{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 8656,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.23,
      "grad_norm": 6.905106067657471,
      "learning_rate": 1.9937951301653444e-05,
      "loss": 2.123,
      "step": 500
    },
    {
      "epoch": 0.46,
      "grad_norm": 8.00811767578125,
      "learning_rate": 1.9561549728661312e-05,
      "loss": 1.1578,
      "step": 1000
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.949448585510254,
      "learning_rate": 1.8856153474512088e-05,
      "loss": 1.0324,
      "step": 1500
    },
    {
      "epoch": 0.92,
      "grad_norm": 21.008249282836914,
      "learning_rate": 1.7846033878299232e-05,
      "loss": 0.9744,
      "step": 2000
    },
    {
      "epoch": 1.16,
      "grad_norm": 13.565423011779785,
      "learning_rate": 1.656594722785445e-05,
      "loss": 0.923,
      "step": 2500
    },
    {
      "epoch": 1.39,
      "grad_norm": 12.738595962524414,
      "learning_rate": 1.5059938862204126e-05,
      "loss": 0.8795,
      "step": 3000
    },
    {
      "epoch": 1.62,
      "grad_norm": 13.901044845581055,
      "learning_rate": 1.3379827655499163e-05,
      "loss": 0.8827,
      "step": 3500
    },
    {
      "epoch": 1.85,
      "grad_norm": 16.066545486450195,
      "learning_rate": 1.1583423028434343e-05,
      "loss": 0.8303,
      "step": 4000
    },
    {
      "epoch": 2.08,
      "grad_norm": 13.391914367675781,
      "learning_rate": 9.732535836392112e-06,
      "loss": 0.8274,
      "step": 4500
    },
    {
      "epoch": 2.31,
      "grad_norm": 13.384796142578125,
      "learning_rate": 7.890851575854108e-06,
      "loss": 0.7901,
      "step": 5000
    },
    {
      "epoch": 2.54,
      "grad_norm": 13.527624130249023,
      "learning_rate": 6.121739087989613e-06,
      "loss": 0.7961,
      "step": 5500
    },
    {
      "epoch": 2.77,
      "grad_norm": 29.323654174804688,
      "learning_rate": 4.486070157749059e-06,
      "loss": 0.7687,
      "step": 6000
    },
    {
      "epoch": 3.0,
      "grad_norm": 10.274979591369629,
      "learning_rate": 3.0401250318972643e-06,
      "loss": 0.7579,
      "step": 6500
    },
    {
      "epoch": 3.23,
      "grad_norm": 14.916624069213867,
      "learning_rate": 1.8336559231141726e-06,
      "loss": 0.7294,
      "step": 7000
    },
    {
      "epoch": 3.47,
      "grad_norm": 27.81003189086914,
      "learning_rate": 9.081751312919406e-07,
      "loss": 0.744,
      "step": 7500
    },
    {
      "epoch": 3.7,
      "grad_norm": 18.139156341552734,
      "learning_rate": 2.9552668450792965e-07,
      "loss": 0.7419,
      "step": 8000
    },
    {
      "epoch": 3.93,
      "grad_norm": 18.67233657836914,
      "learning_rate": 1.679064678218989e-08,
      "loss": 0.7492,
      "step": 8500
    },
    {
      "epoch": 4.0,
      "step": 8656,
      "total_flos": 4.200290487097098e+17,
      "train_loss": 0.9207175602093202,
      "train_runtime": 3689.1664,
      "train_samples_per_second": 18.765,
      "train_steps_per_second": 2.346
    }
  ],
  "logging_steps": 500,
  "max_steps": 8656,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 8224,
  "total_flos": 4.200290487097098e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}