{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 50.0,
  "eval_steps": 500,
  "global_step": 107800,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.23,
      "grad_norm": 0.16410106420516968,
      "learning_rate": 3.110505565862709e-05,
      "loss": 0.8386,
      "step": 500
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.42184513807296753,
      "learning_rate": 3.0960111317254176e-05,
      "loss": 0.8027,
      "step": 1000
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.24314333498477936,
      "learning_rate": 3.0815166975881264e-05,
      "loss": 0.782,
      "step": 1500
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.6049493551254272,
      "learning_rate": 3.067022263450835e-05,
      "loss": 0.7584,
      "step": 2000
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.6963891386985779,
      "learning_rate": 3.0525278293135433e-05,
      "loss": 0.7384,
      "step": 2500
    },
    {
      "epoch": 1.39,
      "grad_norm": 0.4310186803340912,
      "learning_rate": 3.038033395176252e-05,
      "loss": 0.7254,
      "step": 3000
    },
    {
      "epoch": 1.62,
      "grad_norm": 0.713209331035614,
      "learning_rate": 3.0235389610389613e-05,
      "loss": 0.7181,
      "step": 3500
    },
    {
      "epoch": 1.86,
      "grad_norm": 0.6885315775871277,
      "learning_rate": 3.0090445269016697e-05,
      "loss": 0.703,
      "step": 4000
    },
    {
      "epoch": 2.09,
      "grad_norm": 0.48805415630340576,
      "learning_rate": 2.9945500927643785e-05,
      "loss": 0.6923,
      "step": 4500
    },
    {
      "epoch": 2.32,
      "grad_norm": 0.40450435876846313,
      "learning_rate": 2.9800556586270873e-05,
      "loss": 0.691,
      "step": 5000
    },
    {
      "epoch": 2.55,
      "grad_norm": 0.5716784596443176,
      "learning_rate": 2.965561224489796e-05,
      "loss": 0.6831,
      "step": 5500
    },
    {
      "epoch": 2.78,
      "grad_norm": 0.512924075126648,
      "learning_rate": 2.951066790352505e-05,
      "loss": 0.6783,
      "step": 6000
    },
    {
      "epoch": 3.01,
      "grad_norm": 0.5809201598167419,
      "learning_rate": 2.9365723562152134e-05,
      "loss": 0.6682,
      "step": 6500
    },
    {
      "epoch": 3.25,
      "grad_norm": 1.9776902198791504,
      "learning_rate": 2.922077922077922e-05,
      "loss": 0.6654,
      "step": 7000
    },
    {
      "epoch": 3.48,
      "grad_norm": 0.5193782448768616,
      "learning_rate": 2.9075834879406306e-05,
      "loss": 0.6649,
      "step": 7500
    },
    {
      "epoch": 3.71,
      "grad_norm": 0.5309423804283142,
      "learning_rate": 2.8930890538033397e-05,
      "loss": 0.6616,
      "step": 8000
    },
    {
      "epoch": 3.94,
      "grad_norm": 0.5730230212211609,
      "learning_rate": 2.8785946196660482e-05,
      "loss": 0.6559,
      "step": 8500
    },
    {
      "epoch": 4.17,
      "grad_norm": 2.4345550537109375,
      "learning_rate": 2.864100185528757e-05,
      "loss": 0.6535,
      "step": 9000
    },
    {
      "epoch": 4.41,
      "grad_norm": 0.6113856434822083,
      "learning_rate": 2.8496057513914658e-05,
      "loss": 0.6479,
      "step": 9500
    },
    {
      "epoch": 4.64,
      "grad_norm": 0.5557487607002258,
      "learning_rate": 2.8351113172541746e-05,
      "loss": 0.6416,
      "step": 10000
    },
    {
      "epoch": 4.87,
      "grad_norm": 0.5874636173248291,
      "learning_rate": 2.8206168831168834e-05,
      "loss": 0.6399,
      "step": 10500
    },
    {
      "epoch": 5.1,
      "grad_norm": 0.4734204411506653,
      "learning_rate": 2.8061224489795918e-05,
      "loss": 0.6375,
      "step": 11000
    },
    {
      "epoch": 5.33,
      "grad_norm": 0.5736048221588135,
      "learning_rate": 2.7916280148423006e-05,
      "loss": 0.6336,
      "step": 11500
    },
    {
      "epoch": 5.57,
      "grad_norm": 0.5358075499534607,
      "learning_rate": 2.777133580705009e-05,
      "loss": 0.6255,
      "step": 12000
    },
    {
      "epoch": 5.8,
      "grad_norm": 0.39945241808891296,
      "learning_rate": 2.7626391465677182e-05,
      "loss": 0.6245,
      "step": 12500
    },
    {
      "epoch": 6.03,
      "grad_norm": 0.6585483551025391,
      "learning_rate": 2.7481447124304266e-05,
      "loss": 0.619,
      "step": 13000
    },
    {
      "epoch": 6.26,
      "grad_norm": 0.5212920308113098,
      "learning_rate": 2.7336502782931354e-05,
