{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0158730158730158,
  "eval_steps": 500,
  "global_step": 64,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 0.25798711776733396,
      "epoch": 0.07936507936507936,
      "grad_norm": 0.6395628452301025,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.33170533180236816,
      "mean_token_accuracy": 0.9083019256591797,
      "num_tokens": 2249744.0,
      "step": 5
    },
    {
      "entropy": 0.1518770933151245,
      "epoch": 0.15873015873015872,
      "grad_norm": 0.27742329239845276,
      "learning_rate": 4.9966852247120764e-05,
      "loss": 0.14515713453292847,
      "mean_token_accuracy": 0.9476617932319641,
      "num_tokens": 4502357.0,
      "step": 10
    },
    {
      "entropy": 0.1205204889178276,
      "epoch": 0.23809523809523808,
      "grad_norm": 0.15280286967754364,
      "learning_rate": 4.976460088644493e-05,
      "loss": 0.11978392601013184,
      "mean_token_accuracy": 0.9552689790725708,
      "num_tokens": 6750630.0,
      "step": 15
    },
    {
      "entropy": 0.1094709500670433,
      "epoch": 0.31746031746031744,
      "grad_norm": 0.13621626794338226,
      "learning_rate": 4.938000100611456e-05,
      "loss": 0.10746543407440186,
      "mean_token_accuracy": 0.9598726153373718,
      "num_tokens": 8991471.0,
      "step": 20
    },
    {
      "entropy": 0.10419090241193771,
      "epoch": 0.3968253968253968,
      "grad_norm": 0.16320250928401947,
      "learning_rate": 4.881588452008456e-05,
      "loss": 0.10661067962646484,
      "mean_token_accuracy": 0.9599695205688477,
      "num_tokens": 11228489.0,
      "step": 25
    },
    {
      "entropy": 0.09569432139396668,
      "epoch": 0.47619047619047616,
      "grad_norm": 0.15873780846595764,
      "learning_rate": 4.80764051721044e-05,
      "loss": 0.09594419002532958,
      "mean_token_accuracy": 0.9637031435966492,
      "num_tokens": 13464387.0,
      "step": 30
    },
    {
      "entropy": 0.09625398516654968,
      "epoch": 0.5555555555555556,
      "grad_norm": 0.15297284722328186,
      "learning_rate": 4.7167007950568505e-05,
      "loss": 0.09495124220848083,
      "mean_token_accuracy": 0.964220929145813,
      "num_tokens": 15709681.0,
      "step": 35
    },
    {
      "entropy": 0.09195081293582916,
      "epoch": 0.6349206349206349,
      "grad_norm": 0.15219081938266754,
      "learning_rate": 4.609438899557964e-05,
      "loss": 0.09150891304016114,
      "mean_token_accuracy": 0.9652478575706482,
      "num_tokens": 17987403.0,
      "step": 40
    },
    {
      "entropy": 0.08980809897184372,
      "epoch": 0.7142857142857143,
      "grad_norm": 0.1412411779165268,
      "learning_rate": 4.4866446293440626e-05,
      "loss": 0.0898817539215088,
      "mean_token_accuracy": 0.965602707862854,
      "num_tokens": 20231746.0,
      "step": 45
    },
    {
      "entropy": 0.08743101209402085,
      "epoch": 0.7936507936507936,
      "grad_norm": 0.16695629060268402,
      "learning_rate": 4.3492221521624174e-05,
      "loss": 0.08715485334396363,
      "mean_token_accuracy": 0.9668790459632873,
      "num_tokens": 22491762.0,
      "step": 50
    },
    {
      "entropy": 0.08765677958726883,
      "epoch": 0.873015873015873,
      "grad_norm": 0.21143023669719696,
      "learning_rate": 4.198183347243233e-05,
      "loss": 0.08766236305236816,
      "mean_token_accuracy": 0.9668452143669128,
      "num_tokens": 24752715.0,
      "step": 55
    },
    {
      "entropy": 0.08127444982528687,
      "epoch": 0.9523809523809523,
      "grad_norm": 0.18779045343399048,
      "learning_rate": 4.03464035455653e-05,
      "loss": 0.08236484527587891,
      "mean_token_accuracy": 0.9689654350280762,
      "num_tokens": 26982538.0,
      "step": 60
    }
  ],
  "logging_steps": 5,
  "max_steps": 189,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 8,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.604551298047279e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}