trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 29.85745140388769,
  "eval_steps": 500,
  "global_step": 4320,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 2e-05,
      "loss": 2.3508,
      "step": 1
    },
    {
      "epoch": 3.46,
      "learning_rate": 1.7699074074074075e-05,
      "loss": 0.4416,
      "step": 500
    },
    {
      "epoch": 6.91,
      "learning_rate": 1.5412037037037036e-05,
      "loss": 0.0209,
      "step": 1000
    },
    {
      "epoch": 10.37,
      "learning_rate": 1.3125e-05,
      "loss": 0.0105,
      "step": 1500
    },
    {
      "epoch": 13.82,
      "learning_rate": 1.0828703703703705e-05,
      "loss": 0.0155,
      "step": 2000
    },
    {
      "epoch": 17.28,
      "learning_rate": 8.51388888888889e-06,
      "loss": 0.0157,
      "step": 2500
    },
    {
      "epoch": 20.73,
      "learning_rate": 6.199074074074075e-06,
      "loss": 0.0118,
      "step": 3000
    },
    {
      "epoch": 24.19,
      "learning_rate": 3.88425925925926e-06,
      "loss": 0.0117,
      "step": 3500
    },
    {
      "epoch": 27.65,
      "learning_rate": 1.5694444444444446e-06,
      "loss": 0.0112,
      "step": 4000
    },
    {
      "epoch": 29.86,
      "step": 4320,
      "total_flos": 2.83089920921174e+18,
      "train_loss": 0.06351593561746456,
      "train_runtime": 61407.6851,
      "train_samples_per_second": 1.131,
      "train_steps_per_second": 0.07
    }
  ],
  "logging_steps": 500,
  "max_steps": 4320,
  "num_train_epochs": 30,
  "save_steps": 500,
  "total_flos": 2.83089920921174e+18,
  "trial_name": null,
  "trial_params": null
}