{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.8764800560947236,
  "eval_steps": 500,
  "global_step": 55000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.007968000509952033,
      "grad_norm": 118280.5625,
      "learning_rate": 2.9936381561814055e-05,
      "loss": 1.2421,
      "step": 500
    },
    {
      "epoch": 0.015936001019904066,
      "grad_norm": 93972.8125,
      "learning_rate": 2.9856659208197584e-05,
      "loss": 0.7745,
      "step": 1000
    },
    {
      "epoch": 0.023904001529856098,
      "grad_norm": 78047.78125,
      "learning_rate": 2.9776936854581113e-05,
      "loss": 0.73,
      "step": 1500
    },
    {
      "epoch": 0.03187200203980813,
      "grad_norm": 72831.09375,
      "learning_rate": 2.9697214500964642e-05,
      "loss": 0.7087,
      "step": 2000
    },
    {
      "epoch": 0.03984000254976016,
      "grad_norm": 62310.4453125,
      "learning_rate": 2.961749214734817e-05,
      "loss": 0.6964,
      "step": 2500
    },
    {
      "epoch": 0.047808003059712195,
      "grad_norm": 65517.09375,
      "learning_rate": 2.9537769793731697e-05,
      "loss": 0.6858,
      "step": 3000
    },
    {
      "epoch": 0.05577600356966423,
      "grad_norm": 57505.90625,
      "learning_rate": 2.9458047440115226e-05,
      "loss": 0.6761,
      "step": 3500
    },
    {
      "epoch": 0.06374400407961626,
      "grad_norm": 55419.26171875,
      "learning_rate": 2.9378325086498752e-05,
      "loss": 0.6679,
      "step": 4000
    },
    {
      "epoch": 0.07171200458956829,
      "grad_norm": 51254.3671875,
      "learning_rate": 2.929860273288228e-05,
      "loss": 0.6602,
      "step": 4500
    },
    {
      "epoch": 0.07968000509952032,
      "grad_norm": 51469.9765625,
      "learning_rate": 2.9218880379265813e-05,
      "loss": 0.6554,
      "step": 5000
    },
    {
      "epoch": 0.08764800560947236,
      "grad_norm": 69357.2265625,
      "learning_rate": 2.913915802564934e-05,
      "loss": 0.6516,
      "step": 5500
    },
    {
      "epoch": 0.09561600611942439,
      "grad_norm": 48255.7890625,
      "learning_rate": 2.9059435672032868e-05,
      "loss": 0.6482,
      "step": 6000
    },
    {
      "epoch": 0.10358400662937642,
      "grad_norm": 59810.20703125,
      "learning_rate": 2.8979713318416394e-05,
      "loss": 0.6397,
      "step": 6500
    },
    {
      "epoch": 0.11155200713932846,
      "grad_norm": 54012.7109375,
      "learning_rate": 2.8899990964799923e-05,
      "loss": 0.6425,
      "step": 7000
    },
    {
      "epoch": 0.11952000764928049,
      "grad_norm": 47228.74609375,
      "learning_rate": 2.8820268611183452e-05,
      "loss": 0.6392,
      "step": 7500
    },
    {
      "epoch": 0.12748800815923253,
      "grad_norm": 55561.9453125,
      "learning_rate": 2.874054625756698e-05,
      "loss": 0.6342,
      "step": 8000
    },
    {
      "epoch": 0.13545600866918456,
      "grad_norm": 50830.77734375,
      "learning_rate": 2.866082390395051e-05,
      "loss": 0.6288,
      "step": 8500
    },
    {
      "epoch": 0.14342400917913659,
      "grad_norm": 45037.15234375,
      "learning_rate": 2.858110155033404e-05,
      "loss": 0.6283,
      "step": 9000
    },
    {
      "epoch": 0.1513920096890886,
      "grad_norm": 42729.57421875,
      "learning_rate": 2.8501379196717565e-05,
      "loss": 0.6252,
      "step": 9500
    },
    {
      "epoch": 0.15936001019904064,
      "grad_norm": 46592.18359375,
      "learning_rate": 2.8421656843101094e-05,
      "loss": 0.6199,
      "step": 10000
    },
    {
      "epoch": 0.1673280107089927,
      "grad_norm": 40794.70703125,
      "learning_rate": 2.834193448948462e-05,
      "loss": 0.6237,
      "step": 10500
    },
    {
      "epoch": 0.17529601121894473,
      "grad_norm": 40227.96484375,
      "learning_rate": 2.8262212135868152e-05,
