{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 8.010214409722222,
  "eval_steps": 1000,
  "global_step": 36000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0086102430555555,
      "grad_norm": 0.3148804008960724,
      "learning_rate": 0.0009998758966336297,
      "loss": 5.1098,
      "num_input_tokens_seen": 521485760,
      "step": 1000
    },
    {
      "epoch": 1.0086102430555555,
      "eval_loss": 4.390458106994629,
      "eval_runtime": 157.9404,
      "eval_samples_per_second": 116.5,
      "eval_steps_per_second": 7.281,
      "num_input_tokens_seen": 521485760,
      "step": 1000
    },
    {
      "epoch": 2.0169140625,
      "grad_norm": 0.24177567660808563,
      "learning_rate": 0.0009969004950996173,
      "loss": 3.7649,
      "num_input_tokens_seen": 1042810727,
      "step": 2000
    },
    {
      "epoch": 2.0169140625,
      "eval_loss": 4.018285274505615,
      "eval_runtime": 157.619,
      "eval_samples_per_second": 116.737,
      "eval_steps_per_second": 7.296,
      "num_input_tokens_seen": 1042810727,
      "step": 2000
    },
    {
      "epoch": 4.006030381944444,
      "grad_norm": 0.22102612257003784,
      "learning_rate": 0.0009899808525182935,
      "loss": 3.3244,
      "num_input_tokens_seen": 1564614407,
      "step": 3000
    },
    {
      "epoch": 4.006030381944444,
      "eval_loss": 3.883763074874878,
      "eval_runtime": 158.112,
      "eval_samples_per_second": 116.373,
      "eval_steps_per_second": 7.273,
      "num_input_tokens_seen": 1564614407,
      "step": 3000
    },
    {
      "epoch": 5.0143046875,
      "grad_norm": 0.25155559182167053,
      "learning_rate": 0.0009791718948528457,
      "loss": 3.2291,
      "num_input_tokens_seen": 2086545237,
      "step": 4000
    },
    {
      "epoch": 5.0143046875,
      "eval_loss": 3.7907443046569824,
      "eval_runtime": 157.3668,
      "eval_samples_per_second": 116.924,
      "eval_steps_per_second": 7.308,
      "num_input_tokens_seen": 2086545237,
      "step": 4000
    },
    {
      "epoch": 7.002637152777778,
      "grad_norm": 0.23016224801540375,
      "learning_rate": 0.0009645594202357438,
      "loss": 3.1779,
      "num_input_tokens_seen": 2608572303,
      "step": 5000
    },
    {
      "epoch": 7.002637152777778,
      "eval_loss": 3.7254457473754883,
      "eval_runtime": 158.0679,
      "eval_samples_per_second": 116.406,
      "eval_steps_per_second": 7.275,
      "num_input_tokens_seen": 2608572303,
      "step": 5000
    },
    {
      "epoch": 8.010796006944444,
      "grad_norm": 0.1911187618970871,
      "learning_rate": 0.0009462594179299406,
      "loss": 3.1106,
      "num_input_tokens_seen": 3130451615,
      "step": 6000
    },
    {
      "epoch": 8.010796006944444,
      "eval_loss": 3.6576924324035645,
      "eval_runtime": 156.4171,
      "eval_samples_per_second": 117.634,
      "eval_steps_per_second": 7.352,
      "num_input_tokens_seen": 3130451615,
      "step": 6000
    },
    {
      "epoch": 9.0191015625,
      "grad_norm": 0.24357692897319794,
      "learning_rate": 0.0009244171476423036,
      "loss": 3.0493,
      "num_input_tokens_seen": 3652271439,
      "step": 7000
    },
    {
      "epoch": 9.0191015625,
      "eval_loss": 3.6375908851623535,
      "eval_runtime": 157.211,
      "eval_samples_per_second": 117.04,
      "eval_steps_per_second": 7.315,
      "num_input_tokens_seen": 3652271439,
      "step": 7000
    },
    {
      "epoch": 1.0083489583333334,
      "grad_norm": 0.17816534638404846,
