{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.8927444794952681,
  "eval_steps": 500,
  "global_step": 12000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 9.999684542586752e-05,
      "loss": 12.4787,
      "step": 1
    },
    {
      "epoch": 0.02,
      "learning_rate": 9.96845425867508e-05,
      "loss": 1.3146,
      "step": 100
    },
    {
      "epoch": 0.03,
      "learning_rate": 9.936908517350158e-05,
      "loss": 0.0928,
      "step": 200
    },
    {
      "epoch": 0.05,
      "learning_rate": 9.905362776025237e-05,
      "loss": 0.0903,
      "step": 300
    },
    {
      "epoch": 0.06,
      "learning_rate": 9.873817034700316e-05,
      "loss": 0.0675,
      "step": 400
    },
    {
      "epoch": 0.08,
      "learning_rate": 9.842271293375394e-05,
      "loss": 0.0744,
      "step": 500
    },
    {
      "epoch": 0.08,
      "eval_loss": 0.06298290193080902,
      "eval_runtime": 191.3725,
      "eval_samples_per_second": 113.574,
      "eval_steps_per_second": 14.197,
      "step": 500
    },
    {
      "epoch": 0.09,
      "learning_rate": 9.810725552050474e-05,
      "loss": 0.0682,
      "step": 600
    },
    {
      "epoch": 0.11,
      "learning_rate": 9.779179810725552e-05,
      "loss": 0.072,
      "step": 700
    },
    {
      "epoch": 0.13,
      "learning_rate": 9.747634069400632e-05,
      "loss": 0.0731,
      "step": 800
    },
    {
      "epoch": 0.14,
      "learning_rate": 9.71608832807571e-05,
      "loss": 0.0638,
      "step": 900
    },
    {
      "epoch": 0.16,
      "learning_rate": 9.684542586750788e-05,
      "loss": 0.062,
      "step": 1000
    },
    {
      "epoch": 0.16,
      "eval_loss": 0.056409742683172226,
      "eval_runtime": 191.8919,
      "eval_samples_per_second": 113.267,
      "eval_steps_per_second": 14.159,
      "step": 1000
    },
    {
      "epoch": 0.17,
      "learning_rate": 9.652996845425868e-05,
      "loss": 0.0663,
      "step": 1100
    },
    {
      "epoch": 0.19,
      "learning_rate": 9.621451104100947e-05,
      "loss": 0.0598,
      "step": 1200
    },
    {
      "epoch": 0.21,
      "learning_rate": 9.589905362776026e-05,
      "loss": 0.0625,
      "step": 1300
    },
    {
      "epoch": 0.22,
      "learning_rate": 9.558359621451105e-05,
      "loss": 0.0673,
      "step": 1400
    },
    {
      "epoch": 0.24,
      "learning_rate": 9.526813880126184e-05,
      "loss": 0.0625,
      "step": 1500
    },
    {
      "epoch": 0.24,
      "eval_loss": 0.05498537793755531,
      "eval_runtime": 191.898,
      "eval_samples_per_second": 113.263,
      "eval_steps_per_second": 14.159,
      "step": 1500
    },
    {
      "epoch": 0.25,
      "learning_rate": 9.495268138801262e-05,
      "loss": 0.0606,
      "step": 1600
    },
    {
      "epoch": 0.27,
      "learning_rate": 9.463722397476341e-05,
      "loss": 0.0684,
      "step": 1700
    },
    {
      "epoch": 0.28,
      "learning_rate": 9.43217665615142e-05,
      "loss": 0.0607,
      "step": 1800
    },
    {
      "epoch": 0.3,
      "learning_rate": 9.400630914826499e-05,
      "loss": 0.0679,
      "step": 1900
    },
    {
      "epoch": 0.32,
      "learning_rate": 9.369085173501577e-05,
      "loss": 0.0633,
      "step": 2000
    },
    {
      "epoch": 0.32,
