trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "global_step": 1443,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.41519250780437045,
      "eval_accuracy_0": 0.7706576728499157,
      "eval_accuracy_1": 0.18053777208706787,
      "eval_accuracy_2": 0.36496350364963503,
      "eval_loss": 0.7430521249771118,
      "eval_runtime": 2.1482,
      "eval_samples_per_second": 894.711,
      "eval_steps_per_second": 28.396,
      "step": 481
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.41519250780437045,
      "eval_accuracy_0": 0.8010118043844857,
      "eval_accuracy_1": 0.21895006402048656,
      "eval_accuracy_2": 0.2773722627737226,
      "eval_loss": 0.734639585018158,
      "eval_runtime": 2.1556,
      "eval_samples_per_second": 891.64,
      "eval_steps_per_second": 28.299,
      "step": 962
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.4162330905306972,
      "eval_accuracy_0": 0.8145025295109612,
      "eval_accuracy_1": 0.18950064020486557,
      "eval_accuracy_2": 0.3083941605839416,
      "eval_loss": 0.766707181930542,
      "eval_runtime": 2.1595,
      "eval_samples_per_second": 890.029,
      "eval_steps_per_second": 28.248,
      "step": 1443
    },
    {
      "epoch": 3.0,
      "step": 1443,
      "total_flos": 736655902431120.0,
      "train_loss": 0.6979017356824974,
      "train_runtime": 157.0033,
      "train_samples_per_second": 293.936,
      "train_steps_per_second": 9.191
    }
  ],
  "max_steps": 1443,
  "num_train_epochs": 3,
  "total_flos": 736655902431120.0,
  "trial_name": null,
  "trial_params": null
}