trainer_state.json

{
  "best_metric": 0.2906099259853363,
  "best_model_checkpoint": "tinybert_train_book_v2_qqp/checkpoint-4266",
  "epoch": 8.0,
  "eval_steps": 500,
  "global_step": 11376,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 2.6571297645568848,
      "learning_rate": 4.9e-05,
      "loss": 0.397,
      "step": 1422
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8484046500123671,
      "eval_combined_score": 0.8136623401560591,
      "eval_f1": 0.7789200302997511,
      "eval_loss": 0.34448662400245667,
      "eval_runtime": 13.1165,
      "eval_samples_per_second": 3082.378,
      "eval_steps_per_second": 12.046,
      "step": 1422
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.481895923614502,
      "learning_rate": 4.8e-05,
      "loss": 0.2958,
      "step": 2844
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8694781103141231,
      "eval_combined_score": 0.8457658634524586,
      "eval_f1": 0.8220536165907941,
      "eval_loss": 0.3028949499130249,
      "eval_runtime": 12.6986,
      "eval_samples_per_second": 3183.805,
      "eval_steps_per_second": 12.442,
      "step": 2844
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.8351011276245117,
      "learning_rate": 4.7e-05,
      "loss": 0.2432,
      "step": 4266
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8763047242146921,
      "eval_combined_score": 0.8527731836322547,
      "eval_f1": 0.8292416430498173,
      "eval_loss": 0.2906099259853363,
      "eval_runtime": 13.3193,
      "eval_samples_per_second": 3035.444,
      "eval_steps_per_second": 11.862,
      "step": 4266
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.612133502960205,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.1995,
      "step": 5688
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8798417017066534,
      "eval_combined_score": 0.8556741804870671,
      "eval_f1": 0.8315066592674807,
      "eval_loss": 0.2983686625957489,
      "eval_runtime": 13.2337,
      "eval_samples_per_second": 3055.089,
      "eval_steps_per_second": 11.939,
      "step": 5688
    },
    {
      "epoch": 5.0,
      "grad_norm": 3.5917608737945557,
      "learning_rate": 4.5e-05,
      "loss": 0.1629,
      "step": 7110
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8847390551570616,
      "eval_combined_score": 0.8650687512531376,
      "eval_f1": 0.8453984473492137,
      "eval_loss": 0.309291273355484,
      "eval_runtime": 13.3435,
      "eval_samples_per_second": 3029.941,
      "eval_steps_per_second": 11.841,
      "step": 7110
    },
    {
      "epoch": 6.0,
      "grad_norm": 4.064701557159424,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.1334,
      "step": 8532
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8848627256987386,
      "eval_combined_score": 0.8641723937494736,
      "eval_f1": 0.8434820618002086,
      "eval_loss": 0.34539419412612915,
      "eval_runtime": 13.5312,
      "eval_samples_per_second": 2987.905,
      "eval_steps_per_second": 11.677,
      "step": 8532
    },
    {
      "epoch": 7.0,
      "grad_norm": 2.7387797832489014,
      "learning_rate": 4.3e-05,
      "loss": 0.1097,
      "step": 9954
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8846648528320554,
      "eval_combined_score": 0.8649125359959713,
      "eval_f1": 0.8451602191598871,
      "eval_loss": 0.3594832122325897,
      "eval_runtime": 13.2101,
      "eval_samples_per_second": 3060.542,
      "eval_steps_per_second": 11.961,
      "step": 9954
    },
    {
      "epoch": 8.0,
      "grad_norm": 3.645385503768921,
      "learning_rate": 4.2e-05,
      "loss": 0.0925,
      "step": 11376
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8853821419737818,
      "eval_combined_score": 0.8647196078020747,
      "eval_f1": 0.8440570736303675,
      "eval_loss": 0.3933313190937042,
      "eval_runtime": 12.7924,
      "eval_samples_per_second": 3160.468,
      "eval_steps_per_second": 12.351,
      "step": 11376
    },
    {
      "epoch": 8.0,
      "step": 11376,
      "total_flos": 7.633075201391002e+16,
      "train_loss": 0.20422728789338918,
      "train_runtime": 1699.6069,
      "train_samples_per_second": 10703.828,
      "train_steps_per_second": 41.833
    }
  ],
  "logging_steps": 1,
  "max_steps": 71100,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 50,
  "save_steps": 500,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 5
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 7.633075201391002e+16,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}