checkpoint-340/trainer_state.json

{
  "best_global_step": 340,
  "best_metric": 0.9280575539568345,
  "best_model_checkpoint": "D:/Thesis/Taglish_Gaslighting_V3\\model_outputs\\roberta-tagalog_binary\\checkpoint-340",
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 340,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 10.235954284667969,
      "learning_rate": 1.785340314136126e-05,
      "loss": 0.505757724537569,
      "step": 85
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8650519031141869,
      "eval_f1": 0.8539325842696629,
      "eval_loss": 0.2817898094654083,
      "eval_macro_f1": 0.8642653275046064,
      "eval_non_gas_f1": 0.8745980707395499,
      "eval_non_gas_precision": 0.8292682926829268,
      "eval_non_gas_recall": 0.9251700680272109,
      "eval_precision": 0.912,
      "eval_recall": 0.8028169014084507,
      "eval_roc_auc": 0.9591357669828495,
      "eval_runtime": 12.4921,
      "eval_samples_per_second": 23.135,
      "eval_steps_per_second": 1.521,
      "step": 85
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.08221813291311264,
      "learning_rate": 1.3403141361256546e-05,
      "loss": 0.1839109308579389,
      "step": 170
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9204152249134948,
      "eval_f1": 0.9220338983050848,
      "eval_loss": 0.22523701190948486,
      "eval_macro_f1": 0.9203809067497155,
      "eval_non_gas_f1": 0.9187279151943463,
      "eval_non_gas_precision": 0.9558823529411765,
      "eval_non_gas_recall": 0.8843537414965986,
      "eval_precision": 0.8888888888888888,
      "eval_recall": 0.9577464788732394,
      "eval_roc_auc": 0.9769809332183579,
      "eval_runtime": 12.4908,
      "eval_samples_per_second": 23.137,
      "eval_steps_per_second": 1.521,
      "step": 170
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.053263287991285324,
      "learning_rate": 8.952879581151834e-06,
      "loss": 0.0663631607504452,
      "step": 255
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9204152249134948,
      "eval_f1": 0.9163636363636364,
      "eval_loss": 0.297343909740448,
      "eval_macro_f1": 0.9202280228022802,
      "eval_non_gas_f1": 0.9240924092409241,
      "eval_non_gas_precision": 0.8974358974358975,
      "eval_non_gas_recall": 0.9523809523809523,
      "eval_precision": 0.9473684210526315,
      "eval_recall": 0.8873239436619719,
      "eval_roc_auc": 0.9785378940308518,
      "eval_runtime": 12.543,
      "eval_samples_per_second": 23.041,
      "eval_steps_per_second": 1.515,
      "step": 255
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.026363220065832138,
      "learning_rate": 4.502617801047121e-06,
      "loss": 0.026208843904383043,
      "step": 340
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9307958477508651,
      "eval_f1": 0.9280575539568345,
      "eval_loss": 0.3261849582195282,
      "eval_macro_f1": 0.9306954436450839,
      "eval_non_gas_f1": 0.9333333333333333,
      "eval_non_gas_precision": 0.9150326797385621,
      "eval_non_gas_recall": 0.9523809523809523,
      "eval_precision": 0.9485294117647058,
      "eval_recall": 0.9084507042253521,
      "eval_roc_auc": 0.9782504551116222,
      "eval_runtime": 12.4176,
      "eval_samples_per_second": 23.273,
      "eval_steps_per_second": 1.53,
      "step": 340
    }
  ],
  "logging_steps": 500,
  "max_steps": 425,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 2,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 354673702625280.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}