{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.3333333333333333,
  "eval_steps": 50,
  "global_step": 600,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.022222222222222223,
      "grad_norm": 0.6193130612373352,
      "learning_rate": 9.910000000000001e-05,
      "loss": 1.654,
      "mean_token_accuracy": 0.6766653224825859,
      "num_tokens": 23330.0,
      "step": 10
    },
    {
      "epoch": 0.044444444444444446,
      "grad_norm": 0.5931444764137268,
      "learning_rate": 9.81e-05,
      "loss": 1.0695,
      "mean_token_accuracy": 0.7898772016167641,
      "num_tokens": 46070.0,
      "step": 20
    },
    {
      "epoch": 0.06666666666666667,
      "grad_norm": 0.3904027044773102,
      "learning_rate": 9.71e-05,
      "loss": 0.7556,
      "mean_token_accuracy": 0.8578989505767822,
      "num_tokens": 68719.0,
      "step": 30
    },
    {
      "epoch": 0.08888888888888889,
      "grad_norm": 0.32560470700263977,
      "learning_rate": 9.61e-05,
      "loss": 0.6186,
      "mean_token_accuracy": 0.8813577085733414,
      "num_tokens": 92126.0,
      "step": 40
    },
    {
      "epoch": 0.1111111111111111,
      "grad_norm": 0.30384671688079834,
      "learning_rate": 9.51e-05,
      "loss": 0.5761,
      "mean_token_accuracy": 0.8921546742320061,
      "num_tokens": 114738.0,
      "step": 50
    },
    {
      "epoch": 0.1111111111111111,
      "eval_loss": 0.5330827236175537,
      "eval_mean_token_accuracy": 0.9009262701869011,
      "eval_num_tokens": 114738.0,
      "eval_runtime": 40.5094,
      "eval_samples_per_second": 4.937,
      "eval_steps_per_second": 4.937,
      "step": 50
    },
    {
      "epoch": 0.13333333333333333,
      "grad_norm": 0.40981799364089966,
      "learning_rate": 9.41e-05,
      "loss": 0.4978,
      "mean_token_accuracy": 0.9065196111798286,
      "num_tokens": 136310.0,
      "step": 60
    },
    {
      "epoch": 0.15555555555555556,
      "grad_norm": 0.5167299509048462,
      "learning_rate": 9.310000000000001e-05,
      "loss": 0.4466,
      "mean_token_accuracy": 0.9203566908836365,
      "num_tokens": 158193.0,
      "step": 70
    },
    {
      "epoch": 0.17777777777777778,
      "grad_norm": 0.5062010884284973,
      "learning_rate": 9.21e-05,
      "loss": 0.4003,
      "mean_token_accuracy": 0.9284946233034134,
      "num_tokens": 180767.0,
      "step": 80
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.38172802329063416,
      "learning_rate": 9.11e-05,
      "loss": 0.3457,
      "mean_token_accuracy": 0.9404646947979927,
      "num_tokens": 203203.0,
      "step": 90
    },
    {
      "epoch": 0.2222222222222222,
      "grad_norm": 0.4985530376434326,
      "learning_rate": 9.010000000000001e-05,
      "loss": 0.3271,
      "mean_token_accuracy": 0.9434564337134361,
      "num_tokens": 226070.0,
      "step": 100
    },
    {
      "epoch": 0.2222222222222222,
      "eval_loss": 0.3216829001903534,
      "eval_mean_token_accuracy": 0.9481457704305649,
      "eval_num_tokens": 226070.0,
      "eval_runtime": 41.0245,
      "eval_samples_per_second": 4.875,
      "eval_steps_per_second": 4.875,
      "step": 100
    },
    {
      "epoch": 0.24444444444444444,
      "grad_norm": 0.3317733108997345,
      "learning_rate": 8.910000000000001e-05,
      "loss": 0.2834,
      "mean_token_accuracy": 0.9527689620852471,
      "num_tokens": 251659.0,
      "step": 110
    },
    {
      "epoch": 0.26666666666666666,
      "grad_norm": 0.28577539324760437,
      "learning_rate": 8.81e-05,
      "loss": 0.262,
      "mean_token_accuracy": 0.9557135447859764,
      "num_tokens": 276141.0,
      "step": 120
    },
    {
      "epoch": 0.28888888888888886,
