{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 4250,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.5882352941176471,
      "grad_norm": 6.994039058685303,
      "learning_rate": 4.4129411764705886e-05,
      "loss": 0.5798,
      "step": 500
    },
    {
      "epoch": 1.1764705882352942,
      "grad_norm": 7.607770919799805,
      "learning_rate": 3.8247058823529416e-05,
      "loss": 0.4276,
      "step": 1000
    },
    {
      "epoch": 1.7647058823529411,
      "grad_norm": 2.9817395210266113,
      "learning_rate": 3.236470588235294e-05,
      "loss": 0.3671,
      "step": 1500
    },
    {
      "epoch": 2.3529411764705883,
      "grad_norm": 15.393789291381836,
      "learning_rate": 2.648235294117647e-05,
      "loss": 0.315,
      "step": 2000
    },
    {
      "epoch": 2.9411764705882355,
      "grad_norm": 12.855734825134277,
      "learning_rate": 2.06e-05,
      "loss": 0.2736,
      "step": 2500
    },
    {
      "epoch": 3.5294117647058822,
      "grad_norm": 31.535274505615234,
      "learning_rate": 1.4717647058823531e-05,
      "loss": 0.2434,
      "step": 3000
    },
    {
      "epoch": 4.117647058823529,
      "grad_norm": 0.42044809460639954,
      "learning_rate": 8.835294117647059e-06,
      "loss": 0.2018,
      "step": 3500
    },
    {
      "epoch": 4.705882352941177,
      "grad_norm": 32.79914093017578,
      "learning_rate": 2.9529411764705886e-06,
      "loss": 0.1748,
      "step": 4000
    },
    {
      "epoch": 5.0,
      "step": 4250,
      "total_flos": 2.634727646896128e+18,
      "train_loss": 0.3136866392247817,
      "train_runtime": 1820.5916,
      "train_samples_per_second": 18.675,
      "train_steps_per_second": 2.334
    }
  ],
  "logging_steps": 500,
  "max_steps": 4250,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "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": 2.634727646896128e+18,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}