{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 1000,
  "global_step": 111303,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02695345138945042,
      "grad_norm": 1.7572588920593262,
      "learning_rate": 4.955077581017583e-05,
      "loss": 2.2431,
      "step": 1000
    },
    {
      "epoch": 0.05390690277890084,
      "grad_norm": 1.843064308166504,
      "learning_rate": 4.9101551620351654e-05,
      "loss": 1.8256,
      "step": 2000
    },
    {
      "epoch": 0.08086035416835126,
      "grad_norm": 1.7338684797286987,
      "learning_rate": 4.865232743052748e-05,
      "loss": 1.5467,
      "step": 3000
    },
    {
      "epoch": 0.10781380555780168,
      "grad_norm": 1.794568657875061,
      "learning_rate": 4.8203103240703305e-05,
      "loss": 1.3891,
      "step": 4000
    },
    {
      "epoch": 0.13476725694725208,
      "grad_norm": 1.6764180660247803,
      "learning_rate": 4.7753879050879134e-05,
      "loss": 1.3107,
      "step": 5000
    },
    {
      "epoch": 0.16172070833670252,
      "grad_norm": 1.5753543376922607,
      "learning_rate": 4.7304654861054956e-05,
      "loss": 1.2522,
      "step": 6000
    },
    {
      "epoch": 0.18867415972615292,
      "grad_norm": 1.5777322053909302,
      "learning_rate": 4.6855430671230785e-05,
      "loss": 1.2054,
      "step": 7000
    },
    {
      "epoch": 0.21562761111560336,
      "grad_norm": 1.6358872652053833,
      "learning_rate": 4.640620648140661e-05,
      "loss": 1.1639,
      "step": 8000
    },
    {
      "epoch": 0.24258106250505376,
      "grad_norm": 1.4646263122558594,
      "learning_rate": 4.595698229158244e-05,
      "loss": 1.1349,
      "step": 9000
    },
    {
      "epoch": 0.26953451389450417,
      "grad_norm": 1.8237578868865967,
      "learning_rate": 4.5507758101758266e-05,
      "loss": 1.1017,
      "step": 10000
    },
    {
      "epoch": 0.2964879652839546,
      "grad_norm": 4.1678619384765625,
      "learning_rate": 4.5058533911934095e-05,
      "loss": 1.0802,
      "step": 11000
    },
    {
      "epoch": 0.32344141667340504,
      "grad_norm": 1.6083228588104248,
      "learning_rate": 4.4609309722109924e-05,
      "loss": 1.0584,
      "step": 12000
    },
    {
      "epoch": 0.35039486806285547,
      "grad_norm": 1.5944287776947021,
      "learning_rate": 4.4160085532285746e-05,
      "loss": 1.0376,
      "step": 13000
    },
    {
      "epoch": 0.37734831945230585,
      "grad_norm": 1.5678300857543945,
      "learning_rate": 4.3710861342461575e-05,
      "loss": 1.0186,
      "step": 14000
    },
    {
      "epoch": 0.4043017708417563,
      "grad_norm": 1.5760828256607056,
      "learning_rate": 4.32616371526374e-05,
      "loss": 1.0006,
      "step": 15000
    },
    {
      "epoch": 0.4312552222312067,
      "grad_norm": 1.5422290563583374,
      "learning_rate": 4.281241296281323e-05,
      "loss": 0.9868,
      "step": 16000
    },
    {
      "epoch": 0.45820867362065715,
      "grad_norm": 1.8263678550720215,
      "learning_rate": 4.236318877298905e-05,
      "loss": 0.9737,
      "step": 17000
    },
    {
      "epoch": 0.4851621250101075,
      "grad_norm": 1.4930555820465088,
      "learning_rate": 4.191396458316488e-05,
      "loss": 0.9595,
      "step": 18000
    },
    {
      "epoch": 0.512115576399558,
