{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 20,
  "global_step": 481,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.04162330905306972,
      "grad_norm": 20.466537475585938,
      "learning_rate": 4.957081545064378e-05,
      "loss": 8.8593,
      "mean_token_accuracy": 0.6828776024281978,
      "num_tokens": 161548.0,
      "step": 20
    },
    {
      "epoch": 0.04162330905306972,
      "eval_loss": 1.1315333843231201,
      "eval_mean_token_accuracy": 0.8237503719329834,
      "eval_num_tokens": 161548.0,
      "eval_runtime": 34.7122,
      "eval_samples_per_second": 2.881,
      "eval_steps_per_second": 0.72,
      "step": 20
    },
    {
      "epoch": 0.08324661810613944,
      "grad_norm": 2.744436025619507,
      "learning_rate": 4.742489270386266e-05,
      "loss": 1.8553,
      "mean_token_accuracy": 0.9392111219465733,
      "num_tokens": 323168.0,
      "step": 40
    },
    {
      "epoch": 0.08324661810613944,
      "eval_loss": 0.16864576935768127,
      "eval_mean_token_accuracy": 0.9810383248329163,
      "eval_num_tokens": 323168.0,
      "eval_runtime": 34.3084,
      "eval_samples_per_second": 2.915,
      "eval_steps_per_second": 0.729,
      "step": 40
    },
    {
      "epoch": 0.12486992715920915,
      "grad_norm": 1.6022675037384033,
      "learning_rate": 4.527896995708155e-05,
      "loss": 0.3903,
      "mean_token_accuracy": 0.9888123281300067,
      "num_tokens": 484780.0,
      "step": 60
    },
    {
      "epoch": 0.12486992715920915,
      "eval_loss": 0.03530249744653702,
      "eval_mean_token_accuracy": 0.9926152086257934,
      "eval_num_tokens": 484780.0,
      "eval_runtime": 34.7758,
      "eval_samples_per_second": 2.876,
      "eval_steps_per_second": 0.719,
      "step": 60
    },
    {
      "epoch": 0.16649323621227888,
      "grad_norm": 1.2044650316238403,
      "learning_rate": 4.313304721030043e-05,
      "loss": 0.0664,
      "mean_token_accuracy": 0.9934882044792175,
      "num_tokens": 646431.0,
      "step": 80
    },
    {
      "epoch": 0.16649323621227888,
      "eval_loss": 0.012011010199785233,
      "eval_mean_token_accuracy": 0.9936227130889893,
      "eval_num_tokens": 646431.0,
      "eval_runtime": 34.3329,
      "eval_samples_per_second": 2.913,
      "eval_steps_per_second": 0.728,
      "step": 80
    },
    {
      "epoch": 0.2081165452653486,
      "grad_norm": 0.4138229191303253,
      "learning_rate": 4.098712446351932e-05,
      "loss": 0.0513,
      "mean_token_accuracy": 0.994041533768177,
      "num_tokens": 808054.0,
      "step": 100
    },
    {
      "epoch": 0.2081165452653486,
      "eval_loss": 0.011702906340360641,
      "eval_mean_token_accuracy": 0.9942703366279602,
      "eval_num_tokens": 808054.0,
      "eval_runtime": 34.3225,
      "eval_samples_per_second": 2.914,
      "eval_steps_per_second": 0.728,
      "step": 100
    },
    {
      "epoch": 0.2497398543184183,
      "grad_norm": 0.8489187955856323,
      "learning_rate": 3.88412017167382e-05,
      "loss": 0.0416,
      "mean_token_accuracy": 0.9943524189293385,
      "num_tokens": 969623.0,
      "step": 120
    },
    {
      "epoch": 0.2497398543184183,
      "eval_loss": 0.010666043497622013,
      "eval_mean_token_accuracy": 0.9943506979942321,
      "eval_num_tokens": 969623.0,
      "eval_runtime": 34.3761,
      "eval_samples_per_second": 2.909,
      "eval_steps_per_second": 0.727,
      "step": 120
