trainer_state.json

{
  "best_global_step": 20,
  "best_metric": 0.6876339912414551,
  "best_model_checkpoint": "tiny_bert_rand_10_v2_rte/checkpoint-20",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 70,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.0655142068862915,
      "learning_rate": 4.9e-05,
      "loss": 0.7006,
      "step": 10
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.4981949458483754,
      "eval_loss": 0.691949188709259,
      "eval_runtime": 0.0958,
      "eval_samples_per_second": 2892.127,
      "eval_steps_per_second": 20.882,
      "step": 10
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.6061921715736389,
      "learning_rate": 4.8e-05,
      "loss": 0.6849,
      "step": 20
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5667870036101083,
      "eval_loss": 0.6876339912414551,
      "eval_runtime": 0.1099,
      "eval_samples_per_second": 2520.758,
      "eval_steps_per_second": 18.2,
      "step": 20
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.5455423593521118,
      "learning_rate": 4.7e-05,
      "loss": 0.6632,
      "step": 30
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5379061371841155,
      "eval_loss": 0.7043308019638062,
      "eval_runtime": 0.0936,
      "eval_samples_per_second": 2960.592,
      "eval_steps_per_second": 21.376,
      "step": 30
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.3710823059082031,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.6264,
      "step": 40
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5415162454873647,
      "eval_loss": 0.7576820850372314,
      "eval_runtime": 0.1029,
      "eval_samples_per_second": 2690.736,
      "eval_steps_per_second": 19.428,
      "step": 40
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.348634958267212,
      "learning_rate": 4.5e-05,
      "loss": 0.5716,
      "step": 50
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5415162454873647,
      "eval_loss": 0.8213313817977905,
      "eval_runtime": 0.0952,
      "eval_samples_per_second": 2908.657,
      "eval_steps_per_second": 21.001,
      "step": 50
    },
    {
      "epoch": 6.0,
      "grad_norm": 2.3401124477386475,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.5151,
      "step": 60
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.5270758122743683,
      "eval_loss": 0.8917742967605591,
      "eval_runtime": 0.1044,
      "eval_samples_per_second": 2653.816,
      "eval_steps_per_second": 19.161,
      "step": 60
    },
    {
      "epoch": 7.0,
      "grad_norm": 2.4810631275177,
      "learning_rate": 4.3e-05,
      "loss": 0.4083,
      "step": 70
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.4981949458483754,
      "eval_loss": 1.0408666133880615,
      "eval_runtime": 0.1039,
      "eval_samples_per_second": 2666.797,
      "eval_steps_per_second": 19.255,
      "step": 70
    },
    {
      "epoch": 7.0,
      "step": 70,
      "total_flos": 457076966492160.0,
      "train_loss": 0.5957424640655518,
      "train_runtime": 15.6236,
      "train_samples_per_second": 7968.708,
      "train_steps_per_second": 32.003
    }
  ],
  "logging_steps": 1,
  "max_steps": 500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 50,
  "save_steps": 500,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 5
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 457076966492160.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}