trainer_state.json

{
  "best_metric": 1.0963791608810425,
  "best_model_checkpoint": "hBERTv2_new_pretrain_mnli/checkpoint-6136",
  "epoch": 7.0,
  "global_step": 21476,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "learning_rate": 3.9200000000000004e-05,
      "loss": 1.1031,
      "step": 3068
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.3273560876209883,
      "eval_loss": 1.1007808446884155,
      "eval_runtime": 16.7433,
      "eval_samples_per_second": 586.203,
      "eval_steps_per_second": 4.599,
      "step": 3068
    },
    {
      "epoch": 2.0,
      "learning_rate": 3.8400000000000005e-05,
      "loss": 1.0989,
      "step": 6136
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.31818644931227713,
      "eval_loss": 1.0963791608810425,
      "eval_runtime": 16.7646,
      "eval_samples_per_second": 585.459,
      "eval_steps_per_second": 4.593,
      "step": 6136
    },
    {
      "epoch": 3.0,
      "learning_rate": 3.76e-05,
      "loss": 1.0987,
      "step": 9204
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.32654100866021396,
      "eval_loss": 1.0979726314544678,
      "eval_runtime": 16.7522,
      "eval_samples_per_second": 585.894,
      "eval_steps_per_second": 4.596,
      "step": 9204
    },
    {
      "epoch": 4.0,
      "learning_rate": 3.680000000000001e-05,
      "loss": 1.0987,
      "step": 12272
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.31818644931227713,
      "eval_loss": 1.0994985103607178,
      "eval_runtime": 16.749,
      "eval_samples_per_second": 586.006,
      "eval_steps_per_second": 4.597,
      "step": 12272
    },
    {
      "epoch": 5.0,
      "learning_rate": 3.6e-05,
      "loss": 1.0986,
      "step": 15340
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.3273560876209883,
      "eval_loss": 1.0985783338546753,
      "eval_runtime": 16.7396,
      "eval_samples_per_second": 586.335,
      "eval_steps_per_second": 4.6,
      "step": 15340
    },
    {
      "epoch": 6.0,
      "learning_rate": 3.52e-05,
      "loss": 1.0991,
      "step": 18408
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.3544574630667346,
      "eval_loss": 1.1015625,
      "eval_runtime": 16.7478,
      "eval_samples_per_second": 586.047,
      "eval_steps_per_second": 4.598,
      "step": 18408
    },
    {
      "epoch": 7.0,
      "learning_rate": 3.44e-05,
      "loss": 1.0991,
      "step": 21476
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.3273560876209883,
      "eval_loss": 1.099992036819458,
      "eval_runtime": 16.7418,
      "eval_samples_per_second": 586.259,
      "eval_steps_per_second": 4.599,
      "step": 21476
    },
    {
      "epoch": 7.0,
      "step": 21476,
      "total_flos": 4.022770393541509e+17,
      "train_loss": 1.099463806891268,
      "train_runtime": 13736.6021,
      "train_samples_per_second": 1429.4,
      "train_steps_per_second": 11.167
    }
  ],
  "max_steps": 153400,
  "num_train_epochs": 50,
  "total_flos": 4.022770393541509e+17,
  "trial_name": null,
  "trial_params": null
}