      "loss": 0.6195,
      "step": 13500
    },
    {
      "epoch": 6.49,
      "grad_norm": 0.6320118308067322,
      "learning_rate": 2.7191558441558442e-05,
      "loss": 0.6114,
      "step": 14000
    },
    {
      "epoch": 6.73,
      "grad_norm": 0.7715293169021606,
      "learning_rate": 2.704661410018553e-05,
      "loss": 0.6088,
      "step": 14500
    },
    {
      "epoch": 6.96,
      "grad_norm": 0.571773111820221,
      "learning_rate": 2.6901669758812618e-05,
      "loss": 0.6103,
      "step": 15000
    },
    {
      "epoch": 7.19,
      "grad_norm": 0.6421319246292114,
      "learning_rate": 2.6756725417439703e-05,
      "loss": 0.6025,
      "step": 15500
    },
    {
      "epoch": 7.42,
      "grad_norm": 0.5680719017982483,
      "learning_rate": 2.661178107606679e-05,
      "loss": 0.5998,
      "step": 16000
    },
    {
      "epoch": 7.65,
      "grad_norm": 0.5537917613983154,
      "learning_rate": 2.646683673469388e-05,
      "loss": 0.5981,
      "step": 16500
    },
    {
      "epoch": 7.88,
      "grad_norm": 0.4245034158229828,
      "learning_rate": 2.6321892393320966e-05,
      "loss": 0.5959,
      "step": 17000
    },
    {
      "epoch": 8.12,
      "grad_norm": 0.48047196865081787,
      "learning_rate": 2.617694805194805e-05,
      "loss": 0.5952,
      "step": 17500
    },
    {
      "epoch": 8.35,
      "grad_norm": 0.4809471368789673,
      "learning_rate": 2.603200371057514e-05,
      "loss": 0.5923,
      "step": 18000
    },
    {
      "epoch": 8.58,
      "grad_norm": 0.4537654519081116,
      "learning_rate": 2.5887059369202227e-05,
      "loss": 0.5878,
      "step": 18500
    },
    {
      "epoch": 8.81,
      "grad_norm": 0.5159969329833984,
      "learning_rate": 2.5742115027829315e-05,
      "loss": 0.5933,
      "step": 19000
    },
    {
      "epoch": 9.04,
      "grad_norm": 0.9971024990081787,
      "learning_rate": 2.5597170686456403e-05,
      "loss": 0.588,
      "step": 19500
    },
    {
      "epoch": 9.28,
      "grad_norm": 0.4496227204799652,
      "learning_rate": 2.5452226345083487e-05,
      "loss": 0.5877,
      "step": 20000
    },
    {
      "epoch": 9.51,
      "grad_norm": 0.5145105719566345,
      "learning_rate": 2.5307282003710575e-05,
      "loss": 0.5821,
      "step": 20500
    },
    {
      "epoch": 9.74,
      "grad_norm": 0.5195399522781372,
      "learning_rate": 2.5162337662337663e-05,
      "loss": 0.5827,
      "step": 21000
    },
    {
      "epoch": 9.97,
      "grad_norm": 0.5431232452392578,
      "learning_rate": 2.501739332096475e-05,
      "loss": 0.5845,
      "step": 21500
    },
    {
      "epoch": 10.2,
      "grad_norm": 0.4878028333187103,
      "learning_rate": 2.4872448979591835e-05,
      "loss": 0.5828,
      "step": 22000
    },
    {
      "epoch": 10.44,
      "grad_norm": 0.591066837310791,
      "learning_rate": 2.4727504638218923e-05,
      "loss": 0.5819,
      "step": 22500
    },
    {
      "epoch": 10.67,
      "grad_norm": 16.133996963500977,
      "learning_rate": 2.458256029684601e-05,
      "loss": 0.5765,
      "step": 23000
    },
    {
      "epoch": 10.9,
      "grad_norm": 0.5531517863273621,
      "learning_rate": 2.44376159554731e-05,
      "loss": 0.5781,
      "step": 23500
    },
    {
      "epoch": 11.13,
      "grad_norm": 0.4941064417362213,
      "learning_rate": 2.4292671614100187e-05,
      "loss": 0.5774,
      "step": 24000
    },
    {
      "epoch": 11.36,
      "grad_norm": 0.5101198554039001,
      "learning_rate": 2.414772727272727e-05,
      "loss": 0.5717,
      "step": 24500
    },
    {
      "epoch": 11.6,
      "grad_norm": 0.40433865785598755,
      "learning_rate": 2.400278293135436e-05,
      "loss": 0.5729,
      "step": 25000
    },
    {
      "epoch": 11.83,
      "grad_norm": 0.5296735763549805,
      "learning_rate": 2.3857838589981448e-05,
      "loss": 0.5766,
      "step": 25500
    },
    {
      "epoch": 12.06,
      "grad_norm": 0.40421661734580994,
      "learning_rate": 2.3712894248608535e-05,
      "loss": 0.5674,
      "step": 26000
    },
    {
      "epoch": 12.29,
      "grad_norm": 0.48923459649086,
      "learning_rate": 2.356794990723562e-05,
      "loss": 0.5673,
      "step": 26500
    },
    {
      "epoch": 12.52,
      "grad_norm": 0.4543541371822357,
      "learning_rate": 2.3423005565862708e-05,
      "loss": 0.5682,
      "step": 27000
    },
    {
      "epoch": 12.76,
