trainer_state.json

{
  "best_metric": 96.38462406015037,
  "best_model_checkpoint": "xlm-roberta-finetuned-final-squad/checkpoint-2953",
  "epoch": 2.999492127983748,
  "eval_steps": 1000,
  "global_step": 2953,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.51,
      "learning_rate": 1.8983739837398377e-05,
      "loss": 0.3406,
      "step": 500
    },
    {
      "epoch": 1.0,
      "eval_HasAns_exact": 79.25,
      "eval_HasAns_f1": 84.45705266955264,
      "eval_HasAns_total": 800,
      "eval_NoAns_exact": 99.33333333333333,
      "eval_NoAns_f1": 99.33333333333333,
      "eval_NoAns_total": 3000,
      "eval_best_exact": 95.63157894736842,
      "eval_best_exact_thresh": 0.0,
      "eval_best_f1": 96.7278005620109,
      "eval_best_f1_thresh": 0.0,
      "eval_exact": 95.10526315789474,
      "eval_f1": 96.2014847725374,
      "eval_runtime": 258.8129,
      "eval_samples_per_second": 14.682,
      "eval_steps_per_second": 3.671,
      "eval_total": 3800,
      "step": 984
    },
    {
      "epoch": 1.02,
      "learning_rate": 1.796747967479675e-05,
      "loss": 0.2614,
      "step": 1000
    },
    {
      "epoch": 1.52,
      "learning_rate": 1.6951219512195124e-05,
      "loss": 0.1962,
      "step": 1500
    },
    {
      "epoch": 2.0,
      "eval_HasAns_exact": 82.75,
      "eval_HasAns_f1": 85.20308688344682,
      "eval_HasAns_total": 800,
      "eval_NoAns_exact": 98.26666666666667,
      "eval_NoAns_f1": 98.26666666666667,
      "eval_NoAns_total": 3000,
      "eval_best_exact": 96.36842105263158,
      "eval_best_exact_thresh": 0.0,
      "eval_best_f1": 96.88486039651504,
      "eval_best_f1_thresh": 0.0,
      "eval_exact": 95.0,
      "eval_f1": 95.51643934388353,
      "eval_runtime": 258.6634,
      "eval_samples_per_second": 14.691,
      "eval_steps_per_second": 3.673,
      "eval_total": 3800,
      "step": 1969
    },
    {
      "epoch": 2.03,
      "learning_rate": 1.5934959349593496e-05,
      "loss": 0.1842,
      "step": 2000
    },
    {
      "epoch": 2.54,
      "learning_rate": 1.4918699186991872e-05,
      "loss": 0.1307,
      "step": 2500
    },
    {
      "epoch": 3.0,
      "eval_HasAns_exact": 84.75,
      "eval_HasAns_f1": 85.82696428571427,
      "eval_HasAns_total": 800,
      "eval_NoAns_exact": 99.2,
      "eval_NoAns_f1": 99.2,
      "eval_NoAns_total": 3000,
      "eval_best_exact": 96.78947368421052,
      "eval_best_exact_thresh": 0.0,
      "eval_best_f1": 97.01620300751878,
      "eval_best_f1_thresh": 0.0,
      "eval_exact": 96.15789473684211,
      "eval_f1": 96.38462406015037,
      "eval_runtime": 258.7515,
      "eval_samples_per_second": 14.686,
      "eval_steps_per_second": 3.671,
      "eval_total": 3800,
      "step": 2953
    }
  ],
  "logging_steps": 500,
  "max_steps": 9840,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 3.290896444280371e+16,
  "trial_name": null,
  "trial_params": null
}