      "grad_norm": 1.548412561416626,
      "learning_rate": 4.14647403933407e-05,
      "loss": 0.9448,
      "step": 19000
    },
    {
      "epoch": 0.5390690277890083,
      "grad_norm": 1.478004813194275,
      "learning_rate": 4.101551620351653e-05,
      "loss": 0.9351,
      "step": 20000
    },
    {
      "epoch": 0.5660224791784588,
      "grad_norm": 1.5493645668029785,
      "learning_rate": 4.056629201369235e-05,
      "loss": 0.9235,
      "step": 21000
    },
    {
      "epoch": 0.5929759305679092,
      "grad_norm": 1.5261894464492798,
      "learning_rate": 4.011706782386818e-05,
      "loss": 0.9133,
      "step": 22000
    },
    {
      "epoch": 0.6199293819573597,
      "grad_norm": 1.3742462396621704,
      "learning_rate": 3.9667843634044e-05,
      "loss": 0.9027,
      "step": 23000
    },
    {
      "epoch": 0.6468828333468101,
      "grad_norm": 1.451717495918274,
      "learning_rate": 3.921861944421983e-05,
      "loss": 0.8937,
      "step": 24000
    },
    {
      "epoch": 0.6738362847362604,
      "grad_norm": 1.5163180828094482,
      "learning_rate": 3.876939525439566e-05,
      "loss": 0.8837,
      "step": 25000
    },
    {
      "epoch": 0.7007897361257109,
      "grad_norm": 1.312432885169983,
      "learning_rate": 3.832017106457148e-05,
      "loss": 0.8735,
      "step": 26000
    },
    {
      "epoch": 0.7277431875151613,
      "grad_norm": 1.4663609266281128,
      "learning_rate": 3.787094687474731e-05,
      "loss": 0.8663,
      "step": 27000
    },
    {
      "epoch": 0.7546966389046117,
      "grad_norm": 1.457512617111206,
      "learning_rate": 3.742172268492314e-05,
      "loss": 0.8587,
      "step": 28000
    },
    {
      "epoch": 0.7816500902940622,
      "grad_norm": 1.4610168933868408,
      "learning_rate": 3.697249849509897e-05,
      "loss": 0.8514,
      "step": 29000
    },
    {
      "epoch": 0.8086035416835126,
      "grad_norm": 1.6427557468414307,
      "learning_rate": 3.652327430527479e-05,
      "loss": 0.8457,
      "step": 30000
    },
    {
      "epoch": 0.8355569930729629,
      "grad_norm": 1.3497254848480225,
      "learning_rate": 3.607405011545062e-05,
      "loss": 0.8372,
      "step": 31000
    },
    {
      "epoch": 0.8625104444624134,
      "grad_norm": 1.597374677658081,
      "learning_rate": 3.5624825925626444e-05,
      "loss": 0.8308,
      "step": 32000
    },
    {
      "epoch": 0.8894638958518638,
      "grad_norm": 1.5185362100601196,
      "learning_rate": 3.517560173580227e-05,
      "loss": 0.8262,
      "step": 33000
    },
    {
      "epoch": 0.9164173472413143,
      "grad_norm": 1.452099323272705,
      "learning_rate": 3.4726377545978095e-05,
      "loss": 0.8188,
      "step": 34000
    },
    {
      "epoch": 0.9433707986307647,
      "grad_norm": 1.4626882076263428,
      "learning_rate": 3.4277153356153925e-05,
      "loss": 0.8128,
      "step": 35000
    },
    {
      "epoch": 0.970324250020215,
      "grad_norm": 1.329575538635254,
      "learning_rate": 3.3827929166329754e-05,
      "loss": 0.8057,
      "step": 36000
    },
    {
      "epoch": 0.9972777014096655,
      "grad_norm": 1.3917378187179565,
      "learning_rate": 3.3378704976505576e-05,
      "loss": 0.8014,
      "step": 37000
    },
    {
      "epoch": 1.024231152799116,
      "grad_norm": 1.4865970611572266,