      "grad_norm": 0.405658096075058,
      "learning_rate": 2.32780612244898e-05,
      "loss": 0.5654,
      "step": 27500
    },
    {
      "epoch": 12.99,
      "grad_norm": 0.4983844459056854,
      "learning_rate": 2.3133116883116884e-05,
      "loss": 0.5657,
      "step": 28000
    },
    {
      "epoch": 13.22,
      "grad_norm": 1.2178577184677124,
      "learning_rate": 2.2988172541743972e-05,
      "loss": 0.5638,
      "step": 28500
    },
    {
      "epoch": 13.45,
      "grad_norm": 0.4428790509700775,
      "learning_rate": 2.2843228200371056e-05,
      "loss": 0.5658,
      "step": 29000
    },
    {
      "epoch": 13.68,
      "grad_norm": 0.5377063155174255,
      "learning_rate": 2.2698283858998144e-05,
      "loss": 0.5609,
      "step": 29500
    },
    {
      "epoch": 13.91,
      "grad_norm": 0.4518877863883972,
      "learning_rate": 2.2553339517625232e-05,
      "loss": 0.5629,
      "step": 30000
    },
    {
      "epoch": 14.15,
      "grad_norm": 0.5720136761665344,
      "learning_rate": 2.240839517625232e-05,
      "loss": 0.5639,
      "step": 30500
    },
    {
      "epoch": 14.38,
      "grad_norm": 0.6480860114097595,
      "learning_rate": 2.2263450834879408e-05,
      "loss": 0.5622,
      "step": 31000
    },
    {
      "epoch": 14.61,
      "grad_norm": 0.40764591097831726,
      "learning_rate": 2.2118506493506492e-05,
      "loss": 0.5615,
      "step": 31500
    },
    {
      "epoch": 14.84,
      "grad_norm": 0.60052889585495,
      "learning_rate": 2.1973562152133584e-05,
      "loss": 0.5542,
      "step": 32000
    },
    {
      "epoch": 15.07,
      "grad_norm": 0.5227585434913635,
      "learning_rate": 2.182861781076067e-05,
      "loss": 0.5556,
      "step": 32500
    },
    {
      "epoch": 15.31,
      "grad_norm": 0.5119400024414062,
      "learning_rate": 2.1683673469387756e-05,
      "loss": 0.556,
      "step": 33000
    },
    {
      "epoch": 15.54,
      "grad_norm": 0.4895058870315552,
      "learning_rate": 2.153872912801484e-05,
      "loss": 0.5569,
      "step": 33500
    },
    {
      "epoch": 15.77,
      "grad_norm": 0.5205515623092651,
      "learning_rate": 2.1393784786641932e-05,
      "loss": 0.555,
      "step": 34000
    },
    {
      "epoch": 16.0,
      "grad_norm": 0.4375595152378082,
      "learning_rate": 2.1248840445269017e-05,
      "loss": 0.5544,
      "step": 34500
    },
    {
      "epoch": 16.23,
      "grad_norm": 0.4849925637245178,
      "learning_rate": 2.1103896103896105e-05,
      "loss": 0.5536,
      "step": 35000
    },
    {
      "epoch": 16.47,
      "grad_norm": 0.5083441734313965,
      "learning_rate": 2.0958951762523192e-05,
      "loss": 0.5522,
      "step": 35500
    },
    {
      "epoch": 16.7,
      "grad_norm": 0.561335027217865,
      "learning_rate": 2.0814007421150277e-05,
      "loss": 0.5523,
      "step": 36000
    },
    {
      "epoch": 16.93,
      "grad_norm": 0.463858962059021,
      "learning_rate": 2.066906307977737e-05,
      "loss": 0.5506,
      "step": 36500
    },
    {
      "epoch": 17.16,
      "grad_norm": 0.5114089250564575,
      "learning_rate": 2.0524118738404453e-05,
      "loss": 0.5512,
      "step": 37000
    },
    {
      "epoch": 17.39,
      "grad_norm": 0.9562047123908997,
      "learning_rate": 2.037917439703154e-05,
      "loss": 0.5506,
      "step": 37500
    },
    {
      "epoch": 17.63,
      "grad_norm": 0.755348265171051,
      "learning_rate": 2.0234230055658625e-05,
      "loss": 0.5473,
      "step": 38000
    },
    {
      "epoch": 17.86,
      "grad_norm": 0.5000788569450378,
      "learning_rate": 2.0089285714285717e-05,
      "loss": 0.5443,
      "step": 38500
    },
    {
      "epoch": 18.09,
      "grad_norm": 0.5531164407730103,
      "learning_rate": 1.99443413729128e-05,
      "loss": 0.548,
      "step": 39000
    },
    {
      "epoch": 18.32,
      "grad_norm": 0.49754297733306885,
      "learning_rate": 1.979939703153989e-05,
      "loss": 0.5475,
      "step": 39500
    },
    {
      "epoch": 18.55,
      "grad_norm": 0.462971955537796,
      "learning_rate": 1.9654452690166977e-05,
      "loss": 0.5459,
      "step": 40000
    },
    {
      "epoch": 18.78,
      "grad_norm": 0.9089665412902832,
      "learning_rate": 1.950950834879406e-05,
      "loss": 0.5491,
      "step": 40500
    },
    {
      "epoch": 19.02,
      "grad_norm": 0.47444915771484375,
