trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 10.48951048951049,
  "global_step": 1500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 3.5,
      "learning_rate": 4.9800000000000004e-05,
      "loss": 4.6313,
      "step": 500
    },
    {
      "epoch": 3.5,
      "eval_cer": 0.9993038029762423,
      "eval_loss": 3.357034921646118,
      "eval_runtime": 144.6064,
      "eval_samples_per_second": 7.897,
      "eval_steps_per_second": 0.989,
      "step": 500
    },
    {
      "epoch": 6.99,
      "learning_rate": 9.98e-05,
      "loss": 2.8906,
      "step": 1000
    },
    {
      "epoch": 6.99,
      "eval_cer": 0.1851720912018101,
      "eval_loss": 0.7587499022483826,
      "eval_runtime": 144.5652,
      "eval_samples_per_second": 7.9,
      "eval_steps_per_second": 0.989,
      "step": 1000
    },
    {
      "epoch": 10.49,
      "learning_rate": 9.190243902439025e-05,
      "loss": 0.4635,
      "step": 1500
    },
    {
      "epoch": 10.49,
      "eval_cer": 0.04839657122965799,
      "eval_loss": 0.18757499754428864,
      "eval_runtime": 146.4828,
      "eval_samples_per_second": 7.796,
      "eval_steps_per_second": 0.976,
      "step": 1500
    }
  ],
  "max_steps": 7150,
  "num_train_epochs": 50,
  "total_flos": 5.574687107305868e+18,
  "trial_name": null,
  "trial_params": null
}