      "loss": 0.6174,
      "step": 11000
    },
    {
      "epoch": 0.18326401172889675,
      "grad_norm": 49994.6015625,
      "learning_rate": 2.818248978225168e-05,
      "loss": 0.6151,
      "step": 11500
    },
    {
      "epoch": 0.19123201223884878,
      "grad_norm": 55100.75390625,
      "learning_rate": 2.8102767428635207e-05,
      "loss": 0.6117,
      "step": 12000
    },
    {
      "epoch": 0.1992000127488008,
      "grad_norm": 93011.546875,
      "learning_rate": 2.8023045075018736e-05,
      "loss": 0.6122,
      "step": 12500
    },
    {
      "epoch": 0.20716801325875284,
      "grad_norm": 47631.09375,
      "learning_rate": 2.794332272140226e-05,
      "loss": 0.6133,
      "step": 13000
    },
    {
      "epoch": 0.2151360137687049,
      "grad_norm": 42235.15625,
      "learning_rate": 2.786360036778579e-05,
      "loss": 0.6038,
      "step": 13500
    },
    {
      "epoch": 0.22310401427865692,
      "grad_norm": 41144.3359375,
      "learning_rate": 2.7783878014169323e-05,
      "loss": 0.6074,
      "step": 14000
    },
    {
      "epoch": 0.23107201478860895,
      "grad_norm": 47137.19921875,
      "learning_rate": 2.770415566055285e-05,
      "loss": 0.6047,
      "step": 14500
    },
    {
      "epoch": 0.23904001529856098,
      "grad_norm": 37974.65625,
      "learning_rate": 2.7624433306936378e-05,
      "loss": 0.6058,
      "step": 15000
    },
    {
      "epoch": 0.247008015808513,
      "grad_norm": 49424.44140625,
      "learning_rate": 2.7544710953319903e-05,
      "loss": 0.6014,
      "step": 15500
    },
    {
      "epoch": 0.25497601631846506,
      "grad_norm": 57363.96875,
      "learning_rate": 2.7464988599703432e-05,
      "loss": 0.6012,
      "step": 16000
    },
    {
      "epoch": 0.26294401682841706,
      "grad_norm": 48850.90234375,
      "learning_rate": 2.738526624608696e-05,
      "loss": 0.5969,
      "step": 16500
    },
    {
      "epoch": 0.2709120173383691,
      "grad_norm": 60200.43359375,
      "learning_rate": 2.730554389247049e-05,
      "loss": 0.5958,
      "step": 17000
    },
    {
      "epoch": 0.2788800178483211,
      "grad_norm": 48039.16015625,
      "learning_rate": 2.722582153885402e-05,
      "loss": 0.598,
      "step": 17500
    },
    {
      "epoch": 0.28684801835827317,
      "grad_norm": 34431.2265625,
      "learning_rate": 2.7146099185237545e-05,
      "loss": 0.5953,
      "step": 18000
    },
    {
      "epoch": 0.2948160188682252,
      "grad_norm": 34913.16015625,
      "learning_rate": 2.7066376831621074e-05,
      "loss": 0.5932,
      "step": 18500
    },
    {
      "epoch": 0.3027840193781772,
      "grad_norm": 36192.234375,
      "learning_rate": 2.6986654478004603e-05,
      "loss": 0.5939,
      "step": 19000
    },
    {
      "epoch": 0.3107520198881293,
      "grad_norm": 46195.62890625,
      "learning_rate": 2.690693212438813e-05,
      "loss": 0.5914,
      "step": 19500
    },
    {
      "epoch": 0.3187200203980813,
      "grad_norm": 43815.43359375,
      "learning_rate": 2.682720977077166e-05,
      "loss": 0.5904,
      "step": 20000
    },
    {
      "epoch": 0.32668802090803334,
      "grad_norm": 35572.34375,
      "learning_rate": 2.674748741715519e-05,
      "loss": 0.5865,
      "step": 20500
    },
    {
      "epoch": 0.3346560214179854,
      "grad_norm": 36805.02734375,
      "learning_rate": 2.6667765063538716e-05,
      "loss": 0.5904,
      "step": 21000
    },
    {
      "epoch": 0.3426240219279374,
      "grad_norm": 45271.30078125,
      "learning_rate": 2.6588042709922245e-05,
      "loss": 0.5873,
      "step": 21500
    },
    {
      "epoch": 0.35059202243788945,
      "grad_norm": 35245.0234375,
      "learning_rate": 2.650832035630577e-05,
      "loss": 0.5856,
      "step": 22000
    },