      "learning_rate": 1.9364564007421153e-05,
      "loss": 0.5433,
      "step": 41000
    },
    {
      "epoch": 19.25,
      "grad_norm": 0.45020779967308044,
      "learning_rate": 1.9219619666048237e-05,
      "loss": 0.5423,
      "step": 41500
    },
    {
      "epoch": 19.48,
      "grad_norm": 0.6115174293518066,
      "learning_rate": 1.9074675324675325e-05,
      "loss": 0.5455,
      "step": 42000
    },
    {
      "epoch": 19.71,
      "grad_norm": 2.4381721019744873,
      "learning_rate": 1.892973098330241e-05,
      "loss": 0.5433,
      "step": 42500
    },
    {
      "epoch": 19.94,
      "grad_norm": 0.5207477807998657,
      "learning_rate": 1.87847866419295e-05,
      "loss": 0.5414,
      "step": 43000
    },
    {
      "epoch": 20.18,
      "grad_norm": 0.4906873106956482,
      "learning_rate": 1.8639842300556586e-05,
      "loss": 0.5428,
      "step": 43500
    },
    {
      "epoch": 20.41,
      "grad_norm": 0.4558812081813812,
      "learning_rate": 1.8494897959183674e-05,
      "loss": 0.5464,
      "step": 44000
    },
    {
      "epoch": 20.64,
      "grad_norm": 0.5274574756622314,
      "learning_rate": 1.834995361781076e-05,
      "loss": 0.5387,
      "step": 44500
    },
    {
      "epoch": 20.87,
      "grad_norm": 0.46367010474205017,
      "learning_rate": 1.820500927643785e-05,
      "loss": 0.5407,
      "step": 45000
    },
    {
      "epoch": 21.1,
      "grad_norm": 0.45425930619239807,
      "learning_rate": 1.8060064935064937e-05,
      "loss": 0.5444,
      "step": 45500
    },
    {
      "epoch": 21.34,
      "grad_norm": 0.4426890015602112,
      "learning_rate": 1.7915120593692022e-05,
      "loss": 0.5409,
      "step": 46000
    },
    {
      "epoch": 21.57,
      "grad_norm": 0.4001689851284027,
      "learning_rate": 1.777017625231911e-05,
      "loss": 0.5375,
      "step": 46500
    },
    {
      "epoch": 21.8,
      "grad_norm": 0.45908644795417786,
      "learning_rate": 1.7625231910946194e-05,
      "loss": 0.5375,
      "step": 47000
    },
    {
      "epoch": 22.03,
      "grad_norm": 0.41069209575653076,
      "learning_rate": 1.7480287569573286e-05,
      "loss": 0.5364,
      "step": 47500
    },
    {
      "epoch": 22.26,
      "grad_norm": 0.5071306824684143,
      "learning_rate": 1.733534322820037e-05,
      "loss": 0.5392,
      "step": 48000
    },
    {
      "epoch": 22.5,
      "grad_norm": 0.5278964638710022,
      "learning_rate": 1.7190398886827458e-05,
      "loss": 0.5351,
      "step": 48500
    },
    {
      "epoch": 22.73,
      "grad_norm": 0.5334340929985046,
      "learning_rate": 1.7045454545454546e-05,
      "loss": 0.5347,
      "step": 49000
    },
    {
      "epoch": 22.96,
      "grad_norm": 0.47415098547935486,
      "learning_rate": 1.6900510204081634e-05,
      "loss": 0.5378,
      "step": 49500
    },
    {
      "epoch": 23.19,
      "grad_norm": 0.4583585858345032,
      "learning_rate": 1.6755565862708722e-05,
      "loss": 0.5396,
      "step": 50000
    },
    {
      "epoch": 23.42,
      "grad_norm": 0.42494043707847595,
      "learning_rate": 1.6610621521335807e-05,
      "loss": 0.5354,
      "step": 50500
    },
    {
      "epoch": 23.65,
      "grad_norm": 0.46691572666168213,
      "learning_rate": 1.6465677179962894e-05,
      "loss": 0.5383,
      "step": 51000
    },
    {
      "epoch": 23.89,
      "grad_norm": 0.5475151538848877,
      "learning_rate": 1.6320732838589982e-05,
      "loss": 0.5351,
      "step": 51500
    },
    {
      "epoch": 24.12,
      "grad_norm": 0.4535447061061859,
      "learning_rate": 1.617578849721707e-05,
      "loss": 0.5346,
      "step": 52000
    },
    {
      "epoch": 24.35,
      "grad_norm": 0.5584075450897217,
      "learning_rate": 1.6030844155844155e-05,
      "loss": 0.5345,
      "step": 52500
    },
    {
      "epoch": 24.58,
      "grad_norm": 0.7527787685394287,
      "learning_rate": 1.5885899814471243e-05,
      "loss": 0.5273,
      "step": 53000
    },
    {
      "epoch": 24.81,
      "grad_norm": 0.39385178685188293,
      "learning_rate": 1.574095547309833e-05,
      "loss": 0.5365,
      "step": 53500
    },
    {
      "epoch": 25.05,
      "grad_norm": 0.5498375296592712,
      "learning_rate": 1.559601113172542e-05,
      "loss": 0.5375,
      "step": 54000
    },
    {
      "epoch": 25.28,
      "grad_norm": 0.41572290658950806,
      "learning_rate": 1.5451066790352507e-05,
