{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 14.76121562952243,
  "eval_steps": 500,
  "global_step": 2550,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 2.89,
      "grad_norm": 1.0645270347595215,
      "learning_rate": 8.486332109343125e-06,
      "loss": 3.4851,
      "step": 500
    },
    {
      "epoch": 2.89,
      "eval_loss": 0.5308476686477661,
      "eval_runtime": 54.8537,
      "eval_samples_per_second": 17.264,
      "eval_steps_per_second": 1.094,
      "step": 500
    },
    {
      "epoch": 5.79,
      "grad_norm": 1.1767557859420776,
      "learning_rate": 6.446348429212567e-06,
      "loss": 0.4272,
      "step": 1000
    },
    {
      "epoch": 5.79,
      "eval_loss": 0.48605769872665405,
      "eval_runtime": 54.7929,
      "eval_samples_per_second": 17.283,
      "eval_steps_per_second": 1.095,
      "step": 1000
    },
    {
      "epoch": 8.68,
      "grad_norm": 1.228811264038086,
      "learning_rate": 4.406364749082008e-06,
      "loss": 0.3706,
      "step": 1500
    },
    {
      "epoch": 8.68,
      "eval_loss": 0.47189345955848694,
      "eval_runtime": 54.8284,
      "eval_samples_per_second": 17.272,
      "eval_steps_per_second": 1.094,
      "step": 1500
    },
    {
      "epoch": 11.58,
      "grad_norm": 1.4037022590637207,
      "learning_rate": 2.3663810689514487e-06,
      "loss": 0.345,
      "step": 2000
    },
    {
      "epoch": 11.58,
      "eval_loss": 0.4648812413215637,
      "eval_runtime": 54.9124,
      "eval_samples_per_second": 17.246,
      "eval_steps_per_second": 1.093,
      "step": 2000
    },
    {
      "epoch": 14.47,
      "grad_norm": 1.5453025102615356,
      "learning_rate": 3.263973888208895e-07,
      "loss": 0.3283,
      "step": 2500
    },
    {
      "epoch": 14.47,
      "eval_loss": 0.4620264172554016,
      "eval_runtime": 55.0096,
      "eval_samples_per_second": 17.215,
      "eval_steps_per_second": 1.091,
      "step": 2500
    }
  ],
  "logging_steps": 500,
  "max_steps": 2580,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 15,
  "save_steps": 50,
  "total_flos": 9.936487691845632e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}