      "learning_rate": 3.2929480786681405e-05,
      "loss": 0.7947,
      "step": 38000
    },
    {
      "epoch": 1.0511846041885664,
      "grad_norm": 1.4632256031036377,
      "learning_rate": 3.248025659685723e-05,
      "loss": 0.7898,
      "step": 39000
    },
    {
      "epoch": 1.0781380555780167,
      "grad_norm": 1.5164929628372192,
      "learning_rate": 3.2031032407033056e-05,
      "loss": 0.7843,
      "step": 40000
    },
    {
      "epoch": 1.1050915069674672,
      "grad_norm": 1.4825857877731323,
      "learning_rate": 3.158180821720888e-05,
      "loss": 0.7796,
      "step": 41000
    },
    {
      "epoch": 1.1320449583569177,
      "grad_norm": 1.4526509046554565,
      "learning_rate": 3.113258402738471e-05,
      "loss": 0.7771,
      "step": 42000
    },
    {
      "epoch": 1.158998409746368,
      "grad_norm": 1.4606186151504517,
      "learning_rate": 3.068335983756053e-05,
      "loss": 0.7728,
      "step": 43000
    },
    {
      "epoch": 1.1859518611358184,
      "grad_norm": 3.3901302814483643,
      "learning_rate": 3.0234135647736362e-05,
      "loss": 0.7664,
      "step": 44000
    },
    {
      "epoch": 1.212905312525269,
      "grad_norm": 1.4377284049987793,
      "learning_rate": 2.9784911457912185e-05,
      "loss": 0.7656,
      "step": 45000
    },
    {
      "epoch": 1.2398587639147194,
      "grad_norm": 1.432142734527588,
      "learning_rate": 2.9335687268088014e-05,
      "loss": 0.7592,
      "step": 46000
    },
    {
      "epoch": 1.2668122153041697,
      "grad_norm": 1.3727272748947144,
      "learning_rate": 2.8886463078263843e-05,
      "loss": 0.7563,
      "step": 47000
    },
    {
      "epoch": 1.2937656666936201,
      "grad_norm": 1.5601295232772827,
      "learning_rate": 2.8437238888439665e-05,
      "loss": 0.7509,
      "step": 48000
    },
    {
      "epoch": 1.3207191180830704,
      "grad_norm": 1.4392520189285278,
      "learning_rate": 2.7988014698615494e-05,
      "loss": 0.7461,
      "step": 49000
    },
    {
      "epoch": 1.347672569472521,
      "grad_norm": 1.365379810333252,
      "learning_rate": 2.7538790508791316e-05,
      "loss": 0.7416,
      "step": 50000
    },
    {
      "epoch": 1.3746260208619714,
      "grad_norm": 1.352710247039795,
      "learning_rate": 2.7089566318967145e-05,
      "loss": 0.7403,
      "step": 51000
    },
    {
      "epoch": 1.4015794722514219,
      "grad_norm": 1.3562721014022827,
      "learning_rate": 2.664034212914297e-05,
      "loss": 0.7359,
      "step": 52000
    },
    {
      "epoch": 1.4285329236408721,
      "grad_norm": 1.627213954925537,
      "learning_rate": 2.61911179393188e-05,
      "loss": 0.7329,
      "step": 53000
    },
    {
      "epoch": 1.4554863750303226,
      "grad_norm": 1.4729622602462769,
      "learning_rate": 2.5741893749494622e-05,
      "loss": 0.7302,
      "step": 54000
    },
    {
      "epoch": 1.4824398264197731,
      "grad_norm": 1.4408637285232544,
      "learning_rate": 2.529266955967045e-05,
      "loss": 0.7283,
      "step": 55000
    },
    {
      "epoch": 1.5093932778092234,
      "grad_norm": 1.504920482635498,
      "learning_rate": 2.4843445369846277e-05,
      "loss": 0.7233,
      "step": 56000
    },
    {
      "epoch": 1.5363467291986739,
      "grad_norm": 1.4795109033584595,
