trainer_state.json

{
  "best_metric": 0.8403908794788274,
  "best_model_checkpoint": "/home/ubuntu/utah/babylm-24/src/evaluation/results/finetune/DebertaV2-Base-10M_babylm-A/mrpc/checkpoint-2295",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 2295,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6960784196853638,
      "eval_f1": 0.8109756097560976,
      "eval_loss": 0.5777735710144043,
      "eval_mcc": 0.17012280758673753,
      "eval_runtime": 0.347,
      "eval_samples_per_second": 587.838,
      "eval_steps_per_second": 74.921,
      "step": 459
    },
    {
      "epoch": 1.0893246187363834,
      "grad_norm": 2.380143880844116,
      "learning_rate": 1.5642701525054468e-05,
      "loss": 0.6139,
      "step": 500
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.720588207244873,
      "eval_f1": 0.8235294117647058,
      "eval_loss": 0.5808757543563843,
      "eval_mcc": 0.26984949202435965,
      "eval_runtime": 0.3355,
      "eval_samples_per_second": 608.081,
      "eval_steps_per_second": 77.5,
      "step": 918
    },
    {
      "epoch": 2.178649237472767,
      "grad_norm": 6.764563083648682,
      "learning_rate": 1.1285403050108935e-05,
      "loss": 0.5239,
      "step": 1000
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7598039507865906,
      "eval_f1": 0.8382838283828383,
      "eval_loss": 0.6974886655807495,
      "eval_mcc": 0.4042253913596411,
      "eval_runtime": 0.3292,
      "eval_samples_per_second": 619.643,
      "eval_steps_per_second": 78.974,
      "step": 1377
    },
    {
      "epoch": 3.2679738562091503,
      "grad_norm": 30.021303176879883,
      "learning_rate": 6.928104575163399e-06,
      "loss": 0.3466,
      "step": 1500
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7549019455909729,
      "eval_f1": 0.8387096774193549,
      "eval_loss": 1.0106741189956665,
      "eval_mcc": 0.3852755834560997,
      "eval_runtime": 0.3222,
      "eval_samples_per_second": 633.08,
      "eval_steps_per_second": 80.687,
      "step": 1836
    },
    {
      "epoch": 4.357298474945534,
      "grad_norm": 28.98902702331543,
      "learning_rate": 2.570806100217865e-06,
      "loss": 0.2213,
      "step": 2000
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7598039507865906,
      "eval_f1": 0.8403908794788274,
      "eval_loss": 1.1580651998519897,
      "eval_mcc": 0.40096662922574366,
      "eval_runtime": 0.3547,
      "eval_samples_per_second": 575.054,
      "eval_steps_per_second": 73.291,
      "step": 2295
    },
    {
      "epoch": 5.0,
      "step": 2295,
      "total_flos": 1082018524016640.0,
      "train_loss": 0.3891895601692283,
      "train_runtime": 145.114,
      "train_samples_per_second": 126.383,
      "train_steps_per_second": 15.815
    }
  ],
  "logging_steps": 500,
  "max_steps": 2295,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 3,
        "early_stopping_threshold": 0.001
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1082018524016640.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}