trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 10.0,
  "global_step": 5450,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.92,
      "learning_rate": 2.724770642201835e-05,
      "loss": 1.6052,
      "step": 500
    },
    {
      "epoch": 1.83,
      "learning_rate": 2.44954128440367e-05,
      "loss": 1.0401,
      "step": 1000
    },
    {
      "epoch": 2.75,
      "learning_rate": 2.1743119266055045e-05,
      "loss": 0.8638,
      "step": 1500
    },
    {
      "epoch": 3.67,
      "learning_rate": 1.8990825688073394e-05,
      "loss": 0.7357,
      "step": 2000
    },
    {
      "epoch": 4.59,
      "learning_rate": 1.6238532110091743e-05,
      "loss": 0.6302,
      "step": 2500
    },
    {
      "epoch": 5.5,
      "learning_rate": 1.3486238532110092e-05,
      "loss": 0.5428,
      "step": 3000
    },
    {
      "epoch": 6.42,
      "learning_rate": 1.0733944954128442e-05,
      "loss": 0.4863,
      "step": 3500
    },
    {
      "epoch": 7.34,
      "learning_rate": 7.981651376146789e-06,
      "loss": 0.4272,
      "step": 4000
    },
    {
      "epoch": 8.26,
      "learning_rate": 5.229357798165138e-06,
      "loss": 0.391,
      "step": 4500
    },
    {
      "epoch": 9.17,
      "learning_rate": 2.4770642201834866e-06,
      "loss": 0.3596,
      "step": 5000
    },
    {
      "epoch": 10.0,
      "step": 5450,
      "total_flos": 2.205639244649595e+17,
      "train_loss": 0.6770142889460292,
      "train_runtime": 1770.0654,
      "train_samples_per_second": 738.758,
      "train_steps_per_second": 3.079
    }
  ],
  "max_steps": 5450,
  "num_train_epochs": 10,
  "total_flos": 2.205639244649595e+17,
  "trial_name": null,
  "trial_params": null
}