      "loss": 0.5335,
      "step": 54500
    },
    {
      "epoch": 25.51,
      "grad_norm": 0.4475056827068329,
      "learning_rate": 1.530612244897959e-05,
      "loss": 0.5335,
      "step": 55000
    },
    {
      "epoch": 25.74,
      "grad_norm": 0.48311153054237366,
      "learning_rate": 1.516117810760668e-05,
      "loss": 0.5329,
      "step": 55500
    },
    {
      "epoch": 25.97,
      "grad_norm": 0.48509538173675537,
      "learning_rate": 1.5016233766233767e-05,
      "loss": 0.5307,
      "step": 56000
    },
    {
      "epoch": 26.21,
      "grad_norm": 0.5727821588516235,
      "learning_rate": 1.4871289424860855e-05,
      "loss": 0.5316,
      "step": 56500
    },
    {
      "epoch": 26.44,
      "grad_norm": 0.46595117449760437,
      "learning_rate": 1.4726345083487941e-05,
      "loss": 0.5309,
      "step": 57000
    },
    {
      "epoch": 26.67,
      "grad_norm": 0.5022342801094055,
      "learning_rate": 1.4581400742115027e-05,
      "loss": 0.5354,
      "step": 57500
    },
    {
      "epoch": 26.9,
      "grad_norm": 0.5400568842887878,
      "learning_rate": 1.4436456400742115e-05,
      "loss": 0.5268,
      "step": 58000
    },
    {
      "epoch": 27.13,
      "grad_norm": 0.49959874153137207,
      "learning_rate": 1.4291512059369201e-05,
      "loss": 0.5316,
      "step": 58500
    },
    {
      "epoch": 27.37,
      "grad_norm": 0.5435352921485901,
      "learning_rate": 1.414656771799629e-05,
      "loss": 0.53,
      "step": 59000
    },
    {
      "epoch": 27.6,
      "grad_norm": 0.5168443918228149,
      "learning_rate": 1.4001623376623376e-05,
      "loss": 0.5272,
      "step": 59500
    },
    {
      "epoch": 27.83,
      "grad_norm": 0.60955411195755,
      "learning_rate": 1.3856679035250465e-05,
      "loss": 0.5283,
      "step": 60000
    },
    {
      "epoch": 28.06,
      "grad_norm": 0.5063374042510986,
      "learning_rate": 1.3711734693877551e-05,
      "loss": 0.5308,
      "step": 60500
    },
    {
      "epoch": 28.29,
      "grad_norm": 0.46151864528656006,
      "learning_rate": 1.356679035250464e-05,
      "loss": 0.5282,
      "step": 61000
    },
    {
      "epoch": 28.53,
      "grad_norm": 0.5224559903144836,
      "learning_rate": 1.3421846011131726e-05,
      "loss": 0.5261,
      "step": 61500
    },
    {
      "epoch": 28.76,
      "grad_norm": 0.43721848726272583,
      "learning_rate": 1.3276901669758814e-05,
      "loss": 0.525,
      "step": 62000
    },
    {
      "epoch": 28.99,
      "grad_norm": 0.5033512115478516,
      "learning_rate": 1.31319573283859e-05,
      "loss": 0.5304,
      "step": 62500
    },
    {
      "epoch": 29.22,
      "grad_norm": 0.5284975171089172,
      "learning_rate": 1.2987012987012988e-05,
      "loss": 0.5291,
      "step": 63000
    },
    {
      "epoch": 29.45,
      "grad_norm": 0.6417981386184692,
      "learning_rate": 1.2842068645640074e-05,
      "loss": 0.5267,
      "step": 63500
    },
    {
      "epoch": 29.68,
      "grad_norm": 0.5176807045936584,
      "learning_rate": 1.269712430426716e-05,
      "loss": 0.5246,
      "step": 64000
    },
    {
      "epoch": 29.92,
      "grad_norm": 0.46041619777679443,
      "learning_rate": 1.255217996289425e-05,
      "loss": 0.527,
      "step": 64500
    },
    {
      "epoch": 30.15,
      "grad_norm": 0.5162688493728638,
      "learning_rate": 1.2407235621521336e-05,
      "loss": 0.5242,
      "step": 65000
    },
    {
      "epoch": 30.38,
      "grad_norm": 0.46508926153182983,
      "learning_rate": 1.2262291280148424e-05,
      "loss": 0.5234,
      "step": 65500
    },
    {
      "epoch": 30.61,
      "grad_norm": 0.4786316156387329,
      "learning_rate": 1.211734693877551e-05,
      "loss": 0.5287,
      "step": 66000
    },
    {
      "epoch": 30.84,
      "grad_norm": 0.5851900577545166,
      "learning_rate": 1.1972402597402598e-05,
      "loss": 0.5252,
      "step": 66500
    },
    {
      "epoch": 31.08,
      "grad_norm": 0.5234825611114502,
      "learning_rate": 1.1827458256029684e-05,
      "loss": 0.5238,
      "step": 67000
    },
    {
      "epoch": 31.31,
      "grad_norm": 0.4396039843559265,
      "learning_rate": 1.1682513914656772e-05,
      "loss": 0.523,
      "step": 67500
    },
    {
      "epoch": 31.54,
      "grad_norm": 0.4660816788673401,
      "learning_rate": 1.1537569573283858e-05,
      "loss": 0.5254,
      "step": 68000
