trainer_state.json

{
  "best_metric": 0.9394112188195306,
  "best_model_checkpoint": "model_saves/xlnet-large_lemon-spell_5k_2_p3/checkpoint-268",
  "epoch": 4.0,
  "global_step": 1072,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9394112188195306,
      "eval_loss": 0.4460524618625641,
      "eval_runtime": 4.7696,
      "eval_samples_per_second": 918.731,
      "eval_steps_per_second": 7.338,
      "step": 268
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.939321772381176,
      "eval_loss": 0.46568113565444946,
      "eval_runtime": 4.7856,
      "eval_samples_per_second": 915.656,
      "eval_steps_per_second": 7.314,
      "step": 536
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9389695770301546,
      "eval_loss": 0.4946742057800293,
      "eval_runtime": 4.8031,
      "eval_samples_per_second": 912.319,
      "eval_steps_per_second": 7.287,
      "step": 804
    },
    {
      "epoch": 3.73,
      "learning_rate": 1e-05,
      "loss": 0.2992,
      "step": 1000
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9382819575353034,
      "eval_loss": 0.5468558073043823,
      "eval_runtime": 4.7335,
      "eval_samples_per_second": 925.75,
      "eval_steps_per_second": 7.394,
      "step": 1072
    },
    {
      "epoch": 4.0,
      "step": 1072,
      "total_flos": 1.6984497966809088e+16,
      "train_loss": 0.29385928520515786,
      "train_runtime": 438.825,
      "train_samples_per_second": 1172.586,
      "train_steps_per_second": 9.161
    }
  ],
  "max_steps": 4020,
  "num_train_epochs": 15,
  "total_flos": 1.6984497966809088e+16,
  "trial_name": null,
  "trial_params": null
}