      "grad_norm": 0.4030509293079376,
      "learning_rate": 8.71e-05,
      "loss": 0.26,
      "mean_token_accuracy": 0.9594775453209877,
      "num_tokens": 299250.0,
      "step": 130
    },
    {
      "epoch": 0.3111111111111111,
      "grad_norm": 0.4423465132713318,
      "learning_rate": 8.61e-05,
      "loss": 0.2557,
      "mean_token_accuracy": 0.9583914607763291,
      "num_tokens": 323528.0,
      "step": 140
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 0.49233391880989075,
      "learning_rate": 8.510000000000001e-05,
      "loss": 0.2444,
      "mean_token_accuracy": 0.9617681756615639,
      "num_tokens": 346496.0,
      "step": 150
    },
    {
      "epoch": 0.3333333333333333,
      "eval_loss": 0.24680288136005402,
      "eval_mean_token_accuracy": 0.963321838080883,
      "eval_num_tokens": 346496.0,
      "eval_runtime": 41.1525,
      "eval_samples_per_second": 4.86,
      "eval_steps_per_second": 4.86,
      "step": 150
    },
    {
      "epoch": 0.35555555555555557,
      "grad_norm": 0.2696438133716583,
      "learning_rate": 8.41e-05,
      "loss": 0.2196,
      "mean_token_accuracy": 0.9656250089406967,
      "num_tokens": 371258.0,
      "step": 160
    },
    {
      "epoch": 0.37777777777777777,
      "grad_norm": 0.525949239730835,
      "learning_rate": 8.31e-05,
      "loss": 0.2122,
      "mean_token_accuracy": 0.9669450834393502,
      "num_tokens": 395722.0,
      "step": 170
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.38804805278778076,
      "learning_rate": 8.21e-05,
      "loss": 0.2097,
      "mean_token_accuracy": 0.9699977666139603,
      "num_tokens": 418830.0,
      "step": 180
    },
    {
      "epoch": 0.4222222222222222,
      "grad_norm": 0.38709455728530884,
      "learning_rate": 8.11e-05,
      "loss": 0.1914,
      "mean_token_accuracy": 0.9692254871129989,
      "num_tokens": 443015.0,
      "step": 190
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 0.28362640738487244,
      "learning_rate": 8.010000000000001e-05,
      "loss": 0.2051,
      "mean_token_accuracy": 0.968667496740818,
      "num_tokens": 465941.0,
      "step": 200
    },
    {
      "epoch": 0.4444444444444444,
      "eval_loss": 0.21475830674171448,
      "eval_mean_token_accuracy": 0.9686776822805405,
      "eval_num_tokens": 465941.0,
      "eval_runtime": 40.8312,
      "eval_samples_per_second": 4.898,
      "eval_steps_per_second": 4.898,
      "step": 200
    },
    {
      "epoch": 0.4666666666666667,
      "grad_norm": 0.17446358501911163,
      "learning_rate": 7.910000000000001e-05,
      "loss": 0.1992,
      "mean_token_accuracy": 0.9691774398088455,
      "num_tokens": 489869.0,
      "step": 210
    },
    {
      "epoch": 0.4888888888888889,
      "grad_norm": 0.47806766629219055,
      "learning_rate": 7.81e-05,
      "loss": 0.18,
      "mean_token_accuracy": 0.9718323066830635,
      "num_tokens": 514551.0,
      "step": 220
    },
    {
      "epoch": 0.5111111111111111,
      "grad_norm": 0.4450798034667969,
      "learning_rate": 7.71e-05,
      "loss": 0.1818,
      "mean_token_accuracy": 0.9731900364160537,
      "num_tokens": 539497.0,
      "step": 230
    },
    {
      "epoch": 0.5333333333333333,
      "grad_norm": 0.3654688596725464,
      "learning_rate": 7.61e-05,
      "loss": 0.2072,
      "mean_token_accuracy": 0.9681414648890495,
      "num_tokens": 560454.0,
      "step": 240
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 0.26568618416786194,
      "learning_rate": 7.510000000000001e-05,
      "loss": 0.1858,
      "mean_token_accuracy": 0.9720202058553695,
      "num_tokens": 584595.0,
      "step": 250
    },
    {
