trainer_state.json

{
  "best_metric": 0.9050816297531128,
  "best_model_checkpoint": "./mbert_ar_ur/checkpoint-5000",
  "epoch": 4.0,
  "global_step": 5000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.4,
      "learning_rate": 1.885714285714286e-05,
      "loss": 0.4454,
      "step": 500
    },
    {
      "epoch": 0.8,
      "learning_rate": 1.7714285714285717e-05,
      "loss": 0.2955,
      "step": 1000
    },
    {
      "epoch": 1.0,
      "eval_LOC_f1": 0.09510682288077188,
      "eval_ORG_f1": 0.24583663758921492,
      "eval_PER_f1": 0.3119353304802663,
      "eval_loss": 1.4924763441085815,
      "eval_overall_accuracy": 0.4886179434773416,
      "eval_overall_f1": 0.1982125758059368,
      "eval_overall_precision": 0.16568836712913554,
      "eval_overall_recall": 0.2466243050039714,
      "eval_runtime": 2.7088,
      "eval_samples_per_second": 369.163,
      "eval_steps_per_second": 23.257,
      "step": 1250
    },
    {
      "epoch": 1.2,
      "learning_rate": 1.6571428571428574e-05,
      "loss": 0.2182,
      "step": 1500
    },
    {
      "epoch": 1.6,
      "learning_rate": 1.542857142857143e-05,
      "loss": 0.1877,
      "step": 2000
    },
    {
      "epoch": 2.0,
      "learning_rate": 1.4285714285714287e-05,
      "loss": 0.181,
      "step": 2500
    },
    {
      "epoch": 2.0,
      "eval_LOC_f1": 0.16472203157172272,
      "eval_ORG_f1": 0.26800929512006194,
      "eval_PER_f1": 0.6124661246612466,
      "eval_loss": 1.1303032636642456,
      "eval_overall_accuracy": 0.6376770737895553,
      "eval_overall_f1": 0.3045238514346066,
      "eval_overall_precision": 0.2734745494783433,
      "eval_overall_recall": 0.3435266084193805,
      "eval_runtime": 2.7853,
      "eval_samples_per_second": 359.026,
      "eval_steps_per_second": 22.619,
      "step": 2500
    },
    {
      "epoch": 2.4,
      "learning_rate": 1.3142857142857145e-05,
      "loss": 0.1295,
      "step": 3000
    },
    {
      "epoch": 2.8,
      "learning_rate": 1.2e-05,
      "loss": 0.1253,
      "step": 3500
    },
    {
      "epoch": 3.0,
      "eval_LOC_f1": 0.15284677111196027,
      "eval_ORG_f1": 0.3176020408163266,
      "eval_PER_f1": 0.6216577540106952,
      "eval_loss": 1.348933458328247,
      "eval_overall_accuracy": 0.6324617661568821,
      "eval_overall_f1": 0.3217743355043126,
      "eval_overall_precision": 0.2889661713563073,
      "eval_overall_recall": 0.36298649722001586,
      "eval_runtime": 2.5633,
      "eval_samples_per_second": 390.124,
      "eval_steps_per_second": 24.578,
      "step": 3750
    },
    {
      "epoch": 3.2,
      "learning_rate": 1.0857142857142858e-05,
      "loss": 0.098,
      "step": 4000
    },
    {
      "epoch": 3.6,
      "learning_rate": 9.714285714285715e-06,
      "loss": 0.0916,
      "step": 4500
    },
    {
      "epoch": 4.0,
      "learning_rate": 8.571428571428571e-06,
      "loss": 0.0866,
      "step": 5000
    },
    {
      "epoch": 4.0,
      "eval_LOC_f1": 0.42877697841726614,
      "eval_ORG_f1": 0.4028497409326424,
      "eval_PER_f1": 0.6109282422646478,
      "eval_loss": 0.9050816297531128,
      "eval_overall_accuracy": 0.7715836211149482,
      "eval_overall_f1": 0.47474747474747475,
      "eval_overall_precision": 0.464638783269962,
      "eval_overall_recall": 0.4853057982525814,
      "eval_runtime": 2.9178,
      "eval_samples_per_second": 342.725,
      "eval_steps_per_second": 21.592,
      "step": 5000
    }
  ],
  "max_steps": 8750,
  "num_train_epochs": 7,
  "total_flos": 1363663262792160.0,
  "trial_name": null,
  "trial_params": null
}