    },
    {
      "epoch": 0.29136316337148804,
      "grad_norm": 0.49626249074935913,
      "learning_rate": 3.669527896995708e-05,
      "loss": 0.0384,
      "mean_token_accuracy": 0.9943184182047844,
      "num_tokens": 1131222.0,
      "step": 140
    },
    {
      "epoch": 0.29136316337148804,
      "eval_loss": 0.009807135909795761,
      "eval_mean_token_accuracy": 0.9946736145019531,
      "eval_num_tokens": 1131222.0,
      "eval_runtime": 34.34,
      "eval_samples_per_second": 2.912,
      "eval_steps_per_second": 0.728,
      "step": 140
    },
    {
      "epoch": 0.33298647242455776,
      "grad_norm": 0.5716305375099182,
      "learning_rate": 3.454935622317597e-05,
      "loss": 0.0384,
      "mean_token_accuracy": 0.9946457795798779,
      "num_tokens": 1292839.0,
      "step": 160
    },
    {
      "epoch": 0.33298647242455776,
      "eval_loss": 0.009297176264226437,
      "eval_mean_token_accuracy": 0.9946328043937683,
      "eval_num_tokens": 1292839.0,
      "eval_runtime": 34.557,
      "eval_samples_per_second": 2.894,
      "eval_steps_per_second": 0.723,
      "step": 160
    },
    {
      "epoch": 0.37460978147762747,
      "grad_norm": 0.4264802932739258,
      "learning_rate": 3.240343347639485e-05,
      "loss": 0.0373,
      "mean_token_accuracy": 0.9946053452789784,
      "num_tokens": 1454417.0,
      "step": 180
    },
    {
      "epoch": 0.37460978147762747,
      "eval_loss": 0.009250417351722717,
      "eval_mean_token_accuracy": 0.9949553918838501,
      "eval_num_tokens": 1454417.0,
      "eval_runtime": 34.3425,
      "eval_samples_per_second": 2.912,
      "eval_steps_per_second": 0.728,
      "step": 180
    },
    {
      "epoch": 0.4162330905306972,
      "grad_norm": 0.3315845727920532,
      "learning_rate": 3.0257510729613737e-05,
      "loss": 0.0326,
      "mean_token_accuracy": 0.9949214711785317,
      "num_tokens": 1615982.0,
      "step": 200
    },
    {
      "epoch": 0.4162330905306972,
      "eval_loss": 0.00862209778279066,
      "eval_mean_token_accuracy": 0.994794466495514,
      "eval_num_tokens": 1615982.0,
      "eval_runtime": 34.317,
      "eval_samples_per_second": 2.914,
      "eval_steps_per_second": 0.729,
      "step": 200
    },
    {
      "epoch": 0.4578563995837669,
      "grad_norm": 0.3415542542934418,
      "learning_rate": 2.811158798283262e-05,
      "loss": 0.0328,
      "mean_token_accuracy": 0.9948225237429142,
      "num_tokens": 1777586.0,
      "step": 220
    },
    {
      "epoch": 0.4578563995837669,
      "eval_loss": 0.008552273735404015,
      "eval_mean_token_accuracy": 0.9952386736869812,
      "eval_num_tokens": 1777586.0,
      "eval_runtime": 34.357,
      "eval_samples_per_second": 2.911,
      "eval_steps_per_second": 0.728,
      "step": 220
    },
    {
      "epoch": 0.4994797086368366,
      "grad_norm": 0.5801106691360474,
      "learning_rate": 2.59656652360515e-05,
      "loss": 0.0317,
      "mean_token_accuracy": 0.9949369013309479,
      "num_tokens": 1939204.0,
      "step": 240
    },
    {
      "epoch": 0.4994797086368366,
      "eval_loss": 0.008411003276705742,
      "eval_mean_token_accuracy": 0.9951176619529725,
      "eval_num_tokens": 1939204.0,
      "eval_runtime": 34.3383,
      "eval_samples_per_second": 2.912,
      "eval_steps_per_second": 0.728,
      "step": 240
    },
    {
      "epoch": 0.5411030176899063,
      "grad_norm": 0.3876211941242218,
