trainer_state.json

{
  "best_metric": 0.6061349693251534,
  "best_model_checkpoint": "bert-finetuned-sem_eval-georgian/checkpoint-2110",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 2110,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.45023696682464454,
      "eval_f1": 0.5087014725568942,
      "eval_loss": 0.4644582271575928,
      "eval_roc_auc": 0.6718009478672986,
      "eval_runtime": 12.4119,
      "eval_samples_per_second": 33.999,
      "eval_steps_per_second": 4.27,
      "step": 422
    },
    {
      "epoch": 1.1848341232227488,
      "grad_norm": 5.8598737716674805,
      "learning_rate": 1.5260663507109007e-05,
      "loss": 0.4952,
      "step": 500
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.523696682464455,
      "eval_f1": 0.5441176470588235,
      "eval_loss": 0.47526419162750244,
      "eval_roc_auc": 0.6951026856240126,
      "eval_runtime": 12.3933,
      "eval_samples_per_second": 34.051,
      "eval_steps_per_second": 4.277,
      "step": 844
    },
    {
      "epoch": 2.3696682464454977,
      "grad_norm": 2.699039936065674,
      "learning_rate": 1.052132701421801e-05,
      "loss": 0.3968,
      "step": 1000
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5616113744075829,
      "eval_f1": 0.6025316455696202,
      "eval_loss": 0.43833649158477783,
      "eval_roc_auc": 0.7306477093206951,
      "eval_runtime": 12.4177,
      "eval_samples_per_second": 33.984,
      "eval_steps_per_second": 4.268,
      "step": 1266
    },
    {
      "epoch": 3.5545023696682465,
      "grad_norm": 6.871908664703369,
      "learning_rate": 5.7819905213270145e-06,
      "loss": 0.3198,
      "step": 1500
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5734597156398105,
      "eval_f1": 0.6,
      "eval_loss": 0.49058645963668823,
      "eval_roc_auc": 0.7306477093206951,
      "eval_runtime": 12.4414,
      "eval_samples_per_second": 33.919,
      "eval_steps_per_second": 4.26,
      "step": 1688
    },
    {
      "epoch": 4.739336492890995,
      "grad_norm": 3.2302420139312744,
      "learning_rate": 1.042654028436019e-06,
      "loss": 0.2409,
      "step": 2000
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5829383886255924,
      "eval_f1": 0.6061349693251534,
      "eval_loss": 0.4966834783554077,
      "eval_roc_auc": 0.7349921011058452,
      "eval_runtime": 12.2307,
      "eval_samples_per_second": 34.503,
      "eval_steps_per_second": 4.333,
      "step": 2110
    }
  ],
  "logging_steps": 500,
  "max_steps": 2110,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 4440078789120000.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}