      "eval_loss": 0.052233804017305374,
      "eval_runtime": 191.8762,
      "eval_samples_per_second": 113.276,
      "eval_steps_per_second": 14.16,
      "step": 2000
    },
    {
      "epoch": 0.33,
      "learning_rate": 9.337539432176656e-05,
      "loss": 0.061,
      "step": 2100
    },
    {
      "epoch": 0.35,
      "learning_rate": 9.305993690851735e-05,
      "loss": 0.0715,
      "step": 2200
    },
    {
      "epoch": 0.36,
      "learning_rate": 9.274447949526815e-05,
      "loss": 0.0586,
      "step": 2300
    },
    {
      "epoch": 0.38,
      "learning_rate": 9.242902208201893e-05,
      "loss": 0.0581,
      "step": 2400
    },
    {
      "epoch": 0.39,
      "learning_rate": 9.211356466876973e-05,
      "loss": 0.0588,
      "step": 2500
    },
    {
      "epoch": 0.39,
      "eval_loss": 0.05040860176086426,
      "eval_runtime": 191.6049,
      "eval_samples_per_second": 113.437,
      "eval_steps_per_second": 14.18,
      "step": 2500
    },
    {
      "epoch": 0.41,
      "learning_rate": 9.179810725552051e-05,
      "loss": 0.0572,
      "step": 2600
    },
    {
      "epoch": 0.43,
      "learning_rate": 9.148264984227129e-05,
      "loss": 0.0555,
      "step": 2700
    },
    {
      "epoch": 0.44,
      "learning_rate": 9.116719242902209e-05,
      "loss": 0.0547,
      "step": 2800
    },
    {
      "epoch": 0.46,
      "learning_rate": 9.085173501577287e-05,
      "loss": 0.0596,
      "step": 2900
    },
    {
      "epoch": 0.47,
      "learning_rate": 9.053627760252367e-05,
      "loss": 0.0593,
      "step": 3000
    },
    {
      "epoch": 0.47,
      "eval_loss": 0.04637610912322998,
      "eval_runtime": 192.1829,
      "eval_samples_per_second": 113.095,
      "eval_steps_per_second": 14.138,
      "step": 3000
    },
    {
      "epoch": 0.49,
      "learning_rate": 9.022082018927446e-05,
      "loss": 0.0536,
      "step": 3100
    },
    {
      "epoch": 0.5,
      "learning_rate": 8.990536277602523e-05,
      "loss": 0.0506,
      "step": 3200
    },
    {
      "epoch": 0.52,
      "learning_rate": 8.958990536277603e-05,
      "loss": 0.0601,
      "step": 3300
    },
    {
      "epoch": 0.54,
      "learning_rate": 8.927444794952682e-05,
      "loss": 0.05,
      "step": 3400
    },
    {
      "epoch": 0.55,
      "learning_rate": 8.89589905362776e-05,
      "loss": 0.0492,
      "step": 3500
    },
    {
      "epoch": 0.55,
      "eval_loss": 0.0455920547246933,
      "eval_runtime": 192.0796,
      "eval_samples_per_second": 113.156,
      "eval_steps_per_second": 14.145,
      "step": 3500
    },
    {
      "epoch": 0.57,
      "learning_rate": 8.86435331230284e-05,
      "loss": 0.0547,
      "step": 3600
    },
    {
      "epoch": 0.58,
      "learning_rate": 8.832807570977918e-05,
      "loss": 0.0582,
      "step": 3700
    },
    {
      "epoch": 0.6,
      "learning_rate": 8.801261829652997e-05,
      "loss": 0.0535,
      "step": 3800
    },
    {
      "epoch": 0.62,
      "learning_rate": 8.769716088328076e-05,
      "loss": 0.0554,
      "step": 3900
    },
    {
      "epoch": 0.63,
      "learning_rate": 8.738170347003154e-05,
      "loss": 0.0506,
      "step": 4000
    },
    {
      "epoch": 0.63,
      "eval_loss": 0.04419832304120064,
      "eval_runtime": 192.193,
      "eval_samples_per_second": 113.089,
