trainer_state.json

{
  "best_metric": 0.6853141784667969,
  "best_model_checkpoint": "hBERTv2_new_pretrain_rte/checkpoint-60",
  "epoch": 8.0,
  "global_step": 160,
  "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": 0.7512,
      "step": 20
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.4981949458483754,
      "eval_loss": 0.6928728818893433,
      "eval_runtime": 0.4852,
      "eval_samples_per_second": 570.956,
      "eval_steps_per_second": 6.184,
      "step": 20
    },
    {
      "epoch": 2.0,
      "learning_rate": 3.8400000000000005e-05,
      "loss": 0.7204,
      "step": 40
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5270758122743683,
      "eval_loss": 0.6908421516418457,
      "eval_runtime": 0.4837,
      "eval_samples_per_second": 572.644,
      "eval_steps_per_second": 6.202,
      "step": 40
    },
    {
      "epoch": 3.0,
      "learning_rate": 3.76e-05,
      "loss": 0.701,
      "step": 60
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5595667870036101,
      "eval_loss": 0.6853141784667969,
      "eval_runtime": 0.479,
      "eval_samples_per_second": 578.295,
      "eval_steps_per_second": 6.263,
      "step": 60
    },
    {
      "epoch": 4.0,
      "learning_rate": 3.680000000000001e-05,
      "loss": 0.6315,
      "step": 80
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5631768953068592,
      "eval_loss": 0.7080695033073425,
      "eval_runtime": 0.4845,
      "eval_samples_per_second": 571.714,
      "eval_steps_per_second": 6.192,
      "step": 80
    },
    {
      "epoch": 5.0,
      "learning_rate": 3.6e-05,
      "loss": 0.5807,
      "step": 100
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5342960288808665,
      "eval_loss": 0.8746474981307983,
      "eval_runtime": 0.4837,
      "eval_samples_per_second": 572.719,
      "eval_steps_per_second": 6.203,
      "step": 100
    },
    {
      "epoch": 6.0,
      "learning_rate": 3.52e-05,
      "loss": 0.4079,
      "step": 120
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.5631768953068592,
      "eval_loss": 0.8831439018249512,
      "eval_runtime": 0.4787,
      "eval_samples_per_second": 578.592,
      "eval_steps_per_second": 6.266,
      "step": 120
    },
    {
      "epoch": 7.0,
      "learning_rate": 3.44e-05,
      "loss": 0.3077,
      "step": 140
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.5487364620938628,
      "eval_loss": 1.0778553485870361,
      "eval_runtime": 0.479,
      "eval_samples_per_second": 578.245,
      "eval_steps_per_second": 6.263,
      "step": 140
    },
    {
      "epoch": 8.0,
      "learning_rate": 3.3600000000000004e-05,
      "loss": 0.2453,
      "step": 160
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.5415162454873647,
      "eval_loss": 1.1809593439102173,
      "eval_runtime": 0.479,
      "eval_samples_per_second": 578.29,
      "eval_steps_per_second": 6.263,
      "step": 160
    },
    {
      "epoch": 8.0,
      "step": 160,
      "total_flos": 2915076176609280.0,
      "train_loss": 0.5432160198688507,
      "train_runtime": 143.285,
      "train_samples_per_second": 868.898,
      "train_steps_per_second": 6.979
    }
  ],
  "max_steps": 1000,
  "num_train_epochs": 50,
  "total_flos": 2915076176609280.0,
  "trial_name": null,
  "trial_params": null
}