    },
    {
      "epoch": 31.77,
      "grad_norm": 0.4561219811439514,
      "learning_rate": 1.1392625231910948e-05,
      "loss": 0.5202,
      "step": 68500
    },
    {
      "epoch": 32.0,
      "grad_norm": 0.4771936535835266,
      "learning_rate": 1.1247680890538034e-05,
      "loss": 0.526,
      "step": 69000
    },
    {
      "epoch": 32.24,
      "grad_norm": 0.5595078468322754,
      "learning_rate": 1.110273654916512e-05,
      "loss": 0.5233,
      "step": 69500
    },
    {
      "epoch": 32.47,
      "grad_norm": 0.5574076175689697,
      "learning_rate": 1.0957792207792208e-05,
      "loss": 0.5242,
      "step": 70000
    },
    {
      "epoch": 32.7,
      "grad_norm": 0.5540217161178589,
      "learning_rate": 1.0812847866419295e-05,
      "loss": 0.5207,
      "step": 70500
    },
    {
      "epoch": 32.93,
      "grad_norm": 0.4957340359687805,
      "learning_rate": 1.0667903525046383e-05,
      "loss": 0.5234,
      "step": 71000
    },
    {
      "epoch": 33.16,
      "grad_norm": 0.5124360918998718,
      "learning_rate": 1.0522959183673469e-05,
      "loss": 0.5215,
      "step": 71500
    },
    {
      "epoch": 33.4,
      "grad_norm": 0.46945714950561523,
      "learning_rate": 1.0378014842300557e-05,
      "loss": 0.5253,
      "step": 72000
    },
    {
      "epoch": 33.63,
      "grad_norm": 0.379711389541626,
      "learning_rate": 1.0233070500927643e-05,
      "loss": 0.5217,
      "step": 72500
    },
    {
      "epoch": 33.86,
      "grad_norm": 0.6671276092529297,
      "learning_rate": 1.0088126159554733e-05,
      "loss": 0.5184,
      "step": 73000
    },
    {
      "epoch": 34.09,
      "grad_norm": 0.5381418466567993,
      "learning_rate": 9.943181818181819e-06,
      "loss": 0.5273,
      "step": 73500
    },
    {
      "epoch": 34.32,
      "grad_norm": 0.6995874643325806,
      "learning_rate": 9.798237476808907e-06,
      "loss": 0.5203,
      "step": 74000
    },
    {
      "epoch": 34.55,
      "grad_norm": 0.4824086129665375,
      "learning_rate": 9.653293135435993e-06,
      "loss": 0.5186,
      "step": 74500
    },
    {
      "epoch": 34.79,
      "grad_norm": 0.5302020311355591,
      "learning_rate": 9.50834879406308e-06,
      "loss": 0.5194,
      "step": 75000
    },
    {
      "epoch": 35.02,
      "grad_norm": 0.5452972054481506,
      "learning_rate": 9.363404452690167e-06,
      "loss": 0.5244,
      "step": 75500
    },
    {
      "epoch": 35.25,
      "grad_norm": 0.6910521984100342,
      "learning_rate": 9.218460111317253e-06,
      "loss": 0.5185,
      "step": 76000
    },
    {
      "epoch": 35.48,
      "grad_norm": 0.5710629224777222,
      "learning_rate": 9.073515769944341e-06,
      "loss": 0.519,
      "step": 76500
    },
    {
      "epoch": 35.71,
      "grad_norm": 0.4669000804424286,
      "learning_rate": 8.928571428571428e-06,
      "loss": 0.5171,
      "step": 77000
    },
    {
      "epoch": 35.95,
      "grad_norm": 0.508492112159729,
      "learning_rate": 8.783627087198517e-06,
      "loss": 0.5201,
      "step": 77500
    },
    {
      "epoch": 36.18,
      "grad_norm": 0.5005605816841125,
      "learning_rate": 8.638682745825603e-06,
      "loss": 0.5181,
      "step": 78000
    },
    {
      "epoch": 36.41,
      "grad_norm": 0.5115915536880493,
      "learning_rate": 8.493738404452691e-06,
      "loss": 0.517,
      "step": 78500
    },
    {
      "epoch": 36.64,
      "grad_norm": 0.5187366604804993,
      "learning_rate": 8.348794063079778e-06,
      "loss": 0.5197,
      "step": 79000
    },
    {
      "epoch": 36.87,
      "grad_norm": 0.5845080614089966,
      "learning_rate": 8.203849721706865e-06,
      "loss": 0.5209,
      "step": 79500
    },
    {
      "epoch": 37.11,
      "grad_norm": 0.6533055901527405,
      "learning_rate": 8.058905380333952e-06,
      "loss": 0.521,
      "step": 80000
    },
    {
      "epoch": 37.34,
      "grad_norm": 0.4986008107662201,
      "learning_rate": 7.913961038961038e-06,
      "loss": 0.5172,
      "step": 80500
    },
    {
      "epoch": 37.57,
      "grad_norm": 0.557759165763855,
      "learning_rate": 7.769016697588126e-06,
      "loss": 0.5169,
      "step": 81000
    },
    {
      "epoch": 37.8,
      "grad_norm": 0.651156485080719,
      "learning_rate": 7.624072356215214e-06,
      "loss": 0.5179,
      "step": 81500
    },
    {
      "epoch": 38.03,
      "grad_norm": 0.5263897180557251,