      "learning_rate": 2.4394221180022103e-05,
      "loss": 0.7211,
      "step": 57000
    },
    {
      "epoch": 1.5633001805881244,
      "grad_norm": 1.4444007873535156,
      "learning_rate": 2.394499699019793e-05,
      "loss": 0.7162,
      "step": 58000
    },
    {
      "epoch": 1.5902536319775749,
      "grad_norm": 1.4556983709335327,
      "learning_rate": 2.3495772800373754e-05,
      "loss": 0.7156,
      "step": 59000
    },
    {
      "epoch": 1.6172070833670251,
      "grad_norm": 1.5000152587890625,
      "learning_rate": 2.304654861054958e-05,
      "loss": 0.7118,
      "step": 60000
    },
    {
      "epoch": 1.6441605347564756,
      "grad_norm": 1.385746717453003,
      "learning_rate": 2.259732442072541e-05,
      "loss": 0.7091,
      "step": 61000
    },
    {
      "epoch": 1.6711139861459259,
      "grad_norm": 1.5274490118026733,
      "learning_rate": 2.2148100230901235e-05,
      "loss": 0.7073,
      "step": 62000
    },
    {
      "epoch": 1.6980674375353764,
      "grad_norm": 1.411332607269287,
      "learning_rate": 2.169887604107706e-05,
      "loss": 0.703,
      "step": 63000
    },
    {
      "epoch": 1.7250208889248269,
      "grad_norm": 1.6316828727722168,
      "learning_rate": 2.124965185125289e-05,
      "loss": 0.702,
      "step": 64000
    },
    {
      "epoch": 1.7519743403142773,
      "grad_norm": 1.4710853099822998,
      "learning_rate": 2.0800427661428715e-05,
      "loss": 0.6995,
      "step": 65000
    },
    {
      "epoch": 1.7789277917037276,
      "grad_norm": 1.984465479850769,
      "learning_rate": 2.035120347160454e-05,
      "loss": 0.6976,
      "step": 66000
    },
    {
      "epoch": 1.805881243093178,
      "grad_norm": 1.5006201267242432,
      "learning_rate": 1.9901979281780366e-05,
      "loss": 0.6917,
      "step": 67000
    },
    {
      "epoch": 1.8328346944826284,
      "grad_norm": 1.4539721012115479,
      "learning_rate": 1.9452755091956192e-05,
      "loss": 0.6906,
      "step": 68000
    },
    {
      "epoch": 1.8597881458720789,
      "grad_norm": 1.4637404680252075,
      "learning_rate": 1.9003530902132018e-05,
      "loss": 0.6866,
      "step": 69000
    },
    {
      "epoch": 1.8867415972615293,
      "grad_norm": 1.4094914197921753,
      "learning_rate": 1.8554306712307847e-05,
      "loss": 0.6856,
      "step": 70000
    },
    {
      "epoch": 1.9136950486509798,
      "grad_norm": 1.828584909439087,
      "learning_rate": 1.8105082522483672e-05,
      "loss": 0.6838,
      "step": 71000
    },
    {
      "epoch": 1.9406485000404303,
      "grad_norm": 1.3690617084503174,
      "learning_rate": 1.7655858332659498e-05,
      "loss": 0.6815,
      "step": 72000
    },
    {
      "epoch": 1.9676019514298806,
      "grad_norm": 1.438297986984253,
      "learning_rate": 1.7206634142835324e-05,
      "loss": 0.678,
      "step": 73000
    },
    {
      "epoch": 1.9945554028193309,
      "grad_norm": 1.482967495918274,
      "learning_rate": 1.675740995301115e-05,
      "loss": 0.6761,
      "step": 74000
    },
    {
      "epoch": 2.0215088542087813,
      "grad_norm": 1.5024503469467163,
      "learning_rate": 1.6308185763186975e-05,
      "loss": 0.6719,
      "step": 75000
    },
    {
      "epoch": 2.048462305598232,
      "grad_norm": 1.444847583770752,
