{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 2000,
  "global_step": 8340,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0003597769382982551,
      "grad_norm": 43.59233474731445,
      "learning_rate": 0.0,
      "loss": 4.6611,
      "step": 1
    },
    {
      "epoch": 0.17988846914912754,
      "grad_norm": 0.34114521741867065,
      "learning_rate": 4.816339066339067e-05,
      "loss": 0.4796,
      "step": 500
    },
    {
      "epoch": 0.3597769382982551,
      "grad_norm": 0.40485045313835144,
      "learning_rate": 4.5092137592137595e-05,
      "loss": 0.1239,
      "step": 1000
    },
    {
      "epoch": 0.5396654074473827,
      "grad_norm": 0.26643964648246765,
      "learning_rate": 4.202088452088452e-05,
      "loss": 0.0915,
      "step": 1500
    },
    {
      "epoch": 0.7195538765965102,
      "grad_norm": 0.2876332998275757,
      "learning_rate": 3.894963144963145e-05,
      "loss": 0.074,
      "step": 2000
    },
    {
      "epoch": 0.7195538765965102,
      "eval_loss": 0.051152389496564865,
      "eval_runtime": 73.981,
      "eval_samples_per_second": 149.038,
      "eval_steps_per_second": 18.64,
      "step": 2000
    },
    {
      "epoch": 0.8994423457456378,
      "grad_norm": 0.27155473828315735,
      "learning_rate": 3.587837837837838e-05,
      "loss": 0.0643,
      "step": 2500
    },
    {
      "epoch": 1.079150926425616,
      "grad_norm": 0.257112592458725,
      "learning_rate": 3.2807125307125305e-05,
      "loss": 0.0573,
      "step": 3000
    },
    {
      "epoch": 1.2590393955747436,
      "grad_norm": 0.1687777042388916,
      "learning_rate": 2.9735872235872236e-05,
      "loss": 0.0521,
      "step": 3500
    },
    {
      "epoch": 1.4389278647238712,
      "grad_norm": 0.2531055212020874,
      "learning_rate": 2.6664619164619163e-05,
      "loss": 0.0492,
      "step": 4000
    },
    {
      "epoch": 1.4389278647238712,
      "eval_loss": 0.037362292408943176,
      "eval_runtime": 73.4788,
      "eval_samples_per_second": 150.057,
      "eval_steps_per_second": 18.767,
      "step": 4000
    },
    {
      "epoch": 1.6188163338729988,
      "grad_norm": 0.2625373899936676,
      "learning_rate": 2.3593366093366094e-05,
      "loss": 0.0471,
      "step": 4500
    },
    {
      "epoch": 1.7987048030221264,
      "grad_norm": 0.20707161724567413,
      "learning_rate": 2.0522113022113025e-05,
      "loss": 0.0443,
      "step": 5000
    },
    {
      "epoch": 1.9785932721712538,
      "grad_norm": 0.37195566296577454,
      "learning_rate": 1.7450859950859953e-05,
      "loss": 0.0431,
      "step": 5500
    },
    {
      "epoch": 2.158301852851232,
      "grad_norm": 0.2803611755371094,
      "learning_rate": 1.437960687960688e-05,
      "loss": 0.04,
      "step": 6000
    },
    {
      "epoch": 2.158301852851232,
      "eval_loss": 0.032591212540864944,
      "eval_runtime": 73.1424,
      "eval_samples_per_second": 150.747,
      "eval_steps_per_second": 18.854,
      "step": 6000
    },
    {
      "epoch": 2.3381903220003597,
      "grad_norm": 0.26601219177246094,
      "learning_rate": 1.130835380835381e-05,
      "loss": 0.0383,
      "step": 6500
    },
    {
      "epoch": 2.5180787911494873,
      "grad_norm": 0.18378710746765137,
      "learning_rate": 8.237100737100738e-06,
      "loss": 0.0386,
      "step": 7000
    },
    {
      "epoch": 2.697967260298615,
      "grad_norm": 0.2807880938053131,
      "learning_rate": 5.165847665847666e-06,
      "loss": 0.0382,
      "step": 7500
    },
    {
      "epoch": 2.8778557294477425,
      "grad_norm": 0.22750866413116455,
      "learning_rate": 2.094594594594595e-06,
      "loss": 0.0368,
      "step": 8000
    },
    {
      "epoch": 2.8778557294477425,
      "eval_loss": 0.03069758228957653,
      "eval_runtime": 74.1126,
      "eval_samples_per_second": 148.774,
      "eval_steps_per_second": 18.607,
      "step": 8000
    }
  ],
  "logging_steps": 500,
  "max_steps": 8340,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 4000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.941286555704361e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}