    {
      "epoch": 0.35856002294784145,
      "grad_norm": 34578.25,
      "learning_rate": 2.64285980026893e-05,
      "loss": 0.5795,
      "step": 22500
    },
    {
      "epoch": 0.3665280234577935,
      "grad_norm": 40504.375,
      "learning_rate": 2.6348875649072832e-05,
      "loss": 0.5846,
      "step": 23000
    },
    {
      "epoch": 0.37449602396774556,
      "grad_norm": 37683.4453125,
      "learning_rate": 2.6269153295456358e-05,
      "loss": 0.5792,
      "step": 23500
    },
    {
      "epoch": 0.38246402447769756,
      "grad_norm": 35323.41015625,
      "learning_rate": 2.6189430941839887e-05,
      "loss": 0.5803,
      "step": 24000
    },
    {
      "epoch": 0.3904320249876496,
      "grad_norm": 47546.1953125,
      "learning_rate": 2.6109708588223413e-05,
      "loss": 0.5805,
      "step": 24500
    },
    {
      "epoch": 0.3984000254976016,
      "grad_norm": 44604.859375,
      "learning_rate": 2.6029986234606942e-05,
      "loss": 0.5773,
      "step": 25000
    },
    {
      "epoch": 0.4063680260075537,
      "grad_norm": 47108.78125,
      "learning_rate": 2.595026388099047e-05,
      "loss": 0.5825,
      "step": 25500
    },
    {
      "epoch": 0.4143360265175057,
      "grad_norm": 42788.87109375,
      "learning_rate": 2.5870541527374e-05,
      "loss": 0.578,
      "step": 26000
    },
    {
      "epoch": 0.42230402702745773,
      "grad_norm": 33550.89453125,
      "learning_rate": 2.579081917375753e-05,
      "loss": 0.5779,
      "step": 26500
    },
    {
      "epoch": 0.4302720275374098,
      "grad_norm": 36036.8984375,
      "learning_rate": 2.5711096820141055e-05,
      "loss": 0.574,
      "step": 27000
    },
    {
      "epoch": 0.4382400280473618,
      "grad_norm": 61990.4296875,
      "learning_rate": 2.5631374466524584e-05,
      "loss": 0.5751,
      "step": 27500
    },
    {
      "epoch": 0.44620802855731384,
      "grad_norm": 35250.4375,
      "learning_rate": 2.5551652112908113e-05,
      "loss": 0.5761,
      "step": 28000
    },
    {
      "epoch": 0.45417602906726584,
      "grad_norm": 40302.27734375,
      "learning_rate": 2.547192975929164e-05,
      "loss": 0.5723,
      "step": 28500
    },
    {
      "epoch": 0.4621440295772179,
      "grad_norm": 40248.03515625,
      "learning_rate": 2.539220740567517e-05,
      "loss": 0.5754,
      "step": 29000
    },
    {
      "epoch": 0.47011203008716995,
      "grad_norm": 38083.6484375,
      "learning_rate": 2.5312485052058697e-05,
      "loss": 0.5719,
      "step": 29500
    },
    {
      "epoch": 0.47808003059712195,
      "grad_norm": 32291.21484375,
      "learning_rate": 2.5232762698442226e-05,
      "loss": 0.5726,
      "step": 30000
    },
    {
      "epoch": 0.486048031107074,
      "grad_norm": 39065.84375,
      "learning_rate": 2.5153040344825755e-05,
      "loss": 0.5716,
      "step": 30500
    },
    {
      "epoch": 0.494016031617026,
      "grad_norm": 46160.44921875,
      "learning_rate": 2.507331799120928e-05,
      "loss": 0.5723,
      "step": 31000
    },
    {
      "epoch": 0.5019840321269781,
      "grad_norm": 36111.65625,
      "learning_rate": 2.499359563759281e-05,
      "loss": 0.569,
      "step": 31500
    },
    {
      "epoch": 0.5099520326369301,
      "grad_norm": 33012.55859375,
      "learning_rate": 2.4913873283976342e-05,
      "loss": 0.5692,
      "step": 32000
    },
    {
      "epoch": 0.5179200331468822,
      "grad_norm": 40300.83203125,
      "learning_rate": 2.4834150930359868e-05,
      "loss": 0.5686,
      "step": 32500
    },
    {
      "epoch": 0.5258880336568341,
      "grad_norm": 38612.78515625,
      "learning_rate": 2.4754428576743397e-05,
      "loss": 0.5663,
      "step": 33000
    },
    {
      "epoch": 0.5338560341667862,
      "grad_norm": 52813.0234375,