      "learning_rate": 0.0008992059864973972,
      "loss": 3.0005,
      "num_input_tokens_seen": 4174145711,
      "step": 8000
    },
    {
      "epoch": 1.0083489583333334,
      "eval_loss": 3.5855612754821777,
      "eval_runtime": 155.3476,
      "eval_samples_per_second": 118.444,
      "eval_steps_per_second": 7.403,
      "num_input_tokens_seen": 4174145711,
      "step": 8000
    },
    {
      "epoch": 2.016631076388889,
      "grad_norm": 0.1740865856409073,
      "learning_rate": 0.0008708260528239789,
      "loss": 3.0155,
      "num_input_tokens_seen": 4696186623,
      "step": 9000
    },
    {
      "epoch": 2.016631076388889,
      "eval_loss": 3.580111265182495,
      "eval_runtime": 154.9402,
      "eval_samples_per_second": 118.755,
      "eval_steps_per_second": 7.422,
      "num_input_tokens_seen": 4696186623,
      "step": 9000
    },
    {
      "epoch": 4.005178819444445,
      "grad_norm": 0.19117484986782074,
      "learning_rate": 0.0008395026176781626,
      "loss": 2.9609,
      "num_input_tokens_seen": 5217948966,
      "step": 10000
    },
    {
      "epoch": 4.005178819444445,
      "eval_loss": 3.5479941368103027,
      "eval_runtime": 154.297,
      "eval_samples_per_second": 119.251,
      "eval_steps_per_second": 7.453,
      "num_input_tokens_seen": 5217948966,
      "step": 10000
    },
    {
      "epoch": 5.013779513888889,
      "grad_norm": 0.16193148493766785,
      "learning_rate": 0.0008054843167120826,
      "loss": 2.9376,
      "num_input_tokens_seen": 5739503318,
      "step": 11000
    },
    {
      "epoch": 5.013779513888889,
      "eval_loss": 3.5245985984802246,
      "eval_runtime": 154.2866,
      "eval_samples_per_second": 119.259,
      "eval_steps_per_second": 7.454,
      "num_input_tokens_seen": 5739503318,
      "step": 11000
    },
    {
      "epoch": 7.002284722222222,
      "grad_norm": 0.16273680329322815,
      "learning_rate": 0.0007690411765816864,
      "loss": 2.9641,
      "num_input_tokens_seen": 6261603062,
      "step": 12000
    },
    {
      "epoch": 7.002284722222222,
      "eval_loss": 3.5051820278167725,
      "eval_runtime": 153.5212,
      "eval_samples_per_second": 119.853,
      "eval_steps_per_second": 7.491,
      "num_input_tokens_seen": 6261603062,
      "step": 12000
    },
    {
      "epoch": 8.010577256944444,
      "grad_norm": 0.1603420227766037,
      "learning_rate": 0.0007304624715594139,
      "loss": 2.9166,
      "num_input_tokens_seen": 6783593829,
      "step": 13000
    },
    {
      "epoch": 8.010577256944444,
      "eval_loss": 3.497760772705078,
      "eval_runtime": 153.4387,
      "eval_samples_per_second": 119.918,
      "eval_steps_per_second": 7.495,
      "num_input_tokens_seen": 6783593829,
      "step": 13000
    },
    {
      "epoch": 9.018918402777778,
      "grad_norm": 0.1665274053812027,
      "learning_rate": 0.0006900544273653075,
      "loss": 2.8821,
      "num_input_tokens_seen": 7305434352,
      "step": 14000
    },
    {
      "epoch": 9.018918402777778,
      "eval_loss": 3.4890758991241455,
      "eval_runtime": 153.9326,
      "eval_samples_per_second": 119.533,
      "eval_steps_per_second": 7.471,
      "num_input_tokens_seen": 7305434352,
      "step": 14000
    },
    {
      "epoch": 1.0083055555555556,
      "grad_norm": 0.16452226042747498,
      "learning_rate": 0.000648137790442817,
      "loss": 2.8766,
      "num_input_tokens_seen": 7827274720,
      "step": 15000
    },
    {