      "learning_rate": 7.479128014842301e-06,
      "loss": 0.5202,
      "step": 82000
    },
    {
      "epoch": 38.27,
      "grad_norm": 0.497855007648468,
      "learning_rate": 7.334183673469388e-06,
      "loss": 0.5169,
      "step": 82500
    },
    {
      "epoch": 38.5,
      "grad_norm": 0.5888237357139587,
      "learning_rate": 7.189239332096476e-06,
      "loss": 0.5172,
      "step": 83000
    },
    {
      "epoch": 38.73,
      "grad_norm": 0.7321680188179016,
      "learning_rate": 7.044294990723562e-06,
      "loss": 0.5175,
      "step": 83500
    },
    {
      "epoch": 38.96,
      "grad_norm": 0.5042610168457031,
      "learning_rate": 6.899350649350649e-06,
      "loss": 0.5175,
      "step": 84000
    },
    {
      "epoch": 39.19,
      "grad_norm": 0.6449227333068848,
      "learning_rate": 6.754406307977736e-06,
      "loss": 0.5185,
      "step": 84500
    },
    {
      "epoch": 39.42,
      "grad_norm": 0.4842873215675354,
      "learning_rate": 6.609461966604823e-06,
      "loss": 0.5174,
      "step": 85000
    },
    {
      "epoch": 39.66,
      "grad_norm": 0.5853040218353271,
      "learning_rate": 6.464517625231911e-06,
      "loss": 0.5158,
      "step": 85500
    },
    {
      "epoch": 39.89,
      "grad_norm": 0.6231125593185425,
      "learning_rate": 6.319573283858998e-06,
      "loss": 0.518,
      "step": 86000
    },
    {
      "epoch": 40.12,
      "grad_norm": 0.7321327328681946,
      "learning_rate": 6.174628942486085e-06,
      "loss": 0.5101,
      "step": 86500
    },
    {
      "epoch": 40.35,
      "grad_norm": 0.48840710520744324,
      "learning_rate": 6.0296846011131725e-06,
      "loss": 0.5188,
      "step": 87000
    },
    {
      "epoch": 40.58,
      "grad_norm": 0.4609430134296417,
      "learning_rate": 5.88474025974026e-06,
      "loss": 0.5156,
      "step": 87500
    },
    {
      "epoch": 40.82,
      "grad_norm": 0.7623891234397888,
      "learning_rate": 5.7397959183673475e-06,
      "loss": 0.5135,
      "step": 88000
    },
    {
      "epoch": 41.05,
      "grad_norm": 0.6465628147125244,
      "learning_rate": 5.5948515769944346e-06,
      "loss": 0.5148,
      "step": 88500
    },
    {
      "epoch": 41.28,
      "grad_norm": 0.6189027428627014,
      "learning_rate": 5.449907235621521e-06,
      "loss": 0.5162,
      "step": 89000
    },
    {
      "epoch": 41.51,
      "grad_norm": 0.48868080973625183,
      "learning_rate": 5.304962894248609e-06,
      "loss": 0.516,
      "step": 89500
    },
    {
      "epoch": 41.74,
      "grad_norm": 0.5011306405067444,
      "learning_rate": 5.160018552875696e-06,
      "loss": 0.513,
      "step": 90000
    },
    {
      "epoch": 41.98,
      "grad_norm": 0.596338152885437,
      "learning_rate": 5.015074211502783e-06,
      "loss": 0.5149,
      "step": 90500
    },
    {
      "epoch": 42.21,
      "grad_norm": 0.681685209274292,
      "learning_rate": 4.87012987012987e-06,
      "loss": 0.5139,
      "step": 91000
    },
    {
      "epoch": 42.44,
      "grad_norm": 0.6109485030174255,
      "learning_rate": 4.725185528756957e-06,
      "loss": 0.5145,
      "step": 91500
    },
    {
      "epoch": 42.67,
      "grad_norm": 0.6468715071678162,
      "learning_rate": 4.580241187384045e-06,
      "loss": 0.5138,
      "step": 92000
    },
    {
      "epoch": 42.9,
      "grad_norm": 0.47962233424186707,
      "learning_rate": 4.435296846011132e-06,
      "loss": 0.5095,
      "step": 92500
    },
    {
      "epoch": 43.14,
      "grad_norm": 0.6052845120429993,
      "learning_rate": 4.290352504638219e-06,
      "loss": 0.5129,
      "step": 93000
    },
    {
      "epoch": 43.37,
      "grad_norm": 0.49540799856185913,
      "learning_rate": 4.145408163265306e-06,
      "loss": 0.5161,
      "step": 93500
    },
    {
      "epoch": 43.6,
      "grad_norm": 0.5836204290390015,
      "learning_rate": 4.000463821892394e-06,
      "loss": 0.509,
      "step": 94000
    },
    {
      "epoch": 43.83,
      "grad_norm": 0.42629340291023254,
      "learning_rate": 3.85551948051948e-06,
      "loss": 0.5135,
      "step": 94500
    },
    {
      "epoch": 44.06,
      "grad_norm": 0.4765370488166809,
      "learning_rate": 3.710575139146568e-06,
      "loss": 0.5107,
      "step": 95000
    },
    {
      "epoch": 44.29,
      "grad_norm": 0.6870307326316833,
      "learning_rate": 3.565630797773655e-06,