      "epoch": 0.5555555555555556,
      "eval_loss": 0.19870425760746002,
      "eval_mean_token_accuracy": 0.9720033320784569,
      "eval_num_tokens": 584595.0,
      "eval_runtime": 40.9104,
      "eval_samples_per_second": 4.889,
      "eval_steps_per_second": 4.889,
      "step": 250
    },
    {
      "epoch": 0.5777777777777777,
      "grad_norm": 0.21654823422431946,
      "learning_rate": 7.41e-05,
      "loss": 0.1765,
      "mean_token_accuracy": 0.9739127770066262,
      "num_tokens": 608352.0,
      "step": 260
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.3529929220676422,
      "learning_rate": 7.31e-05,
      "loss": 0.1726,
      "mean_token_accuracy": 0.9745927527546883,
      "num_tokens": 632550.0,
      "step": 270
    },
    {
      "epoch": 0.6222222222222222,
      "grad_norm": 0.2798812985420227,
      "learning_rate": 7.21e-05,
      "loss": 0.1814,
      "mean_token_accuracy": 0.9733540549874306,
      "num_tokens": 655957.0,
      "step": 280
    },
    {
      "epoch": 0.6444444444444445,
      "grad_norm": 0.28510963916778564,
      "learning_rate": 7.11e-05,
      "loss": 0.1789,
      "mean_token_accuracy": 0.9735615655779839,
      "num_tokens": 680743.0,
      "step": 290
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.22315602004528046,
      "learning_rate": 7.01e-05,
      "loss": 0.1777,
      "mean_token_accuracy": 0.9739167138934135,
      "num_tokens": 704844.0,
      "step": 300
    },
    {
      "epoch": 0.6666666666666666,
      "eval_loss": 0.18998412787914276,
      "eval_mean_token_accuracy": 0.9746258324384689,
      "eval_num_tokens": 704844.0,
      "eval_runtime": 40.6816,
      "eval_samples_per_second": 4.916,
      "eval_steps_per_second": 4.916,
      "step": 300
    },
    {
      "epoch": 0.6888888888888889,
      "grad_norm": 0.25007760524749756,
      "learning_rate": 6.91e-05,
      "loss": 0.1913,
      "mean_token_accuracy": 0.9710126295685768,
      "num_tokens": 724403.0,
      "step": 310
    },
    {
      "epoch": 0.7111111111111111,
      "grad_norm": 0.19620826840400696,
      "learning_rate": 6.81e-05,
      "loss": 0.1782,
      "mean_token_accuracy": 0.9740939602255821,
      "num_tokens": 748460.0,
      "step": 320
    },
    {
      "epoch": 0.7333333333333333,
      "grad_norm": 0.2543444037437439,
      "learning_rate": 6.71e-05,
      "loss": 0.1825,
      "mean_token_accuracy": 0.972811259329319,
      "num_tokens": 769992.0,
      "step": 330
    },
    {
      "epoch": 0.7555555555555555,
      "grad_norm": 0.12709860503673553,
      "learning_rate": 6.610000000000001e-05,
      "loss": 0.1828,
      "mean_token_accuracy": 0.9728233471512795,
      "num_tokens": 792260.0,
      "step": 340
    },
    {
      "epoch": 0.7777777777777778,
      "grad_norm": 0.24585242569446564,
      "learning_rate": 6.510000000000001e-05,
      "loss": 0.1798,
      "mean_token_accuracy": 0.9745096892118454,
      "num_tokens": 815593.0,
      "step": 350
    },
    {
      "epoch": 0.7777777777777778,
      "eval_loss": 0.18533459305763245,
      "eval_mean_token_accuracy": 0.9745631077885628,
      "eval_num_tokens": 815593.0,
      "eval_runtime": 40.662,
      "eval_samples_per_second": 4.919,
      "eval_steps_per_second": 4.919,
      "step": 350
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.22581206262111664,
      "learning_rate": 6.41e-05,
      "loss": 0.1807,
      "mean_token_accuracy": 0.9736085414886475,
      "num_tokens": 837982.0,
      "step": 360
    },
    {
      "epoch": 0.8222222222222222,
      "grad_norm": 0.2912260591983795,
      "learning_rate": 6.31e-05,
      "loss": 0.1668,
      "mean_token_accuracy": 0.9759353816509246,