      "epoch": 1.0083055555555556,
      "eval_loss": 3.4698522090911865,
      "eval_runtime": 153.8232,
      "eval_samples_per_second": 119.618,
      "eval_steps_per_second": 7.476,
      "num_input_tokens_seen": 7827274720,
      "step": 15000
    },
    {
      "epoch": 2.0166302083333334,
      "grad_norm": 0.17168712615966797,
      "learning_rate": 0.0006050452819736389,
      "loss": 2.8809,
      "num_input_tokens_seen": 8348682560,
      "step": 16000
    },
    {
      "epoch": 2.0166302083333334,
      "eval_loss": 3.4530444145202637,
      "eval_runtime": 152.7846,
      "eval_samples_per_second": 120.431,
      "eval_steps_per_second": 7.527,
      "num_input_tokens_seen": 8348682560,
      "step": 16000
    },
    {
      "epoch": 4.005391493055556,
      "grad_norm": 0.16454088687896729,
      "learning_rate": 0.0005611189568408173,
      "loss": 2.885,
      "num_input_tokens_seen": 8870476792,
      "step": 17000
    },
    {
      "epoch": 4.005391493055556,
      "eval_loss": 3.442610025405884,
      "eval_runtime": 153.4312,
      "eval_samples_per_second": 119.923,
      "eval_steps_per_second": 7.495,
      "num_input_tokens_seen": 8870476792,
      "step": 17000
    },
    {
      "epoch": 5.013413194444444,
      "grad_norm": 0.1566428244113922,
      "learning_rate": 0.0005167074885038374,
      "loss": 2.8616,
      "num_input_tokens_seen": 9392395768,
      "step": 18000
    },
    {
      "epoch": 5.013413194444444,
      "eval_loss": 3.4214282035827637,
      "eval_runtime": 153.1872,
      "eval_samples_per_second": 120.115,
      "eval_steps_per_second": 7.507,
      "num_input_tokens_seen": 9392395768,
      "step": 18000
    },
    {
      "epoch": 7.001969618055556,
      "grad_norm": 0.16596154868602753,
      "learning_rate": 0.000472163401337526,
      "loss": 2.8546,
      "num_input_tokens_seen": 9914322356,
      "step": 19000
    },
    {
      "epoch": 7.001969618055556,
      "eval_loss": 3.399761438369751,
      "eval_runtime": 153.8483,
      "eval_samples_per_second": 119.598,
      "eval_steps_per_second": 7.475,
      "num_input_tokens_seen": 9914322356,
      "step": 19000
    },
    {
      "epoch": 8.010303819444445,
      "grad_norm": 0.16305798292160034,
      "learning_rate": 0.00042784027240358674,
      "loss": 2.8039,
      "num_input_tokens_seen": 10436233756,
      "step": 20000
    },
    {
      "epoch": 8.010303819444445,
      "eval_loss": 3.3940844535827637,
      "eval_runtime": 158.2602,
      "eval_samples_per_second": 116.264,
      "eval_steps_per_second": 7.267,
      "num_input_tokens_seen": 10436233756,
      "step": 20000
    },
    {
      "epoch": 1.0082630208333334,
      "grad_norm": 0.2344750463962555,
      "learning_rate": 0.00038408992486623584,
      "loss": 2.7959,
      "num_input_tokens_seen": 10958156105,
      "step": 21000
    },
    {
      "epoch": 1.0082630208333334,
      "eval_loss": 3.4070310592651367,
      "eval_runtime": 155.4587,
      "eval_samples_per_second": 118.359,
      "eval_steps_per_second": 7.397,
      "num_input_tokens_seen": 10958156105,
      "step": 21000
    },
    {
      "epoch": 2.0165078125,
      "grad_norm": 0.16524210572242737,
      "learning_rate": 0.0003412596353297288,
      "loss": 2.791,
      "num_input_tokens_seen": 11479508780,
      "step": 22000
    },
    {
      "epoch": 2.0165078125,
      "eval_loss": 3.3836262226104736,
      "eval_runtime": 155.5662,
      "eval_samples_per_second": 118.278,
      "eval_steps_per_second": 7.392,
