training_checkpoints/checkpoint-3215/trainer_state.json

{
  "best_metric": 0.7392877210524833,
  "best_model_checkpoint": "bart-large-lora-no-grad/checkpoint-2572",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 3215,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.78,
      "learning_rate": 4.740798341109383e-05,
      "loss": 1.3548,
      "step": 500
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7567776917118513,
      "eval_f1_macro": 0.43635159996281114,
      "eval_f1_micro": 0.7567776917118513,
      "eval_f1_weighted": 0.735942387508421,
      "eval_loss": 0.7811183333396912,
      "eval_macro_fpr": 0.023397002359423266,
      "eval_macro_sensitivity": 0.4733929701344221,
      "eval_macro_specificity": 0.98166320866911,
      "eval_precision": 0.7272429944970519,
      "eval_precision_macro": 0.4205947342353388,
      "eval_recall": 0.7567776917118513,
      "eval_recall_macro": 0.4733929701344221,
      "eval_runtime": 124.629,
      "eval_samples_per_second": 10.359,
      "eval_steps_per_second": 1.3,
      "eval_weighted_fpr": 0.022441395082904516,
      "eval_weighted_sensitivity": 0.7567776917118513,
      "eval_weighted_specificity": 0.9681704383248004,
      "step": 643
    },
    {
      "epoch": 1.56,
      "learning_rate": 4.481596682218767e-05,
      "loss": 0.7738,
      "step": 1000
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7893106119287374,
      "eval_f1_macro": 0.561796090533545,
      "eval_f1_micro": 0.7893106119287374,
      "eval_f1_weighted": 0.7783408570207113,
      "eval_loss": 0.657169759273529,
      "eval_macro_fpr": 0.019577279999420293,
      "eval_macro_sensitivity": 0.5639220943833532,
      "eval_macro_specificity": 0.9841641947022086,
      "eval_precision": 0.7847538302510335,
      "eval_precision_macro": 0.652919812505265,
      "eval_recall": 0.7893106119287374,
      "eval_recall_macro": 0.5639220943833532,
      "eval_runtime": 120.3505,
      "eval_samples_per_second": 10.727,
      "eval_steps_per_second": 1.346,
      "eval_weighted_fpr": 0.018709588664190396,
      "eval_weighted_sensitivity": 0.7893106119287374,
      "eval_weighted_specificity": 0.9731523086043933,
      "step": 1286
    },
    {
      "epoch": 2.33,
      "learning_rate": 4.22239502332815e-05,
      "loss": 0.6874,
      "step": 1500
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8009295120061968,
      "eval_f1_macro": 0.6248193338597906,
      "eval_f1_micro": 0.8009295120061968,
      "eval_f1_weighted": 0.7948048170690496,
      "eval_loss": 0.6484545469284058,
      "eval_macro_fpr": 0.01794652035398996,
      "eval_macro_sensitivity": 0.6498133528886191,
      "eval_macro_specificity": 0.985178423920643,
      "eval_precision": 0.7993513486718621,
      "eval_precision_macro": 0.622395994467453,
      "eval_recall": 0.8009295120061968,
      "eval_recall_macro": 0.6498133528886191,
      "eval_runtime": 122.1243,
      "eval_samples_per_second": 10.571,
      "eval_steps_per_second": 1.327,
      "eval_weighted_fpr": 0.017443833570895267,
      "eval_weighted_sensitivity": 0.8009295120061968,
      "eval_weighted_specificity": 0.9767468468034478,
      "step": 1929
    },
    {
      "epoch": 3.11,
      "learning_rate": 3.963193364437533e-05,
      "loss": 0.5681,
      "step": 2000
    },
    {
      "epoch": 3.89,
      "learning_rate": 3.7039917055469156e-05,
      "loss": 0.502,
      "step": 2500
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.82571649883811,
      "eval_f1_macro": 0.7392877210524833,
      "eval_f1_micro": 0.82571649883811,
      "eval_f1_weighted": 0.818185291800718,
      "eval_loss": 0.6912310123443604,
      "eval_macro_fpr": 0.015790725498781473,
      "eval_macro_sensitivity": 0.73988416100595,
      "eval_macro_specificity": 0.9866316247243733,
      "eval_precision": 0.8216256605414862,
      "eval_precision_macro": 0.7660625024843042,
      "eval_recall": 0.82571649883811,
      "eval_recall_macro": 0.73988416100595,
      "eval_runtime": 122.0618,
      "eval_samples_per_second": 10.577,
      "eval_steps_per_second": 1.327,
      "eval_weighted_fpr": 0.01485246550927454,
      "eval_weighted_sensitivity": 0.82571649883811,
      "eval_weighted_specificity": 0.9737578720274908,
      "step": 2572
    },
    {
      "epoch": 4.67,
      "learning_rate": 3.4447900466562985e-05,
      "loss": 0.4443,
      "step": 3000
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8350116189000775,
      "eval_f1_macro": 0.7352088783341215,
      "eval_f1_micro": 0.8350116189000775,
      "eval_f1_weighted": 0.8307677449867239,
      "eval_loss": 0.6655119061470032,
      "eval_macro_fpr": 0.014583446099727526,
      "eval_macro_sensitivity": 0.7343745083798942,
      "eval_macro_specificity": 0.9875414771562052,
      "eval_precision": 0.8323934212037492,
      "eval_precision_macro": 0.758378992160777,
      "eval_recall": 0.8350116189000775,
      "eval_recall_macro": 0.7343745083798942,
      "eval_runtime": 120.1795,
      "eval_samples_per_second": 10.742,
      "eval_steps_per_second": 1.348,
      "eval_weighted_fpr": 0.01391702058150931,
      "eval_weighted_sensitivity": 0.8350116189000775,
      "eval_weighted_specificity": 0.9781105384430038,
      "step": 3215
    }
  ],
  "logging_steps": 500,
  "max_steps": 9645,
  "num_train_epochs": 15,
  "save_steps": 500,
  "total_flos": 2.860711868077056e+16,
  "trial_name": null,
  "trial_params": null
}