "epoch": 19.267414778742324,
      "grad_norm": 0.5687286853790283,
      "learning_rate": 1.251531545866815e-05,
      "loss": 1.7415,
      "step": 273000
    },
    {
      "epoch": 19.337991389653467,
      "grad_norm": 0.48967909812927246,
      "learning_rate": 1.1321903094916061e-05,
      "loss": 1.7452,
      "step": 274000
    },
    {
      "epoch": 19.408568000564614,
      "grad_norm": 0.4711904227733612,
      "learning_rate": 1.0129684143527726e-05,
      "loss": 1.7565,
      "step": 275000
    },
    {
      "epoch": 19.479144611475757,
      "grad_norm": 0.5071347951889038,
      "learning_rate": 8.936271779775638e-06,
      "loss": 1.7424,
      "step": 276000
    },
    {
      "epoch": 19.5497212223869,
      "grad_norm": 0.494548499584198,
      "learning_rate": 7.744052828387302e-06,
      "loss": 1.7508,
      "step": 277000
    },
    {
      "epoch": 19.620297833298046,
      "grad_norm": 0.5207987427711487,
      "learning_rate": 6.5506404646352136e-06,
      "loss": 1.742,
      "step": 278000
    },
    {
      "epoch": 19.69087444420919,
      "grad_norm": 0.4780532121658325,
      "learning_rate": 5.358421513246877e-06,
      "loss": 1.7425,
      "step": 279000
    },
    {
      "epoch": 19.761451055120332,
      "grad_norm": 0.49671173095703125,
      "learning_rate": 4.165009149494788e-06,
      "loss": 1.7454,
      "step": 280000
    },
    {
      "epoch": 19.83202766603148,
      "grad_norm": 0.5458129644393921,
      "learning_rate": 2.972790198106452e-06,
      "loss": 1.749,
      "step": 281000
    },
    {
      "epoch": 19.902604276942622,
      "grad_norm": 0.5356537103652954,
      "learning_rate": 1.7793778343543638e-06,
      "loss": 1.744,
      "step": 282000
    },
    {
      "epoch": 19.973180887853765,
      "grad_norm": 0.5553900599479675,
      "learning_rate": 5.859654706022754e-07,
      "loss": 1.7481,
      "step": 283000
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.5309790230016276,
      "eval_loss": 2.4374380111694336,
      "eval_runtime": 124.6359,
      "eval_samples_per_second": 376.705,
      "eval_steps_per_second": 5.889,
      "step": 283380
    },
    {
      "epoch": 20.0,
      "step": 283380,
      "total_flos": 1.19390876011008e+18,
      "train_loss": 2.0466036078604977,
      "train_runtime": 62067.1587,
      "train_samples_per_second": 146.094,
      "train_steps_per_second": 4.566
    }
  ],
  "logging_steps": 1000,
  "max_steps": 283380,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "save_steps": 5000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.19390876011008e+18,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}