      "learning_rate": 2.3819742489270388e-05,
      "loss": 0.0322,
      "mean_token_accuracy": 0.9952259331941604,
      "num_tokens": 2100821.0,
      "step": 260
    },
    {
      "epoch": 0.5411030176899063,
      "eval_loss": 0.008060808293521404,
      "eval_mean_token_accuracy": 0.9952384281158447,
      "eval_num_tokens": 2100821.0,
      "eval_runtime": 34.5381,
      "eval_samples_per_second": 2.895,
      "eval_steps_per_second": 0.724,
      "step": 260
    },
    {
      "epoch": 0.5827263267429761,
      "grad_norm": 0.3304857015609741,
      "learning_rate": 2.1673819742489272e-05,
      "loss": 0.0326,
      "mean_token_accuracy": 0.9948949187994003,
      "num_tokens": 2262372.0,
      "step": 280
    },
    {
      "epoch": 0.5827263267429761,
      "eval_loss": 0.008030685596168041,
      "eval_mean_token_accuracy": 0.9953594398498535,
      "eval_num_tokens": 2262372.0,
      "eval_runtime": 34.2589,
      "eval_samples_per_second": 2.919,
      "eval_steps_per_second": 0.73,
      "step": 280
    },
    {
      "epoch": 0.6243496357960457,
      "grad_norm": 0.262820839881897,
      "learning_rate": 1.9527896995708157e-05,
      "loss": 0.0295,
      "mean_token_accuracy": 0.9954247616231442,
      "num_tokens": 2423931.0,
      "step": 300
    },
    {
      "epoch": 0.6243496357960457,
      "eval_loss": 0.0076558589935302734,
      "eval_mean_token_accuracy": 0.9952384281158447,
      "eval_num_tokens": 2423931.0,
      "eval_runtime": 34.2994,
      "eval_samples_per_second": 2.915,
      "eval_steps_per_second": 0.729,
      "step": 300
    },
    {
      "epoch": 0.6659729448491155,
      "grad_norm": 0.4573795199394226,
      "learning_rate": 1.7381974248927038e-05,
      "loss": 0.0308,
      "mean_token_accuracy": 0.9953766994178295,
      "num_tokens": 2585532.0,
      "step": 320
    },
    {
      "epoch": 0.6659729448491155,
      "eval_loss": 0.007896814495325089,
      "eval_mean_token_accuracy": 0.9952785062789917,
      "eval_num_tokens": 2585532.0,
      "eval_runtime": 34.2979,
      "eval_samples_per_second": 2.916,
      "eval_steps_per_second": 0.729,
      "step": 320
    },
    {
      "epoch": 0.7075962539021852,
      "grad_norm": 0.3572976887226105,
      "learning_rate": 1.5236051502145923e-05,
      "loss": 0.0311,
      "mean_token_accuracy": 0.9952577523887157,
      "num_tokens": 2747068.0,
      "step": 340
    },
    {
      "epoch": 0.7075962539021852,
      "eval_loss": 0.007795471698045731,
      "eval_mean_token_accuracy": 0.9952789568901061,
      "eval_num_tokens": 2747068.0,
      "eval_runtime": 34.3024,
      "eval_samples_per_second": 2.915,
      "eval_steps_per_second": 0.729,
      "step": 340
    },
    {
      "epoch": 0.7492195629552549,
      "grad_norm": 0.25351160764694214,
      "learning_rate": 1.3090128755364809e-05,
      "loss": 0.0302,
      "mean_token_accuracy": 0.9953413404524326,
      "num_tokens": 2908697.0,
      "step": 360
    },
    {
      "epoch": 0.7492195629552549,
      "eval_loss": 0.007526129484176636,
      "eval_mean_token_accuracy": 0.9951572942733765,
      "eval_num_tokens": 2908697.0,
      "eval_runtime": 34.4802,
      "eval_samples_per_second": 2.9,
      "eval_steps_per_second": 0.725,
      "step": 360
    },
    {
      "epoch": 0.7908428720083247,
      "grad_norm": 0.45476171374320984,
      "learning_rate": 1.0944206008583692e-05,
      "loss": 0.0301,
