{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 333,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 1.7490966856479644,
      "epoch": 0.09009009009009009,
      "grad_norm": 4.318613052368164,
      "learning_rate": 9e-05,
      "loss": 6.8516,
      "mean_token_accuracy": 0.31817560344934465,
      "num_tokens": 5273.0,
      "step": 10
    },
    {
      "entropy": 2.358074116706848,
      "epoch": 0.18018018018018017,
      "grad_norm": 2.591895341873169,
      "learning_rate": 0.00019,
      "loss": 4.7791,
      "mean_token_accuracy": 0.4064402043819427,
      "num_tokens": 10323.0,
      "step": 20
    },
    {
      "entropy": 2.7442060232162477,
      "epoch": 0.2702702702702703,
      "grad_norm": 2.085822105407715,
      "learning_rate": 0.00019424920127795528,
      "loss": 2.528,
      "mean_token_accuracy": 0.6114133253693581,
      "num_tokens": 15657.0,
      "step": 30
    },
    {
      "entropy": 1.6624038934707641,
      "epoch": 0.36036036036036034,
      "grad_norm": 1.898437261581421,
      "learning_rate": 0.0001878594249201278,
      "loss": 1.5967,
      "mean_token_accuracy": 0.75713110268116,
      "num_tokens": 20566.0,
      "step": 40
    },
    {
      "entropy": 1.3454128205776215,
      "epoch": 0.45045045045045046,
      "grad_norm": 2.2532074451446533,
      "learning_rate": 0.00018146964856230032,
      "loss": 1.3224,
      "mean_token_accuracy": 0.7765294492244721,
      "num_tokens": 25625.0,
      "step": 50
    },
    {
      "entropy": 1.1165625154972076,
      "epoch": 0.5405405405405406,
      "grad_norm": 2.300675868988037,
      "learning_rate": 0.00017507987220447287,
      "loss": 1.0734,
      "mean_token_accuracy": 0.803844365477562,
      "num_tokens": 30624.0,
      "step": 60
    },
    {
      "entropy": 0.975060197710991,
      "epoch": 0.6306306306306306,
      "grad_norm": 1.954689383506775,
      "learning_rate": 0.00016869009584664536,
      "loss": 0.9733,
      "mean_token_accuracy": 0.818100169301033,
      "num_tokens": 35746.0,
      "step": 70
    },
    {
      "entropy": 0.8530668556690216,
      "epoch": 0.7207207207207207,
      "grad_norm": 2.083409309387207,
      "learning_rate": 0.0001623003194888179,
      "loss": 0.8131,
      "mean_token_accuracy": 0.8403199791908265,
      "num_tokens": 41031.0,
      "step": 80
    },
    {
      "entropy": 0.8242652952671051,
      "epoch": 0.8108108108108109,
      "grad_norm": 2.1543054580688477,
      "learning_rate": 0.00015591054313099042,
      "loss": 0.7285,
      "mean_token_accuracy": 0.8522787302732467,
      "num_tokens": 45910.0,
      "step": 90
    },
    {
      "entropy": 0.7216496258974076,
      "epoch": 0.9009009009009009,
      "grad_norm": 1.8068711757659912,
      "learning_rate": 0.00014952076677316297,
      "loss": 0.632,
      "mean_token_accuracy": 0.8595692455768585,
      "num_tokens": 51027.0,
      "step": 100
    },
    {
      "entropy": 0.6904670834541321,
      "epoch": 0.990990990990991,
      "grad_norm": 2.6211395263671875,
      "learning_rate": 0.00014313099041533546,
      "loss": 0.6219,
      "mean_token_accuracy": 0.8704036623239517,
      "num_tokens": 55880.0,
      "step": 110
    },
    {
      "entropy": 0.5557049199938774,
      "epoch": 1.0810810810810811,
      "grad_norm": 2.952716112136841,
      "learning_rate": 0.000136741214057508,
      "loss": 0.5041,
      "mean_token_accuracy": 0.8821422606706619,
      "num_tokens": 60665.0,
      "step": 120
    },
    {
      "entropy": 0.4335968837141991,
      "epoch": 1.1711711711711712,
      "grad_norm": 1.972198724746704,
      "learning_rate": 0.00013035143769968052,
      "loss": 0.4319,
      "mean_token_accuracy": 0.8849613904953003,
      "num_tokens": 65647.0,
      "step": 130
    },
    {
      "entropy": 0.44969954043626786,
      "epoch": 1.2612612612612613,
      "grad_norm": 1.9271284341812134,
      "learning_rate": 0.00012396166134185304,
      "loss": 0.4252,
      "mean_token_accuracy": 0.8931547611951828,
      "num_tokens": 70510.0,
      "step": 140
    },
    {
      "entropy": 0.40625376254320145,
      "epoch": 1.3513513513513513,
      "grad_norm": 2.1082940101623535,
      "learning_rate": 0.00011757188498402556,
      "loss": 0.4017,
      "mean_token_accuracy": 0.8913865387439728,
      "num_tokens": 75632.0,
      "step": 150
    },
    {
      "entropy": 0.45655421912670135,
      "epoch": 1.4414414414414414,
      "grad_norm": 1.5840785503387451,
      "learning_rate": 0.00011118210862619809,
      "loss": 0.4614,
      "mean_token_accuracy": 0.8811476260423661,
      "num_tokens": 80976.0,
      "step": 160
    },
    {
      "entropy": 0.42031795233488084,
      "epoch": 1.5315315315315314,
      "grad_norm": 1.667891502380371,
      "learning_rate": 0.00010479233226837062,
      "loss": 0.395,
      "mean_token_accuracy": 0.894879949092865,