      "learning_rate": 1.5858961573362804e-05,
      "loss": 0.6715,
      "step": 76000
    },
    {
      "epoch": 2.0754157569876823,
      "grad_norm": 1.4710804224014282,
      "learning_rate": 1.540973738353863e-05,
      "loss": 0.6697,
      "step": 77000
    },
    {
      "epoch": 2.102369208377133,
      "grad_norm": 1.4638535976409912,
      "learning_rate": 1.4960513193714457e-05,
      "loss": 0.6676,
      "step": 78000
    },
    {
      "epoch": 2.1293226597665833,
      "grad_norm": 1.4491604566574097,
      "learning_rate": 1.4511289003890283e-05,
      "loss": 0.666,
      "step": 79000
    },
    {
      "epoch": 2.1562761111560333,
      "grad_norm": 1.5296227931976318,
      "learning_rate": 1.4062064814066108e-05,
      "loss": 0.6638,
      "step": 80000
    },
    {
      "epoch": 2.183229562545484,
      "grad_norm": 1.4576656818389893,
      "learning_rate": 1.3612840624241934e-05,
      "loss": 0.6603,
      "step": 81000
    },
    {
      "epoch": 2.2101830139349343,
      "grad_norm": 1.3864960670471191,
      "learning_rate": 1.3163616434417761e-05,
      "loss": 0.6593,
      "step": 82000
    },
    {
      "epoch": 2.237136465324385,
      "grad_norm": 1.7719892263412476,
      "learning_rate": 1.2714392244593587e-05,
      "loss": 0.6602,
      "step": 83000
    },
    {
      "epoch": 2.2640899167138353,
      "grad_norm": 1.3673619031906128,
      "learning_rate": 1.2265168054769415e-05,
      "loss": 0.656,
      "step": 84000
    },
    {
      "epoch": 2.291043368103286,
      "grad_norm": 1.4584944248199463,
      "learning_rate": 1.181594386494524e-05,
      "loss": 0.6536,
      "step": 85000
    },
    {
      "epoch": 2.317996819492736,
      "grad_norm": 1.3770402669906616,
      "learning_rate": 1.1366719675121066e-05,
      "loss": 0.6519,
      "step": 86000
    },
    {
      "epoch": 2.3449502708821863,
      "grad_norm": 1.3435842990875244,
      "learning_rate": 1.0917495485296893e-05,
      "loss": 0.6498,
      "step": 87000
    },
    {
      "epoch": 2.371903722271637,
      "grad_norm": 1.4926843643188477,
      "learning_rate": 1.0468271295472719e-05,
      "loss": 0.6497,
      "step": 88000
    },
    {
      "epoch": 2.3988571736610873,
      "grad_norm": 1.4428088665008545,
      "learning_rate": 1.0019047105648545e-05,
      "loss": 0.648,
      "step": 89000
    },
    {
      "epoch": 2.425810625050538,
      "grad_norm": 1.3555138111114502,
      "learning_rate": 9.569822915824372e-06,
      "loss": 0.6463,
      "step": 90000
    },
    {
      "epoch": 2.4527640764399883,
      "grad_norm": 1.8231887817382812,
      "learning_rate": 9.1205987260002e-06,
      "loss": 0.6441,
      "step": 91000
    },
    {
      "epoch": 2.4797175278294388,
      "grad_norm": 1.4294542074203491,
      "learning_rate": 8.671374536176025e-06,
      "loss": 0.6443,
      "step": 92000
    },
    {
      "epoch": 2.506670979218889,
      "grad_norm": 1.4158700704574585,
      "learning_rate": 8.22215034635185e-06,
      "loss": 0.6408,
      "step": 93000
    },
    {
      "epoch": 2.5336244306083393,
      "grad_norm": 1.3847190141677856,
      "learning_rate": 7.772926156527676e-06,
      "loss": 0.6407,
      "step": 94000
    },
    {
      "epoch": 2.56057788199779,
      "grad_norm": 1.4408886432647705,
