trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "global_step": 3000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.5,
      "learning_rate": 1.6673333333333335e-05,
      "loss": 0.4114,
      "step": 500
    },
    {
      "epoch": 1.0,
      "learning_rate": 1.3340000000000001e-05,
      "loss": 0.3333,
      "step": 1000
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.86925,
      "eval_classification_report": "              precision    recall  f1-score     support\n0              0.902408  0.927723  0.914890  3030.00000\n1              0.752542  0.686598  0.718059   970.00000\naccuracy       0.869250  0.869250  0.869250     0.86925\nmacro avg      0.827475  0.807160  0.816475  4000.00000\nweighted avg   0.866065  0.869250  0.867159  4000.00000",
      "eval_confusion_matrix": "[[2811  219]\n [ 304  666]]",
      "eval_confusion_matrix_norm": "[[0.92772277 0.07227723]\n [0.31340206 0.68659794]]",
      "eval_f1": 0.7180592991913748,
      "eval_f1_macro": 0.8164747268943042,
      "eval_f1_weighted": 0.8671586721613128,
      "eval_loss": 0.3063889145851135,
      "eval_runtime": 9.6316,
      "eval_samples_per_second": 415.3,
      "eval_steps_per_second": 25.956,
      "step": 1000
    },
    {
      "epoch": 1.5,
      "learning_rate": 1.0013333333333335e-05,
      "loss": 0.2455,
      "step": 1500
    },
    {
      "epoch": 2.0,
      "learning_rate": 6.6866666666666665e-06,
      "loss": 0.2271,
      "step": 2000
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.871,
      "eval_classification_report": "              precision    recall  f1-score   support\n0              0.913487  0.916502  0.914992  3030.000\n1              0.736458  0.728866  0.732642   970.000\naccuracy       0.871000  0.871000  0.871000     0.871\nmacro avg      0.824973  0.822684  0.823817  4000.000\nweighted avg   0.870557  0.871000  0.870772  4000.000",
      "eval_confusion_matrix": "[[2777  253]\n [ 263  707]]",
      "eval_confusion_matrix_norm": "[[0.91650165 0.08349835]\n [0.27113402 0.72886598]]",
      "eval_f1": 0.732642487046632,
      "eval_f1_macro": 0.8238171249071711,
      "eval_f1_weighted": 0.8707720634053486,
      "eval_loss": 0.3905148506164551,
      "eval_runtime": 9.6371,
      "eval_samples_per_second": 415.062,
      "eval_steps_per_second": 25.941,
      "step": 2000
    },
    {
      "epoch": 2.5,
      "learning_rate": 3.3533333333333336e-06,
      "loss": 0.1517,
      "step": 2500
    },
    {
      "epoch": 3.0,
      "learning_rate": 2e-08,
      "loss": 0.1435,
      "step": 3000
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.87925,
      "eval_classification_report": "              precision    recall  f1-score     support\n0              0.908042  0.935314  0.921476  3030.00000\n1              0.777019  0.704124  0.738778   970.00000\naccuracy       0.879250  0.879250  0.879250     0.87925\nmacro avg      0.842531  0.819719  0.830127  4000.00000\nweighted avg   0.876269  0.879250  0.877172  4000.00000",
      "eval_confusion_matrix": "[[2834  196]\n [ 287  683]]",
      "eval_confusion_matrix_norm": "[[0.93531353 0.06468647]\n [0.29587629 0.70412371]]",
      "eval_f1": 0.7387777176852353,
      "eval_f1_macro": 0.8301269502098749,
      "eval_f1_weighted": 0.8771718049600645,
      "eval_loss": 0.4823172092437744,
      "eval_runtime": 9.6475,
      "eval_samples_per_second": 414.617,
      "eval_steps_per_second": 25.914,
      "step": 3000
    },
    {
      "epoch": 3.0,
      "step": 3000,
      "total_flos": 1.262933065728e+16,
      "train_loss": 0.2520793151855469,
      "train_runtime": 430.4281,
      "train_samples_per_second": 111.517,
      "train_steps_per_second": 6.97
    }
  ],
  "max_steps": 3000,
  "num_train_epochs": 3,
  "total_flos": 1.262933065728e+16,
  "trial_name": null,
  "trial_params": null
}