run-21/checkpoint-688/trainer_state.json

{
  "best_metric": 0.8093333333333333,
  "best_model_checkpoint": "Bert_tinybert-distilled/run-21/checkpoint-688",
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 688,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.8566362261772156,
      "learning_rate": 0.00023417386468774877,
      "loss": 0.5506,
      "step": 172
    },
    {
      "epoch": 1.0,
      "eval_AUC": 0.8244817777777778,
      "eval_BAC": 0.7586666666666667,
      "eval_MCC": 0.5194728884404217,
      "eval_PR_AUC": 0.7813767310995844,
      "eval_accuracy": 0.7586666666666667,
      "eval_loss": 0.4952147603034973,
      "eval_runtime": 6.2747,
      "eval_samples_per_second": 239.054,
      "eval_steps_per_second": 7.49,
      "step": 172
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.2101809978485107,
      "learning_rate": 0.00020490213160178017,
      "loss": 0.5088,
      "step": 344
    },
    {
      "epoch": 2.0,
      "eval_AUC": 0.8558417777777777,
      "eval_BAC": 0.7833333333333333,
      "eval_MCC": 0.598402241494527,
      "eval_PR_AUC": 0.8083122328065501,
      "eval_accuracy": 0.7833333333333333,
      "eval_loss": 0.47188517451286316,
      "eval_runtime": 6.2564,
      "eval_samples_per_second": 239.754,
      "eval_steps_per_second": 7.512,
      "step": 344
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.8857367038726807,
      "learning_rate": 0.00017563039851581157,
      "loss": 0.4883,
      "step": 516
    },
    {
      "epoch": 3.0,
      "eval_AUC": 0.8698346666666668,
      "eval_BAC": 0.804,
      "eval_MCC": 0.6118496314310189,
      "eval_PR_AUC": 0.8393731728166693,
      "eval_accuracy": 0.804,
      "eval_loss": 0.4446364641189575,
      "eval_runtime": 6.227,
      "eval_samples_per_second": 240.887,
      "eval_steps_per_second": 7.548,
      "step": 516
    },
    {
      "epoch": 4.0,
      "grad_norm": 3.583167314529419,
      "learning_rate": 0.00014635866542984298,
      "loss": 0.4758,
      "step": 688
    },
    {
      "epoch": 4.0,
      "eval_AUC": 0.8714311111111109,
      "eval_BAC": 0.8093333333333333,
      "eval_MCC": 0.6240167633484885,
      "eval_PR_AUC": 0.8444772811224848,
      "eval_accuracy": 0.8093333333333333,
      "eval_loss": 0.4367929697036743,
      "eval_runtime": 6.2174,
      "eval_samples_per_second": 241.258,
      "eval_steps_per_second": 7.559,
      "step": 688
    }
  ],
  "logging_steps": 500,
  "max_steps": 1548,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1801510920000.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.8568716109861949,
    "learning_rate": 0.00026344559777371736,
    "num_train_epochs": 9,
    "temperature": 2
  }
}