      "mean_token_accuracy": 0.9954766884446145,
      "num_tokens": 3070284.0,
      "step": 380
    },
    {
      "epoch": 0.7908428720083247,
      "eval_loss": 0.007462680339813232,
      "eval_mean_token_accuracy": 0.9952787923812866,
      "eval_num_tokens": 3070284.0,
      "eval_runtime": 34.2958,
      "eval_samples_per_second": 2.916,
      "eval_steps_per_second": 0.729,
      "step": 380
    },
    {
      "epoch": 0.8324661810613944,
      "grad_norm": 0.37078818678855896,
      "learning_rate": 8.798283261802575e-06,
      "loss": 0.0298,
      "mean_token_accuracy": 0.9952507764101028,
      "num_tokens": 3231892.0,
      "step": 400
    },
    {
      "epoch": 0.8324661810613944,
      "eval_loss": 0.0073690456338226795,
      "eval_mean_token_accuracy": 0.9953188729286194,
      "eval_num_tokens": 3231892.0,
      "eval_runtime": 34.3243,
      "eval_samples_per_second": 2.913,
      "eval_steps_per_second": 0.728,
      "step": 400
    },
    {
      "epoch": 0.8740894901144641,
      "grad_norm": 0.4091401696205139,
      "learning_rate": 6.65236051502146e-06,
      "loss": 0.0296,
      "mean_token_accuracy": 0.9953283965587616,
      "num_tokens": 3393525.0,
      "step": 420
    },
    {
      "epoch": 0.8740894901144641,
      "eval_loss": 0.007452231831848621,
      "eval_mean_token_accuracy": 0.9953588700294494,
      "eval_num_tokens": 3393525.0,
      "eval_runtime": 34.2775,
      "eval_samples_per_second": 2.917,
      "eval_steps_per_second": 0.729,
      "step": 420
    },
    {
      "epoch": 0.9157127991675338,
      "grad_norm": 0.2830151915550232,
      "learning_rate": 4.5064377682403434e-06,
      "loss": 0.0296,
      "mean_token_accuracy": 0.9951381236314774,
      "num_tokens": 3555132.0,
      "step": 440
    },
    {
      "epoch": 0.9157127991675338,
      "eval_loss": 0.0072263493202626705,
      "eval_mean_token_accuracy": 0.9954402089118958,
      "eval_num_tokens": 3555132.0,
      "eval_runtime": 34.3128,
      "eval_samples_per_second": 2.914,
      "eval_steps_per_second": 0.729,
      "step": 440
    },
    {
      "epoch": 0.9573361082206036,
      "grad_norm": 0.23723173141479492,
      "learning_rate": 2.3605150214592277e-06,
      "loss": 0.0294,
      "mean_token_accuracy": 0.9953089639544487,
      "num_tokens": 3716673.0,
      "step": 460
    },
    {
      "epoch": 0.9573361082206036,
      "eval_loss": 0.007136988453567028,
      "eval_mean_token_accuracy": 0.9953193616867065,
      "eval_num_tokens": 3716673.0,
      "eval_runtime": 34.5004,
      "eval_samples_per_second": 2.899,
      "eval_steps_per_second": 0.725,
      "step": 460
    },
    {
      "epoch": 0.9989594172736732,
      "grad_norm": 0.34731465578079224,
      "learning_rate": 2.145922746781116e-07,
      "loss": 0.0289,
      "mean_token_accuracy": 0.9953534014523029,
      "num_tokens": 3878323.0,
      "step": 480
    },
    {
      "epoch": 0.9989594172736732,
      "eval_loss": 0.007239112630486488,
      "eval_mean_token_accuracy": 0.99535968542099,
      "eval_num_tokens": 3878323.0,
      "eval_runtime": 34.2833,
      "eval_samples_per_second": 2.917,
      "eval_steps_per_second": 0.729,
      "step": 480
    }
  ],
  "logging_steps": 20,
  "max_steps": 481,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 50,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.013575302866206e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}