run-41/checkpoint-688/trainer_state.json

{
  "best_metric": 0.8053333333333333,
  "best_model_checkpoint": "Bert_tinybert-distilled/run-41/checkpoint-516",
  "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": 1.023292899131775,
      "learning_rate": 0.00013414703695256762,
      "loss": 0.5108,
      "step": 172
    },
    {
      "epoch": 1.0,
      "eval_AUC": 0.8195857777777777,
      "eval_BAC": 0.7413333333333334,
      "eval_MCC": 0.4826941273803735,
      "eval_PR_AUC": 0.7801986097000582,
      "eval_accuracy": 0.7413333333333333,
      "eval_loss": 0.4535185992717743,
      "eval_runtime": 6.2428,
      "eval_samples_per_second": 240.277,
      "eval_steps_per_second": 7.529,
      "step": 172
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.2800709009170532,
      "learning_rate": 0.00011498317453077225,
      "loss": 0.476,
      "step": 344
    },
    {
      "epoch": 2.0,
      "eval_AUC": 0.8486666666666666,
      "eval_BAC": 0.762,
      "eval_MCC": 0.5619684351279509,
      "eval_PR_AUC": 0.8033125619568726,
      "eval_accuracy": 0.762,
      "eval_loss": 0.4463256895542145,
      "eval_runtime": 6.2256,
      "eval_samples_per_second": 240.94,
      "eval_steps_per_second": 7.549,
      "step": 344
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.8949135541915894,
      "learning_rate": 9.581931210897687e-05,
      "loss": 0.4613,
      "step": 516
    },
    {
      "epoch": 3.0,
      "eval_AUC": 0.866936888888889,
      "eval_BAC": 0.8053333333333332,
      "eval_MCC": 0.6112951294292618,
      "eval_PR_AUC": 0.8308009588757836,
      "eval_accuracy": 0.8053333333333333,
      "eval_loss": 0.4179477393627167,
      "eval_runtime": 6.2211,
      "eval_samples_per_second": 241.113,
      "eval_steps_per_second": 7.555,
      "step": 516
    },
    {
      "epoch": 4.0,
      "grad_norm": 3.031830072402954,
      "learning_rate": 7.66554496871815e-05,
      "loss": 0.4495,
      "step": 688
    },
    {
      "epoch": 4.0,
      "eval_AUC": 0.8737368888888889,
      "eval_BAC": 0.8053333333333333,
      "eval_MCC": 0.6175917781815502,
      "eval_PR_AUC": 0.8462363913140127,
      "eval_accuracy": 0.8053333333333333,
      "eval_loss": 0.4130651354789734,
      "eval_runtime": 6.2386,
      "eval_samples_per_second": 240.439,
      "eval_steps_per_second": 7.534,
      "step": 688
    }
  ],
  "logging_steps": 500,
  "max_steps": 1376,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "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": 0.7837073808557267,
    "learning_rate": 0.000153310899374363,
    "num_train_epochs": 8,
    "temperature": 5
  }
}