      "learning_rate": 7.323701966703503e-06,
      "loss": 0.6408,
      "step": 95000
    },
    {
      "epoch": 2.5875313333872403,
      "grad_norm": 1.5943797826766968,
      "learning_rate": 6.87447777687933e-06,
      "loss": 0.6379,
      "step": 96000
    },
    {
      "epoch": 2.6144847847766908,
      "grad_norm": 1.4184848070144653,
      "learning_rate": 6.425253587055157e-06,
      "loss": 0.6367,
      "step": 97000
    },
    {
      "epoch": 2.641438236166141,
      "grad_norm": 1.4443740844726562,
      "learning_rate": 5.976029397230982e-06,
      "loss": 0.635,
      "step": 98000
    },
    {
      "epoch": 2.6683916875555918,
      "grad_norm": 1.5342421531677246,
      "learning_rate": 5.526805207406809e-06,
      "loss": 0.6349,
      "step": 99000
    },
    {
      "epoch": 2.695345138945042,
      "grad_norm": 1.4228054285049438,
      "learning_rate": 5.077581017582635e-06,
      "loss": 0.6349,
      "step": 100000
    },
    {
      "epoch": 2.7222985903344923,
      "grad_norm": 1.5616410970687866,
      "learning_rate": 4.628356827758461e-06,
      "loss": 0.6345,
      "step": 101000
    },
    {
      "epoch": 2.7492520417239428,
      "grad_norm": 1.4892140626907349,
      "learning_rate": 4.1791326379342876e-06,
      "loss": 0.6322,
      "step": 102000
    },
    {
      "epoch": 2.7762054931133933,
      "grad_norm": 1.48268723487854,
      "learning_rate": 3.7299084481101145e-06,
      "loss": 0.6318,
      "step": 103000
    },
    {
      "epoch": 2.8031589445028438,
      "grad_norm": 1.6489795446395874,
      "learning_rate": 3.2806842582859406e-06,
      "loss": 0.6295,
      "step": 104000
    },
    {
      "epoch": 2.830112395892294,
      "grad_norm": 1.458978295326233,
      "learning_rate": 2.8314600684617667e-06,
      "loss": 0.6291,
      "step": 105000
    },
    {
      "epoch": 2.8570658472817443,
      "grad_norm": 1.338670015335083,
      "learning_rate": 2.3822358786375928e-06,
      "loss": 0.6265,
      "step": 106000
    },
    {
      "epoch": 2.8840192986711948,
      "grad_norm": 1.3654705286026,
      "learning_rate": 1.9330116888134193e-06,
      "loss": 0.6259,
      "step": 107000
    },
    {
      "epoch": 2.9109727500606453,
      "grad_norm": 1.3295788764953613,
      "learning_rate": 1.4837874989892456e-06,
      "loss": 0.6274,
      "step": 108000
    },
    {
      "epoch": 2.9379262014500958,
      "grad_norm": 1.3103362321853638,
      "learning_rate": 1.0345633091650721e-06,
      "loss": 0.6261,
      "step": 109000
    },
    {
      "epoch": 2.9648796528395462,
      "grad_norm": 1.3937286138534546,
      "learning_rate": 5.853391193408983e-07,
      "loss": 0.6253,
      "step": 110000
    },
    {
      "epoch": 2.9918331042289967,
      "grad_norm": 1.3265410661697388,
      "learning_rate": 1.3611492951672462e-07,
      "loss": 0.6257,
      "step": 111000
    },
    {
      "epoch": 3.0,
      "step": 111303,
      "total_flos": 2.6556045146219667e+19,
      "train_loss": 0.8057731239829722,
      "train_runtime": 323766.3615,
      "train_samples_per_second": 88.005,
      "train_steps_per_second": 0.344
    }
  ],
  "logging_steps": 1000,
  "max_steps": 111303,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.6556045146219667e+19,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}