      "learning_rate": 2.4674706223126922e-05,
      "loss": 0.5662,
      "step": 33500
    },
    {
      "epoch": 0.5418240346767382,
      "grad_norm": 36266.23046875,
      "learning_rate": 2.459498386951045e-05,
      "loss": 0.5635,
      "step": 34000
    },
    {
      "epoch": 0.5497920351866903,
      "grad_norm": 38791.453125,
      "learning_rate": 2.451526151589398e-05,
      "loss": 0.5677,
      "step": 34500
    },
    {
      "epoch": 0.5577600356966422,
      "grad_norm": 43926.4296875,
      "learning_rate": 2.443553916227751e-05,
      "loss": 0.5688,
      "step": 35000
    },
    {
      "epoch": 0.5657280362065943,
      "grad_norm": 32247.173828125,
      "learning_rate": 2.435581680866104e-05,
      "loss": 0.5686,
      "step": 35500
    },
    {
      "epoch": 0.5736960367165463,
      "grad_norm": 35230.44140625,
      "learning_rate": 2.4276094455044564e-05,
      "loss": 0.566,
      "step": 36000
    },
    {
      "epoch": 0.5816640372264984,
      "grad_norm": 33197.58203125,
      "learning_rate": 2.4196372101428093e-05,
      "loss": 0.5629,
      "step": 36500
    },
    {
      "epoch": 0.5896320377364505,
      "grad_norm": 41267.5390625,
      "learning_rate": 2.4116649747811623e-05,
      "loss": 0.5642,
      "step": 37000
    },
    {
      "epoch": 0.5976000382464024,
      "grad_norm": 41852.71875,
      "learning_rate": 2.4036927394195148e-05,
      "loss": 0.5635,
      "step": 37500
    },
    {
      "epoch": 0.6055680387563545,
      "grad_norm": 32820.1015625,
      "learning_rate": 2.395720504057868e-05,
      "loss": 0.5639,
      "step": 38000
    },
    {
      "epoch": 0.6135360392663065,
      "grad_norm": 39303.70703125,
      "learning_rate": 2.3877482686962206e-05,
      "loss": 0.5605,
      "step": 38500
    },
    {
      "epoch": 0.6215040397762586,
      "grad_norm": 53008.375,
      "learning_rate": 2.3797760333345735e-05,
      "loss": 0.562,
      "step": 39000
    },
    {
      "epoch": 0.6294720402862106,
      "grad_norm": 41845.28515625,
      "learning_rate": 2.3718037979729264e-05,
      "loss": 0.5641,
      "step": 39500
    },
    {
      "epoch": 0.6374400407961626,
      "grad_norm": 49620.7734375,
      "learning_rate": 2.363831562611279e-05,
      "loss": 0.5582,
      "step": 40000
    },
    {
      "epoch": 0.6454080413061146,
      "grad_norm": 46115.32421875,
      "learning_rate": 2.355859327249632e-05,
      "loss": 0.5616,
      "step": 40500
    },
    {
      "epoch": 0.6533760418160667,
      "grad_norm": 33357.73828125,
      "learning_rate": 2.3478870918879848e-05,
      "loss": 0.5633,
      "step": 41000
    },
    {
      "epoch": 0.6613440423260187,
      "grad_norm": 48345.0859375,
      "learning_rate": 2.3399148565263377e-05,
      "loss": 0.5588,
      "step": 41500
    },
    {
      "epoch": 0.6693120428359708,
      "grad_norm": 46256.62890625,
      "learning_rate": 2.3319426211646906e-05,
      "loss": 0.5584,
      "step": 42000
    },
    {
      "epoch": 0.6772800433459227,
      "grad_norm": 34897.8828125,
      "learning_rate": 2.3239703858030432e-05,
      "loss": 0.5577,
      "step": 42500
    },
    {
      "epoch": 0.6852480438558748,
      "grad_norm": 86223.203125,
      "learning_rate": 2.315998150441396e-05,
      "loss": 0.5549,
      "step": 43000
    },
    {
      "epoch": 0.6932160443658268,
      "grad_norm": 32481.861328125,
      "learning_rate": 2.308025915079749e-05,
      "loss": 0.5623,
      "step": 43500
    },
    {
      "epoch": 0.7011840448757789,
      "grad_norm": 33890.0859375,
      "learning_rate": 2.300053679718102e-05,
      "loss": 0.5618,
      "step": 44000
    },
    {
      "epoch": 0.709152045385731,
      "grad_norm": 32936.66015625,
      "learning_rate": 2.2920814443564548e-05,
      "loss": 0.5613,