      "eval_steps_per_second": 14.137,
      "step": 4000
    },
    {
      "epoch": 0.65,
      "learning_rate": 8.706624605678234e-05,
      "loss": 0.0508,
      "step": 4100
    },
    {
      "epoch": 0.66,
      "learning_rate": 8.675078864353314e-05,
      "loss": 0.0566,
      "step": 4200
    },
    {
      "epoch": 0.68,
      "learning_rate": 8.64353312302839e-05,
      "loss": 0.0556,
      "step": 4300
    },
    {
      "epoch": 0.69,
      "learning_rate": 8.61198738170347e-05,
      "loss": 0.0595,
      "step": 4400
    },
    {
      "epoch": 0.71,
      "learning_rate": 8.58044164037855e-05,
      "loss": 0.0584,
      "step": 4500
    },
    {
      "epoch": 0.71,
      "eval_loss": 0.043363627046346664,
      "eval_runtime": 191.6803,
      "eval_samples_per_second": 113.392,
      "eval_steps_per_second": 14.175,
      "step": 4500
    },
    {
      "epoch": 0.73,
      "learning_rate": 8.548895899053628e-05,
      "loss": 0.0546,
      "step": 4600
    },
    {
      "epoch": 0.74,
      "learning_rate": 8.517350157728708e-05,
      "loss": 0.0457,
      "step": 4700
    },
    {
      "epoch": 0.76,
      "learning_rate": 8.485804416403787e-05,
      "loss": 0.0531,
      "step": 4800
    },
    {
      "epoch": 0.77,
      "learning_rate": 8.454258675078864e-05,
      "loss": 0.0506,
      "step": 4900
    },
    {
      "epoch": 0.79,
      "learning_rate": 8.422712933753944e-05,
      "loss": 0.0455,
      "step": 5000
    },
    {
      "epoch": 0.79,
      "eval_loss": 0.04175921157002449,
      "eval_runtime": 192.0012,
      "eval_samples_per_second": 113.202,
      "eval_steps_per_second": 14.151,
      "step": 5000
    },
    {
      "epoch": 0.8,
      "learning_rate": 8.391167192429022e-05,
      "loss": 0.053,
      "step": 5100
    },
    {
      "epoch": 0.82,
      "learning_rate": 8.359621451104101e-05,
      "loss": 0.0499,
      "step": 5200
    },
    {
      "epoch": 0.84,
      "learning_rate": 8.328075709779181e-05,
      "loss": 0.0407,
      "step": 5300
    },
    {
      "epoch": 0.85,
      "learning_rate": 8.296529968454258e-05,
      "loss": 0.0559,
      "step": 5400
    },
    {
      "epoch": 0.87,
      "learning_rate": 8.264984227129337e-05,
      "loss": 0.0477,
      "step": 5500
    },
    {
      "epoch": 0.87,
      "eval_loss": 0.042047169059515,
      "eval_runtime": 192.1,
      "eval_samples_per_second": 113.144,
      "eval_steps_per_second": 14.144,
      "step": 5500
    },
    {
      "epoch": 0.88,
      "learning_rate": 8.233438485804417e-05,
      "loss": 0.0488,
      "step": 5600
    },
    {
      "epoch": 0.9,
      "learning_rate": 8.201892744479495e-05,
      "loss": 0.0537,
      "step": 5700
    },
    {
      "epoch": 0.91,
      "learning_rate": 8.170347003154575e-05,
      "loss": 0.0515,
      "step": 5800
    },
    {
      "epoch": 0.93,
      "learning_rate": 8.138801261829655e-05,
      "loss": 0.0511,
      "step": 5900
    },
    {
      "epoch": 0.95,
      "learning_rate": 8.107255520504731e-05,
      "loss": 0.0548,
      "step": 6000
    },
    {
      "epoch": 0.95,
      "eval_loss": 0.039724551141262054,
      "eval_runtime": 191.9769,
      "eval_samples_per_second": 113.217,
      "eval_steps_per_second": 14.153,
      "step": 6000