      "loss": 0.5134,
      "step": 95500
    },
    {
      "epoch": 44.53,
      "grad_norm": 0.4927411675453186,
      "learning_rate": 3.4206864564007424e-06,
      "loss": 0.5083,
      "step": 96000
    },
    {
      "epoch": 44.76,
      "grad_norm": 0.5409857034683228,
      "learning_rate": 3.275742115027829e-06,
      "loss": 0.5133,
      "step": 96500
    },
    {
      "epoch": 44.99,
      "grad_norm": 0.5635824799537659,
      "learning_rate": 3.1307977736549166e-06,
      "loss": 0.5111,
      "step": 97000
    },
    {
      "epoch": 45.22,
      "grad_norm": 0.6565483212471008,
      "learning_rate": 2.9858534322820037e-06,
      "loss": 0.5099,
      "step": 97500
    },
    {
      "epoch": 45.45,
      "grad_norm": 0.659949779510498,
      "learning_rate": 2.840909090909091e-06,
      "loss": 0.5124,
      "step": 98000
    },
    {
      "epoch": 45.69,
      "grad_norm": 0.551925539970398,
      "learning_rate": 2.6959647495361782e-06,
      "loss": 0.5105,
      "step": 98500
    },
    {
      "epoch": 45.92,
      "grad_norm": 0.6591493487358093,
      "learning_rate": 2.5510204081632653e-06,
      "loss": 0.5106,
      "step": 99000
    },
    {
      "epoch": 46.15,
      "grad_norm": 0.6934979557991028,
      "learning_rate": 2.4060760667903524e-06,
      "loss": 0.5109,
      "step": 99500
    },
    {
      "epoch": 46.38,
      "grad_norm": 0.5429938435554504,
      "learning_rate": 2.26113172541744e-06,
      "loss": 0.5081,
      "step": 100000
    },
    {
      "epoch": 46.61,
      "grad_norm": 0.5906868577003479,
      "learning_rate": 2.116187384044527e-06,
      "loss": 0.5106,
      "step": 100500
    },
    {
      "epoch": 46.85,
      "grad_norm": 0.7781776785850525,
      "learning_rate": 1.971243042671614e-06,
      "loss": 0.5131,
      "step": 101000
    },
    {
      "epoch": 47.08,
      "grad_norm": 0.5847499966621399,
      "learning_rate": 1.8262987012987013e-06,
      "loss": 0.5101,
      "step": 101500
    },
    {
      "epoch": 47.31,
      "grad_norm": 0.49553003907203674,
      "learning_rate": 1.6813543599257886e-06,
      "loss": 0.5129,
      "step": 102000
    },
    {
      "epoch": 47.54,
      "grad_norm": 0.581206738948822,
      "learning_rate": 1.5364100185528757e-06,
      "loss": 0.5122,
      "step": 102500
    },
    {
      "epoch": 47.77,
      "grad_norm": 0.7169119715690613,
      "learning_rate": 1.391465677179963e-06,
      "loss": 0.5075,
      "step": 103000
    },
    {
      "epoch": 48.01,
      "grad_norm": 0.7762653231620789,
      "learning_rate": 1.24652133580705e-06,
      "loss": 0.5121,
      "step": 103500
    },
    {
      "epoch": 48.24,
      "grad_norm": 0.6211891770362854,
      "learning_rate": 1.1015769944341372e-06,
      "loss": 0.5096,
      "step": 104000
    },
    {
      "epoch": 48.47,
      "grad_norm": 0.6150048971176147,
      "learning_rate": 9.566326530612244e-07,
      "loss": 0.5141,
      "step": 104500
    },
    {
      "epoch": 48.7,
      "grad_norm": 0.5785318613052368,
      "learning_rate": 8.116883116883117e-07,
      "loss": 0.5066,
      "step": 105000
    },
    {
      "epoch": 48.93,
      "grad_norm": 0.64094078540802,
      "learning_rate": 6.667439703153989e-07,
      "loss": 0.5066,
      "step": 105500
    },
    {
      "epoch": 49.17,
      "grad_norm": 0.5458254218101501,
      "learning_rate": 5.217996289424861e-07,
      "loss": 0.5083,
      "step": 106000
    },
    {
      "epoch": 49.4,
      "grad_norm": 0.6012634634971619,
      "learning_rate": 3.768552875695733e-07,
      "loss": 0.5112,
      "step": 106500
    },
    {
      "epoch": 49.63,
      "grad_norm": 0.6441764235496521,
      "learning_rate": 2.319109461966605e-07,
      "loss": 0.5094,
      "step": 107000
    },
    {
      "epoch": 49.86,
      "grad_norm": 0.5359529852867126,
      "learning_rate": 8.696660482374769e-08,
      "loss": 0.5095,
      "step": 107500
    },
    {
      "epoch": 50.0,
      "step": 107800,
      "total_flos": 8.716413062401229e+19,
      "train_loss": 0.5550972292020724,
      "train_runtime": 13976.6946,
      "train_samples_per_second": 61.703,
      "train_steps_per_second": 7.713
    }
  ],
  "logging_steps": 500,
  "max_steps": 107800,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 50,
  "save_steps": 1000000000,
  "total_flos": 8.716413062401229e+19,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}