      "num_tokens": 860757.0,
      "step": 370
    },
    {
      "epoch": 0.8444444444444444,
      "grad_norm": 0.1845020055770874,
      "learning_rate": 6.21e-05,
      "loss": 0.1839,
      "mean_token_accuracy": 0.9732057675719261,
      "num_tokens": 881639.0,
      "step": 380
    },
    {
      "epoch": 0.8666666666666667,
      "grad_norm": 0.2601327896118164,
      "learning_rate": 6.110000000000001e-05,
      "loss": 0.1652,
      "mean_token_accuracy": 0.9748824626207352,
      "num_tokens": 905643.0,
      "step": 390
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 0.16098515689373016,
      "learning_rate": 6.0100000000000004e-05,
      "loss": 0.171,
      "mean_token_accuracy": 0.9756452158093453,
      "num_tokens": 929665.0,
      "step": 400
    },
    {
      "epoch": 0.8888888888888888,
      "eval_loss": 0.18200141191482544,
      "eval_mean_token_accuracy": 0.9752507287263871,
      "eval_num_tokens": 929665.0,
      "eval_runtime": 40.698,
      "eval_samples_per_second": 4.914,
      "eval_steps_per_second": 4.914,
      "step": 400
    },
    {
      "epoch": 0.9111111111111111,
      "grad_norm": 0.16805872321128845,
      "learning_rate": 5.91e-05,
      "loss": 0.1718,
      "mean_token_accuracy": 0.9748784646391868,
      "num_tokens": 951866.0,
      "step": 410
    },
    {
      "epoch": 0.9333333333333333,
      "grad_norm": 0.14501382410526276,
      "learning_rate": 5.8099999999999996e-05,
      "loss": 0.1662,
      "mean_token_accuracy": 0.9753882080316544,
      "num_tokens": 975946.0,
      "step": 420
    },
    {
      "epoch": 0.9555555555555556,
      "grad_norm": 0.17429794371128082,
      "learning_rate": 5.71e-05,
      "loss": 0.1719,
      "mean_token_accuracy": 0.9735226526856422,
      "num_tokens": 997417.0,
      "step": 430
    },
    {
      "epoch": 0.9777777777777777,
      "grad_norm": 0.21708282828330994,
      "learning_rate": 5.610000000000001e-05,
      "loss": 0.1679,
      "mean_token_accuracy": 0.9750987365841866,
      "num_tokens": 1021762.0,
      "step": 440
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.11724573373794556,
      "learning_rate": 5.5100000000000004e-05,
      "loss": 0.1648,
      "mean_token_accuracy": 0.9775774568319321,
      "num_tokens": 1046469.0,
      "step": 450
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.18049028515815735,
      "eval_mean_token_accuracy": 0.9776099815964698,
      "eval_num_tokens": 1046469.0,
      "eval_runtime": 40.8205,
      "eval_samples_per_second": 4.899,
      "eval_steps_per_second": 4.899,
      "step": 450
    },
    {
      "epoch": 1.0222222222222221,
      "grad_norm": 0.13565418124198914,
      "learning_rate": 5.410000000000001e-05,
      "loss": 0.161,
      "mean_token_accuracy": 0.9767818033695221,
      "num_tokens": 1070354.0,
      "step": 460
    },
    {
      "epoch": 1.0444444444444445,
      "grad_norm": 0.14690855145454407,
      "learning_rate": 5.31e-05,
      "loss": 0.1616,
      "mean_token_accuracy": 0.9762020215392113,
      "num_tokens": 1095160.0,
      "step": 470
    },
    {
      "epoch": 1.0666666666666667,
      "grad_norm": 0.17638139426708221,
      "learning_rate": 5.2100000000000006e-05,
      "loss": 0.1633,
      "mean_token_accuracy": 0.9768583804368973,
      "num_tokens": 1118117.0,
      "step": 480
    },
    {
      "epoch": 1.0888888888888888,
      "grad_norm": 0.18202361464500427,
      "learning_rate": 5.11e-05,
      "loss": 0.165,
      "mean_token_accuracy": 0.9771157950162888,
      "num_tokens": 1140967.0,
      "step": 490
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 0.1454939991235733,
