checkpoint-5000/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.6469031233456857,
  "global_step": 5000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.26,
      "learning_rate": 1.8235397917769543e-05,
      "loss": 21.3637,
      "step": 500
    },
    {
      "epoch": 0.53,
      "learning_rate": 1.6470795835539088e-05,
      "loss": 10.4706,
      "step": 1000
    },
    {
      "epoch": 0.79,
      "learning_rate": 1.470619375330863e-05,
      "loss": 5.7273,
      "step": 1500
    },
    {
      "epoch": 1.0,
      "eval_loss": 1.1415340900421143,
      "eval_runtime": 9.212,
      "eval_samples_per_second": 182.262,
      "eval_steps_per_second": 11.398,
      "step": 1889
    },
    {
      "epoch": 1.06,
      "learning_rate": 1.2941591671078173e-05,
      "loss": 3.094,
      "step": 2000
    },
    {
      "epoch": 1.32,
      "learning_rate": 1.1176989588847715e-05,
      "loss": 1.946,
      "step": 2500
    },
    {
      "epoch": 1.59,
      "learning_rate": 9.41238750661726e-06,
      "loss": 1.3926,
      "step": 3000
    },
    {
      "epoch": 1.85,
      "learning_rate": 7.6477854243868e-06,
      "loss": 1.0576,
      "step": 3500
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.2452467828989029,
      "eval_runtime": 9.2184,
      "eval_samples_per_second": 182.137,
      "eval_steps_per_second": 11.39,
      "step": 3778
    },
    {
      "epoch": 2.12,
      "learning_rate": 5.883183342156344e-06,
      "loss": 0.8478,
      "step": 4000
    },
    {
      "epoch": 2.38,
      "learning_rate": 4.118581259925887e-06,
      "loss": 0.71,
      "step": 4500
    },
    {
      "epoch": 2.65,
      "learning_rate": 2.3539791776954297e-06,
      "loss": 0.6478,
      "step": 5000
    }
  ],
  "max_steps": 5667,
  "num_train_epochs": 3,
  "total_flos": 5285910659727360.0,
  "trial_name": null,
  "trial_params": null
}