    },
    {
      "epoch": 0.96,
      "learning_rate": 8.075709779179811e-05,
      "loss": 0.0482,
      "step": 6100
    },
    {
      "epoch": 0.98,
      "learning_rate": 8.04416403785489e-05,
      "loss": 0.0474,
      "step": 6200
    },
    {
      "epoch": 0.99,
      "learning_rate": 8.012618296529969e-05,
      "loss": 0.0488,
      "step": 6300
    },
    {
      "epoch": 1.01,
      "learning_rate": 7.981072555205048e-05,
      "loss": 0.0346,
      "step": 6400
    },
    {
      "epoch": 1.03,
      "learning_rate": 7.949526813880127e-05,
      "loss": 0.035,
      "step": 6500
    },
    {
      "epoch": 1.03,
      "eval_loss": 0.039325978606939316,
      "eval_runtime": 191.6608,
      "eval_samples_per_second": 113.403,
      "eval_steps_per_second": 14.176,
      "step": 6500
    },
    {
      "epoch": 1.04,
      "learning_rate": 7.917981072555205e-05,
      "loss": 0.0355,
      "step": 6600
    },
    {
      "epoch": 1.06,
      "learning_rate": 7.886435331230284e-05,
      "loss": 0.0363,
      "step": 6700
    },
    {
      "epoch": 1.07,
      "learning_rate": 7.854889589905363e-05,
      "loss": 0.0407,
      "step": 6800
    },
    {
      "epoch": 1.09,
      "learning_rate": 7.823343848580442e-05,
      "loss": 0.0385,
      "step": 6900
    },
    {
      "epoch": 1.1,
      "learning_rate": 7.791798107255522e-05,
      "loss": 0.0347,
      "step": 7000
    },
    {
      "epoch": 1.1,
      "eval_loss": 0.03924637660384178,
      "eval_runtime": 192.1827,
      "eval_samples_per_second": 113.096,
      "eval_steps_per_second": 14.138,
      "step": 7000
    },
    {
      "epoch": 1.12,
      "learning_rate": 7.760252365930599e-05,
      "loss": 0.0357,
      "step": 7100
    },
    {
      "epoch": 1.14,
      "learning_rate": 7.728706624605678e-05,
      "loss": 0.041,
      "step": 7200
    },
    {
      "epoch": 1.15,
      "learning_rate": 7.697160883280758e-05,
      "loss": 0.0313,
      "step": 7300
    },
    {
      "epoch": 1.17,
      "learning_rate": 7.665615141955836e-05,
      "loss": 0.0336,
      "step": 7400
    },
    {
      "epoch": 1.18,
      "learning_rate": 7.634069400630916e-05,
      "loss": 0.0303,
      "step": 7500
    },
    {
      "epoch": 1.18,
      "eval_loss": 0.03979608044028282,
      "eval_runtime": 192.103,
      "eval_samples_per_second": 113.142,
      "eval_steps_per_second": 14.143,
      "step": 7500
    },
    {
      "epoch": 1.2,
      "learning_rate": 7.602523659305994e-05,
      "loss": 0.0356,
      "step": 7600
    },
    {
      "epoch": 1.21,
      "learning_rate": 7.570977917981072e-05,
      "loss": 0.0379,
      "step": 7700
    },
    {
      "epoch": 1.23,
      "learning_rate": 7.539432176656152e-05,
      "loss": 0.0365,
      "step": 7800
    },
    {
      "epoch": 1.25,
      "learning_rate": 7.50788643533123e-05,
      "loss": 0.0393,
      "step": 7900
    },
    {
      "epoch": 1.26,
      "learning_rate": 7.47634069400631e-05,
      "loss": 0.0374,
      "step": 8000
    },
    {
      "epoch": 1.26,
      "eval_loss": 0.038463614881038666,
      "eval_runtime": 192.1601,
      "eval_samples_per_second": 113.109,
      "eval_steps_per_second": 14.139,
      "step": 8000
    },
    {
      "epoch": 1.28,
      "learning_rate": 7.444794952681389e-05,
