{
  "best_global_step": 1560,
  "best_metric": 0.8301976323127747,
  "best_model_checkpoint": "./bert_scratch/checkpoint-1560",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 1560,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5404862366887684,
      "eval_loss": 0.956041157245636,
      "eval_runtime": 65.1158,
      "eval_samples_per_second": 152.866,
      "eval_steps_per_second": 1.198,
      "step": 312
    },
    {
      "epoch": 1.6025641025641026,
      "grad_norm": 10.819154739379883,
      "learning_rate": 1.3602564102564105e-05,
      "loss": 0.9773,
      "step": 500
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5778581474784007,
      "eval_loss": 0.9190859198570251,
      "eval_runtime": 65.1597,
      "eval_samples_per_second": 152.763,
      "eval_steps_per_second": 1.197,
      "step": 624
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5923246935905164,
      "eval_loss": 0.8675242066383362,
      "eval_runtime": 65.1616,
      "eval_samples_per_second": 152.759,
      "eval_steps_per_second": 1.197,
      "step": 936
    },
    {
      "epoch": 3.2051282051282053,
      "grad_norm": 12.551536560058594,
      "learning_rate": 7.192307692307693e-06,
      "loss": 0.8403,
      "step": 1000
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5965441028732168,
      "eval_loss": 0.8640562891960144,
      "eval_runtime": 65.1621,
      "eval_samples_per_second": 152.757,
      "eval_steps_per_second": 1.197,
      "step": 1248
    },
    {
      "epoch": 4.8076923076923075,
      "grad_norm": 8.695352554321289,
      "learning_rate": 7.820512820512821e-07,
      "loss": 0.7812,
      "step": 1500
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6231665662045409,
      "eval_loss": 0.8301976323127747,
      "eval_runtime": 65.1825,
      "eval_samples_per_second": 152.71,
      "eval_steps_per_second": 1.197,
      "step": 1560
    }
  ],
  "logging_steps": 500,
  "max_steps": 1560,
  "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": 1.309515480032256e+16,
  "train_batch_size": 128,
  "trial_name": null,
  "trial_params": null
}