      "num_input_tokens_seen": 11479508780,
      "step": 22000
    },
    {
      "epoch": 4.005475694444445,
      "grad_norm": 0.18112713098526,
      "learning_rate": 0.0002996893772650602,
      "loss": 2.7732,
      "num_input_tokens_seen": 12001335662,
      "step": 23000
    },
    {
      "epoch": 4.005475694444445,
      "eval_loss": 3.374772548675537,
      "eval_runtime": 158.5198,
      "eval_samples_per_second": 116.074,
      "eval_steps_per_second": 7.255,
      "num_input_tokens_seen": 12001335662,
      "step": 23000
    },
    {
      "epoch": 5.013543402777778,
      "grad_norm": 0.20000149309635162,
      "learning_rate": 0.0002597091224066581,
      "loss": 2.776,
      "num_input_tokens_seen": 12523295262,
      "step": 24000
    },
    {
      "epoch": 5.013543402777778,
      "eval_loss": 3.3714466094970703,
      "eval_runtime": 156.2603,
      "eval_samples_per_second": 117.752,
      "eval_steps_per_second": 7.36,
      "num_input_tokens_seen": 12523295262,
      "step": 24000
    },
    {
      "epoch": 1.0082881944444444,
      "grad_norm": 0.17906545102596283,
      "learning_rate": 0.0002216362215397393,
      "loss": 2.7496,
      "num_input_tokens_seen": 13045103310,
      "step": 25000
    },
    {
      "epoch": 1.0082881944444444,
      "eval_loss": 3.3571267127990723,
      "eval_runtime": 156.601,
      "eval_samples_per_second": 117.496,
      "eval_steps_per_second": 7.344,
      "num_input_tokens_seen": 13045103310,
      "step": 25000
    },
    {
      "epoch": 2.0165208333333333,
      "grad_norm": 0.17018218338489532,
      "learning_rate": 0.00018577288546882165,
      "loss": 2.7435,
      "num_input_tokens_seen": 13566410636,
      "step": 26000
    },
    {
      "epoch": 2.0165208333333333,
      "eval_loss": 3.3441579341888428,
      "eval_runtime": 157.974,
      "eval_samples_per_second": 116.475,
      "eval_steps_per_second": 7.28,
      "num_input_tokens_seen": 13566410636,
      "step": 26000
    },
    {
      "epoch": 4.005661458333333,
      "grad_norm": 0.16987864673137665,
      "learning_rate": 0.00015240378616267886,
      "loss": 2.7321,
      "num_input_tokens_seen": 14088234739,
      "step": 27000
    },
    {
      "epoch": 4.005661458333333,
      "eval_loss": 3.335869550704956,
      "eval_runtime": 156.5095,
      "eval_samples_per_second": 117.565,
      "eval_steps_per_second": 7.348,
      "num_input_tokens_seen": 14088234739,
      "step": 27000
    },
    {
      "epoch": 5.013948784722222,
      "grad_norm": 0.18387989699840546,
      "learning_rate": 0.00012179379711709738,
      "loss": 2.7137,
      "num_input_tokens_seen": 14609933027,
      "step": 28000
    },
    {
      "epoch": 5.013948784722222,
      "eval_loss": 3.323387861251831,
      "eval_runtime": 156.9047,
      "eval_samples_per_second": 117.269,
      "eval_steps_per_second": 7.329,
      "num_input_tokens_seen": 14609933027,
      "step": 28000
    },
    {
      "epoch": 7.002777777777778,
      "grad_norm": 0.17292186617851257,
      "learning_rate": 9.418589087173441e-05,
      "loss": 2.7152,
      "num_input_tokens_seen": 15131802301,
      "step": 29000
    },
    {
      "epoch": 7.002777777777778,
      "eval_loss": 3.318844795227051,
      "eval_runtime": 156.7029,
      "eval_samples_per_second": 117.42,
      "eval_steps_per_second": 7.339,
      "num_input_tokens_seen": 15131802301,
      "step": 29000
    },
    {
      "epoch": 8.0111015625,
      "grad_norm": 0.17619270086288452,