      "loss": 0.0303,
      "step": 8100
    },
    {
      "epoch": 1.29,
      "learning_rate": 7.413249211356468e-05,
      "loss": 0.0384,
      "step": 8200
    },
    {
      "epoch": 1.31,
      "learning_rate": 7.381703470031546e-05,
      "loss": 0.0383,
      "step": 8300
    },
    {
      "epoch": 1.32,
      "learning_rate": 7.350157728706625e-05,
      "loss": 0.0378,
      "step": 8400
    },
    {
      "epoch": 1.34,
      "learning_rate": 7.318611987381704e-05,
      "loss": 0.0343,
      "step": 8500
    },
    {
      "epoch": 1.34,
      "eval_loss": 0.03961439058184624,
      "eval_runtime": 192.1114,
      "eval_samples_per_second": 113.137,
      "eval_steps_per_second": 14.143,
      "step": 8500
    },
    {
      "epoch": 1.36,
      "learning_rate": 7.287066246056783e-05,
      "loss": 0.0362,
      "step": 8600
    },
    {
      "epoch": 1.37,
      "learning_rate": 7.255520504731861e-05,
      "loss": 0.0435,
      "step": 8700
    },
    {
      "epoch": 1.39,
      "learning_rate": 7.22397476340694e-05,
      "loss": 0.0382,
      "step": 8800
    },
    {
      "epoch": 1.4,
      "learning_rate": 7.192429022082019e-05,
      "loss": 0.0439,
      "step": 8900
    },
    {
      "epoch": 1.42,
      "learning_rate": 7.160883280757098e-05,
      "loss": 0.0374,
      "step": 9000
    },
    {
      "epoch": 1.42,
      "eval_loss": 0.037410151213407516,
      "eval_runtime": 192.1572,
      "eval_samples_per_second": 113.111,
      "eval_steps_per_second": 14.139,
      "step": 9000
    },
    {
      "epoch": 1.44,
      "learning_rate": 7.129337539432177e-05,
      "loss": 0.0342,
      "step": 9100
    },
    {
      "epoch": 1.45,
      "learning_rate": 7.097791798107257e-05,
      "loss": 0.0337,
      "step": 9200
    },
    {
      "epoch": 1.47,
      "learning_rate": 7.066246056782335e-05,
      "loss": 0.0358,
      "step": 9300
    },
    {
      "epoch": 1.48,
      "learning_rate": 7.034700315457413e-05,
      "loss": 0.0301,
      "step": 9400
    },
    {
      "epoch": 1.5,
      "learning_rate": 7.003154574132493e-05,
      "loss": 0.0362,
      "step": 9500
    },
    {
      "epoch": 1.5,
      "eval_loss": 0.03637044504284859,
      "eval_runtime": 192.1621,
      "eval_samples_per_second": 113.108,
      "eval_steps_per_second": 14.139,
      "step": 9500
    },
    {
      "epoch": 1.51,
      "learning_rate": 6.971608832807571e-05,
      "loss": 0.037,
      "step": 9600
    },
    {
      "epoch": 1.53,
      "learning_rate": 6.94006309148265e-05,
      "loss": 0.0374,
      "step": 9700
    },
    {
      "epoch": 1.55,
      "learning_rate": 6.908517350157729e-05,
      "loss": 0.0379,
      "step": 9800
    },
    {
      "epoch": 1.56,
      "learning_rate": 6.876971608832808e-05,
      "loss": 0.0328,
      "step": 9900
    },
    {
      "epoch": 1.58,
      "learning_rate": 6.845425867507887e-05,
      "loss": 0.0315,
      "step": 10000
    },
    {
      "epoch": 1.58,
      "eval_loss": 0.03636159375309944,
      "eval_runtime": 192.0641,
      "eval_samples_per_second": 113.165,
      "eval_steps_per_second": 14.146,
      "step": 10000
    },
    {
      "epoch": 1.59,
      "learning_rate": 6.813880126182965e-05,
      "loss": 0.0328,
      "step": 10100
    },
    {
