{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 108,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 1.180859375,
      "epoch": 0.37037037037037035,
      "grad_norm": 1.9087318181991577,
      "learning_rate": 0.00019886154122075343,
      "loss": 1.3855,
      "mean_token_accuracy": 0.709491191804409,
      "num_tokens": 20480.0,
      "step": 10
    },
    {
      "entropy": 0.77197265625,
      "epoch": 0.7407407407407407,
      "grad_norm": 1.5744595527648926,
      "learning_rate": 0.00018990881137654258,
      "loss": 0.8126,
      "mean_token_accuracy": 0.8044520571827889,
      "num_tokens": 40960.0,
      "step": 20
    },
    {
      "epoch": 1.0,
      "eval_entropy": 0.666015625,
      "eval_loss": 0.664973795413971,
      "eval_mean_token_accuracy": 0.8258317112922668,
      "eval_num_tokens": 55296.0,
      "eval_runtime": 104.8748,
      "eval_samples_per_second": 0.114,
      "eval_steps_per_second": 0.019,
      "step": 27
    },
    {
      "entropy": 0.67705078125,
      "epoch": 1.1111111111111112,
      "grad_norm": 1.3851813077926636,
      "learning_rate": 0.00017281409538757883,
      "loss": 0.6667,
      "mean_token_accuracy": 0.8270058631896973,
      "num_tokens": 61440.0,
      "step": 30
    },
    {
      "entropy": 0.3906982421875,
      "epoch": 1.4814814814814814,
      "grad_norm": 1.8623592853546143,
      "learning_rate": 0.00014912546110838775,
      "loss": 0.3976,
      "mean_token_accuracy": 0.8884540066123009,
      "num_tokens": 81920.0,
      "step": 40
    },
    {
      "entropy": 0.4400634765625,
      "epoch": 1.8518518518518519,
      "grad_norm": 1.3342927694320679,
      "learning_rate": 0.00012098811020648475,
      "loss": 0.4118,
      "mean_token_accuracy": 0.8866927579045296,
      "num_tokens": 102400.0,
      "step": 50
    },
    {
      "epoch": 2.0,
      "eval_entropy": 0.447265625,
      "eval_loss": 0.594461977481842,
      "eval_mean_token_accuracy": 0.8510273993015289,
      "eval_num_tokens": 110592.0,
      "eval_runtime": 158.7424,
      "eval_samples_per_second": 0.076,
      "eval_steps_per_second": 0.013,
      "step": 54
    },
    {
      "entropy": 0.3321044921875,
      "epoch": 2.2222222222222223,
      "grad_norm": 1.1872409582138062,
      "learning_rate": 9.095011241703623e-05,
      "loss": 0.2912,
      "mean_token_accuracy": 0.9179549887776375,
      "num_tokens": 122880.0,
      "step": 60
    },
    {
      "entropy": 0.24451904296875,
      "epoch": 2.5925925925925926,
      "grad_norm": 1.2892343997955322,
      "learning_rate": 6.173165676349103e-05,
      "loss": 0.237,
      "mean_token_accuracy": 0.9271037086844445,
      "num_tokens": 143360.0,
      "step": 70
    },
    {
      "entropy": 0.24892578125,
      "epoch": 2.962962962962963,
      "grad_norm": 1.2926205396652222,
      "learning_rate": 3.597871595375121e-05,
      "loss": 0.2154,
      "mean_token_accuracy": 0.9361056685447693,
      "num_tokens": 163840.0,
      "step": 80
    },
    {
      "epoch": 3.0,
      "eval_entropy": 0.3505859375,
      "eval_loss": 0.6058798432350159,
      "eval_mean_token_accuracy": 0.852372795343399,
      "eval_num_tokens": 165888.0,
      "eval_runtime": 453.1986,
      "eval_samples_per_second": 0.026,
      "eval_steps_per_second": 0.004,
      "step": 81
    },
    {
      "entropy": 0.18564453125,
      "epoch": 3.3333333333333335,
      "grad_norm": 0.9748417735099792,
      "learning_rate": 1.6023431677260214e-05,
      "loss": 0.1381,
      "mean_token_accuracy": 0.9602250471711159,
      "num_tokens": 184320.0,
      "step": 90
    },
    {
      "entropy": 0.15009765625,
      "epoch": 3.7037037037037037,
      "grad_norm": 0.9260263442993164,
      "learning_rate": 3.6729198952483724e-06,
      "loss": 0.1271,
      "mean_token_accuracy": 0.9631604671478271,
      "num_tokens": 204800.0,
      "step": 100
    },
    {
      "epoch": 4.0,
      "eval_entropy": 0.29296875,
      "eval_loss": 0.6591059565544128,
      "eval_mean_token_accuracy": 0.8539628088474274,
      "eval_num_tokens": 221184.0,
      "eval_runtime": 170.1635,
      "eval_samples_per_second": 0.071,
      "eval_steps_per_second": 0.012,
      "step": 108
    }
  ],
  "logging_steps": 10,
  "max_steps": 108,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 9496422940409856.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}