      "num_tokens": 86111.0,
      "step": 170
    },
    {
      "entropy": 0.42292200922966006,
      "epoch": 1.6216216216216215,
      "grad_norm": 1.7975671291351318,
      "learning_rate": 9.840255591054314e-05,
      "loss": 0.4021,
      "mean_token_accuracy": 0.8952047973871231,
      "num_tokens": 91226.0,
      "step": 180
    },
    {
      "entropy": 0.3855786919593811,
      "epoch": 1.7117117117117115,
      "grad_norm": 2.244504690170288,
      "learning_rate": 9.201277955271566e-05,
      "loss": 0.387,
      "mean_token_accuracy": 0.8959153234958649,
      "num_tokens": 96405.0,
      "step": 190
    },
    {
      "entropy": 0.4205052852630615,
      "epoch": 1.8018018018018018,
      "grad_norm": 1.7068063020706177,
      "learning_rate": 8.562300319488819e-05,
      "loss": 0.419,
      "mean_token_accuracy": 0.888062196969986,
      "num_tokens": 101805.0,
      "step": 200
    },
    {
      "entropy": 0.37721182107925416,
      "epoch": 1.8918918918918919,
      "grad_norm": 1.7175687551498413,
      "learning_rate": 7.923322683706071e-05,
      "loss": 0.342,
      "mean_token_accuracy": 0.9077515214681625,
      "num_tokens": 106683.0,
      "step": 210
    },
    {
      "entropy": 0.33771649897098543,
      "epoch": 1.981981981981982,
      "grad_norm": 1.5167031288146973,
      "learning_rate": 7.284345047923323e-05,
      "loss": 0.3184,
      "mean_token_accuracy": 0.9064931035041809,
      "num_tokens": 111616.0,
      "step": 220
    },
    {
      "entropy": 0.3555280163884163,
      "epoch": 2.0720720720720722,
      "grad_norm": 1.660430908203125,
      "learning_rate": 6.645367412140575e-05,
      "loss": 0.3216,
      "mean_token_accuracy": 0.9094264894723892,
      "num_tokens": 116395.0,
      "step": 230
    },
    {
      "entropy": 0.33757986277341845,
      "epoch": 2.1621621621621623,
      "grad_norm": 1.3866914510726929,
      "learning_rate": 6.006389776357828e-05,
      "loss": 0.3271,
      "mean_token_accuracy": 0.9032522082328797,
      "num_tokens": 121543.0,
      "step": 240
    },
    {
      "entropy": 0.3460941515862942,
      "epoch": 2.2522522522522523,
      "grad_norm": 1.5430920124053955,
      "learning_rate": 5.36741214057508e-05,
      "loss": 0.3345,
      "mean_token_accuracy": 0.9088279217481613,
      "num_tokens": 126607.0,
      "step": 250
    },
    {
      "entropy": 0.30904691815376284,
      "epoch": 2.3423423423423424,
      "grad_norm": 1.9197715520858765,
      "learning_rate": 4.728434504792332e-05,
      "loss": 0.2831,
      "mean_token_accuracy": 0.9148383587598801,
      "num_tokens": 131589.0,
      "step": 260
    },
    {
      "entropy": 0.29833986386656763,
      "epoch": 2.4324324324324325,
      "grad_norm": 2.148452043533325,
      "learning_rate": 4.089456869009585e-05,
      "loss": 0.2785,
      "mean_token_accuracy": 0.9198061287403106,
      "num_tokens": 136431.0,
      "step": 270
    },
    {
      "entropy": 0.30950267389416697,
      "epoch": 2.5225225225225225,
      "grad_norm": 1.3571665287017822,
      "learning_rate": 3.450479233226837e-05,
      "loss": 0.2827,
      "mean_token_accuracy": 0.9169433653354645,
      "num_tokens": 141739.0,
      "step": 280
    },
    {
      "entropy": 0.27893091589212415,
      "epoch": 2.6126126126126126,
      "grad_norm": 1.4515094757080078,
      "learning_rate": 2.8115015974440894e-05,
      "loss": 0.2708,
      "mean_token_accuracy": 0.9156843811273575,
      "num_tokens": 146814.0,
      "step": 290
    },
    {
      "entropy": 0.3063665457069874,
      "epoch": 2.7027027027027026,
      "grad_norm": 2.0274174213409424,
      "learning_rate": 2.172523961661342e-05,
      "loss": 0.2873,
      "mean_token_accuracy": 0.9122885495424271,
      "num_tokens": 152037.0,
      "step": 300
    },
    {
      "entropy": 0.3225431203842163,
      "epoch": 2.7927927927927927,
      "grad_norm": 1.8013384342193604,
      "learning_rate": 1.533546325878594e-05,
      "loss": 0.2961,
      "mean_token_accuracy": 0.9138898193836212,
      "num_tokens": 157245.0,
      "step": 310
    },
    {
      "entropy": 0.2981728859245777,
      "epoch": 2.8828828828828827,
      "grad_norm": 1.7200424671173096,
      "learning_rate": 8.945686900958466e-06,
      "loss": 0.2758,
      "mean_token_accuracy": 0.9152151584625244,
      "num_tokens": 162185.0,
      "step": 320
    },
    {
      "entropy": 0.2874752961099148,
      "epoch": 2.972972972972973,
      "grad_norm": 1.8970165252685547,
      "learning_rate": 2.5559105431309904e-06,
      "loss": 0.2718,
      "mean_token_accuracy": 0.9175638914108276,
      "num_tokens": 167004.0,
      "step": 330
    }
  ],
  "logging_steps": 10,
  "max_steps": 333,
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1848157889673216.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}