      "epoch": 1.61,
      "learning_rate": 6.782334384858045e-05,
      "loss": 0.0355,
      "step": 10200
    },
    {
      "epoch": 1.62,
      "learning_rate": 6.750788643533124e-05,
      "loss": 0.0398,
      "step": 10300
    },
    {
      "epoch": 1.64,
      "learning_rate": 6.719242902208202e-05,
      "loss": 0.0372,
      "step": 10400
    },
    {
      "epoch": 1.66,
      "learning_rate": 6.68769716088328e-05,
      "loss": 0.0399,
      "step": 10500
    },
    {
      "epoch": 1.66,
      "eval_loss": 0.03677404299378395,
      "eval_runtime": 191.6576,
      "eval_samples_per_second": 113.405,
      "eval_steps_per_second": 14.176,
      "step": 10500
    },
    {
      "epoch": 1.67,
      "learning_rate": 6.65615141955836e-05,
      "loss": 0.0394,
      "step": 10600
    },
    {
      "epoch": 1.69,
      "learning_rate": 6.624605678233438e-05,
      "loss": 0.0355,
      "step": 10700
    },
    {
      "epoch": 1.7,
      "learning_rate": 6.593059936908518e-05,
      "loss": 0.0326,
      "step": 10800
    },
    {
      "epoch": 1.72,
      "learning_rate": 6.561514195583596e-05,
      "loss": 0.0307,
      "step": 10900
    },
    {
      "epoch": 1.74,
      "learning_rate": 6.529968454258676e-05,
      "loss": 0.0372,
      "step": 11000
    },
    {
      "epoch": 1.74,
      "eval_loss": 0.03506915271282196,
      "eval_runtime": 192.4202,
      "eval_samples_per_second": 112.956,
      "eval_steps_per_second": 14.12,
      "step": 11000
    },
    {
      "epoch": 1.75,
      "learning_rate": 6.498422712933754e-05,
      "loss": 0.0321,
      "step": 11100
    },
    {
      "epoch": 1.77,
      "learning_rate": 6.466876971608832e-05,
      "loss": 0.0272,
      "step": 11200
    },
    {
      "epoch": 1.78,
      "learning_rate": 6.435331230283912e-05,
      "loss": 0.0322,
      "step": 11300
    },
    {
      "epoch": 1.8,
      "learning_rate": 6.403785488958992e-05,
      "loss": 0.0355,
      "step": 11400
    },
    {
      "epoch": 1.81,
      "learning_rate": 6.37223974763407e-05,
      "loss": 0.0324,
      "step": 11500
    },
    {
      "epoch": 1.81,
      "eval_loss": 0.034638404846191406,
      "eval_runtime": 192.5205,
      "eval_samples_per_second": 112.897,
      "eval_steps_per_second": 14.113,
      "step": 11500
    },
    {
      "epoch": 1.83,
      "learning_rate": 6.34069400630915e-05,
      "loss": 0.0362,
      "step": 11600
    },
    {
      "epoch": 1.85,
      "learning_rate": 6.309148264984228e-05,
      "loss": 0.03,
      "step": 11700
    },
    {
      "epoch": 1.86,
      "learning_rate": 6.277602523659306e-05,
      "loss": 0.0313,
      "step": 11800
    },
    {
      "epoch": 1.88,
      "learning_rate": 6.246056782334385e-05,
      "loss": 0.0348,
      "step": 11900
    },
    {
      "epoch": 1.89,
      "learning_rate": 6.214511041009464e-05,
      "loss": 0.0351,
      "step": 12000
    },
    {
      "epoch": 1.89,
      "eval_loss": 0.03646688908338547,
      "eval_runtime": 192.2872,
      "eval_samples_per_second": 113.034,
      "eval_steps_per_second": 14.13,
      "step": 12000
    }
  ],
  "logging_steps": 100,
  "max_steps": 31700,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 4226972988026880.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}