      "learning_rate": 5.0100000000000005e-05,
      "loss": 0.1647,
      "mean_token_accuracy": 0.9771055221557617,
      "num_tokens": 1165040.0,
      "step": 500
    },
    {
      "epoch": 1.1111111111111112,
      "eval_loss": 0.17858880758285522,
      "eval_mean_token_accuracy": 0.9758283907175064,
      "eval_num_tokens": 1165040.0,
      "eval_runtime": 40.5796,
      "eval_samples_per_second": 4.929,
      "eval_steps_per_second": 4.929,
      "step": 500
    },
    {
      "epoch": 1.1333333333333333,
      "grad_norm": 0.18715986609458923,
      "learning_rate": 4.91e-05,
      "loss": 0.1647,
      "mean_token_accuracy": 0.9763637855648994,
      "num_tokens": 1188748.0,
      "step": 510
    },
    {
      "epoch": 1.1555555555555554,
      "grad_norm": 0.1779111623764038,
      "learning_rate": 4.8100000000000004e-05,
      "loss": 0.1696,
      "mean_token_accuracy": 0.9764641597867012,
      "num_tokens": 1211411.0,
      "step": 520
    },
    {
      "epoch": 1.1777777777777778,
      "grad_norm": 0.19597475230693817,
      "learning_rate": 4.71e-05,
      "loss": 0.1707,
      "mean_token_accuracy": 0.976306001842022,
      "num_tokens": 1233545.0,
      "step": 530
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.15781964361667633,
      "learning_rate": 4.61e-05,
      "loss": 0.1617,
      "mean_token_accuracy": 0.9768658638000488,
      "num_tokens": 1257597.0,
      "step": 540
    },
    {
      "epoch": 1.2222222222222223,
      "grad_norm": 0.16846869885921478,
      "learning_rate": 4.5100000000000005e-05,
      "loss": 0.1643,
      "mean_token_accuracy": 0.9779975593090058,
      "num_tokens": 1280857.0,
      "step": 550
    },
    {
      "epoch": 1.2222222222222223,
      "eval_loss": 0.17742373049259186,
      "eval_mean_token_accuracy": 0.975758826136589,
      "eval_num_tokens": 1280857.0,
      "eval_runtime": 40.7021,
      "eval_samples_per_second": 4.914,
      "eval_steps_per_second": 4.914,
      "step": 550
    },
    {
      "epoch": 1.2444444444444445,
      "grad_norm": 0.09051565825939178,
      "learning_rate": 4.41e-05,
      "loss": 0.1672,
      "mean_token_accuracy": 0.9775181457400322,
      "num_tokens": 1304923.0,
      "step": 560
    },
    {
      "epoch": 1.2666666666666666,
      "grad_norm": 0.12825968861579895,
      "learning_rate": 4.3100000000000004e-05,
      "loss": 0.1709,
      "mean_token_accuracy": 0.976058243215084,
      "num_tokens": 1326584.0,
      "step": 570
    },
    {
      "epoch": 1.2888888888888888,
      "grad_norm": 0.11256464570760727,
      "learning_rate": 4.21e-05,
      "loss": 0.1566,
      "mean_token_accuracy": 0.9770786449313164,
      "num_tokens": 1351586.0,
      "step": 580
    },
    {
      "epoch": 1.3111111111111111,
      "grad_norm": 0.19633373618125916,
      "learning_rate": 4.11e-05,
      "loss": 0.1815,
      "mean_token_accuracy": 0.9755366548895836,
      "num_tokens": 1372532.0,
      "step": 590
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.1573602855205536,
      "learning_rate": 4.0100000000000006e-05,
      "loss": 0.1666,
      "mean_token_accuracy": 0.9768186181783676,
      "num_tokens": 1395159.0,
      "step": 600
    },
    {
      "epoch": 1.3333333333333333,
      "eval_loss": 0.17687298357486725,
      "eval_mean_token_accuracy": 0.9778035506606102,
      "eval_num_tokens": 1395159.0,
      "eval_runtime": 40.6718,
      "eval_samples_per_second": 4.917,
      "eval_steps_per_second": 4.917,
      "step": 600
    }
  ],
  "logging_steps": 10,
  "max_steps": 1000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 6751944528768000.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}