      "step": 44500
    },
    {
      "epoch": 0.7171200458956829,
      "grad_norm": 36461.4453125,
      "learning_rate": 2.2841092089948074e-05,
      "loss": 0.5581,
      "step": 45000
    },
    {
      "epoch": 0.725088046405635,
      "grad_norm": 35596.3046875,
      "learning_rate": 2.2761369736331603e-05,
      "loss": 0.5514,
      "step": 45500
    },
    {
      "epoch": 0.733056046915587,
      "grad_norm": 46988.7890625,
      "learning_rate": 2.2681647382715132e-05,
      "loss": 0.5511,
      "step": 46000
    },
    {
      "epoch": 0.7410240474255391,
      "grad_norm": 34446.23046875,
      "learning_rate": 2.2601925029098658e-05,
      "loss": 0.556,
      "step": 46500
    },
    {
      "epoch": 0.7489920479354911,
      "grad_norm": 30729.033203125,
      "learning_rate": 2.252220267548219e-05,
      "loss": 0.5568,
      "step": 47000
    },
    {
      "epoch": 0.7569600484454431,
      "grad_norm": 35997.25390625,
      "learning_rate": 2.2442480321865716e-05,
      "loss": 0.5538,
      "step": 47500
    },
    {
      "epoch": 0.7649280489553951,
      "grad_norm": 36570.2578125,
      "learning_rate": 2.2362757968249245e-05,
      "loss": 0.5524,
      "step": 48000
    },
    {
      "epoch": 0.7728960494653472,
      "grad_norm": 38073.46484375,
      "learning_rate": 2.2283035614632774e-05,
      "loss": 0.5564,
      "step": 48500
    },
    {
      "epoch": 0.7808640499752992,
      "grad_norm": 32213.791015625,
      "learning_rate": 2.22033132610163e-05,
      "loss": 0.5531,
      "step": 49000
    },
    {
      "epoch": 0.7888320504852512,
      "grad_norm": 51021.55078125,
      "learning_rate": 2.212359090739983e-05,
      "loss": 0.5551,
      "step": 49500
    },
    {
      "epoch": 0.7968000509952032,
      "grad_norm": 34028.9609375,
      "learning_rate": 2.2043868553783358e-05,
      "loss": 0.5522,
      "step": 50000
    },
    {
      "epoch": 0.8047680515051553,
      "grad_norm": 35102.0390625,
      "learning_rate": 2.1964146200166887e-05,
      "loss": 0.551,
      "step": 50500
    },
    {
      "epoch": 0.8127360520151073,
      "grad_norm": 38054.03125,
      "learning_rate": 2.1884423846550416e-05,
      "loss": 0.5564,
      "step": 51000
    },
    {
      "epoch": 0.8207040525250594,
      "grad_norm": 34621.86328125,
      "learning_rate": 2.180470149293394e-05,
      "loss": 0.5505,
      "step": 51500
    },
    {
      "epoch": 0.8286720530350113,
      "grad_norm": 34099.5859375,
      "learning_rate": 2.172497913931747e-05,
      "loss": 0.5504,
      "step": 52000
    },
    {
      "epoch": 0.8366400535449634,
      "grad_norm": 36866.79296875,
      "learning_rate": 2.1645256785701e-05,
      "loss": 0.5473,
      "step": 52500
    },
    {
      "epoch": 0.8446080540549155,
      "grad_norm": 38420.94140625,
      "learning_rate": 2.156553443208453e-05,
      "loss": 0.547,
      "step": 53000
    },
    {
      "epoch": 0.8525760545648675,
      "grad_norm": 35430.8984375,
      "learning_rate": 2.1485812078468058e-05,
      "loss": 0.5504,
      "step": 53500
    },
    {
      "epoch": 0.8605440550748196,
      "grad_norm": 35433.71875,
      "learning_rate": 2.1406089724851584e-05,
      "loss": 0.5547,
      "step": 54000
    },
    {
      "epoch": 0.8685120555847715,
      "grad_norm": 41701.05859375,
      "learning_rate": 2.1326367371235113e-05,
      "loss": 0.5517,
      "step": 54500
    },
    {
      "epoch": 0.8764800560947236,
      "grad_norm": 34275.7265625,
      "learning_rate": 2.124664501761864e-05,
      "loss": 0.5496,
      "step": 55000
    }
  ],
  "logging_steps": 500,
  "max_steps": 188253,
  "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": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.1496849408e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}