      "learning_rate": 6.979921036993042e-05,
      "loss": 2.7042,
      "num_input_tokens_seen": 15653732413,
      "step": 30000
    },
    {
      "epoch": 8.0111015625,
      "eval_loss": 3.3149850368499756,
      "eval_runtime": 157.8074,
      "eval_samples_per_second": 116.598,
      "eval_steps_per_second": 7.287,
      "num_input_tokens_seen": 15653732413,
      "step": 30000
    },
    {
      "epoch": 1.0080494791666668,
      "grad_norm": 0.20021264255046844,
      "learning_rate": 4.882732947041818e-05,
      "loss": 2.7114,
      "num_input_tokens_seen": 16175778125,
      "step": 31000
    },
    {
      "epoch": 1.0080494791666668,
      "eval_loss": 3.3138480186462402,
      "eval_runtime": 155.6732,
      "eval_samples_per_second": 118.196,
      "eval_steps_per_second": 7.387,
      "num_input_tokens_seen": 16175778125,
      "step": 31000
    },
    {
      "epoch": 2.0162526041666666,
      "grad_norm": 0.16703246533870697,
      "learning_rate": 3.143671641844831e-05,
      "loss": 2.6986,
      "num_input_tokens_seen": 16697712121,
      "step": 32000
    },
    {
      "epoch": 2.0162526041666666,
      "eval_loss": 3.3022472858428955,
      "eval_runtime": 156.1087,
      "eval_samples_per_second": 117.867,
      "eval_steps_per_second": 7.367,
      "num_input_tokens_seen": 16697712121,
      "step": 32000
    },
    {
      "epoch": 4.0051015625,
      "grad_norm": 0.16916561126708984,
      "learning_rate": 1.776541247281177e-05,
      "loss": 2.6798,
      "num_input_tokens_seen": 17219624329,
      "step": 33000
    },
    {
      "epoch": 4.0051015625,
      "eval_loss": 3.301964521408081,
      "eval_runtime": 154.9317,
      "eval_samples_per_second": 118.762,
      "eval_steps_per_second": 7.423,
      "num_input_tokens_seen": 17219624329,
      "step": 33000
    },
    {
      "epoch": 5.013414930555555,
      "grad_norm": 0.16186361014842987,
      "learning_rate": 7.921936177411049e-06,
      "loss": 2.6798,
      "num_input_tokens_seen": 17741407609,
      "step": 34000
    },
    {
      "epoch": 5.013414930555555,
      "eval_loss": 3.300182580947876,
      "eval_runtime": 155.5972,
      "eval_samples_per_second": 118.254,
      "eval_steps_per_second": 7.391,
      "num_input_tokens_seen": 17741407609,
      "step": 34000
    },
    {
      "epoch": 7.002003472222222,
      "grad_norm": 0.1644630879163742,
      "learning_rate": 1.984421974927375e-06,
      "loss": 2.7195,
      "num_input_tokens_seen": 18263254841,
      "step": 35000
    },
    {
      "epoch": 7.002003472222222,
      "eval_loss": 3.3019583225250244,
      "eval_runtime": 155.6463,
      "eval_samples_per_second": 118.217,
      "eval_steps_per_second": 7.389,
      "num_input_tokens_seen": 18263254841,
      "step": 35000
    },
    {
      "epoch": 8.010214409722222,
      "grad_norm": 0.1579432338476181,
      "learning_rate": 0.0,
      "loss": 2.683,
      "num_input_tokens_seen": 18785262825,
      "step": 36000
    },
    {
      "epoch": 8.010214409722222,
      "eval_loss": 3.3019728660583496,
      "eval_runtime": 155.7512,
      "eval_samples_per_second": 118.137,
      "eval_steps_per_second": 7.384,
      "num_input_tokens_seen": 18785262825,
      "step": 36000
    }
  ],
  "logging_steps": 1000,
  "max_steps": 36000,
  "num_input_tokens_seen": 18785262825,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 1000,
  "total_flos": 1.2950229157653494e+19,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}