trainer_state.json

{
  "best_global_step": 10,
  "best_metric": 0.6917728781700134,
  "best_model_checkpoint": "tiny_bert_rand_10_v1_rte/checkpoint-10",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 60,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.118343472480774,
      "learning_rate": 4.9e-05,
      "loss": 0.7007,
      "step": 10
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.516245487364621,
      "eval_loss": 0.6917728781700134,
      "eval_runtime": 0.1044,
      "eval_samples_per_second": 2653.931,
      "eval_steps_per_second": 19.162,
      "step": 10
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.4658519923686981,
      "learning_rate": 4.8e-05,
      "loss": 0.6835,
      "step": 20
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5595667870036101,
      "eval_loss": 0.6919773817062378,
      "eval_runtime": 0.1027,
      "eval_samples_per_second": 2698.116,
      "eval_steps_per_second": 19.481,
      "step": 20
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.5628891587257385,
      "learning_rate": 4.7e-05,
      "loss": 0.6616,
      "step": 30
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5740072202166066,
      "eval_loss": 0.7088081240653992,
      "eval_runtime": 0.1004,
      "eval_samples_per_second": 2759.129,
      "eval_steps_per_second": 19.922,
      "step": 30
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.0160233974456787,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.6219,
      "step": 40
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5523465703971119,
      "eval_loss": 0.7802593111991882,
      "eval_runtime": 0.1098,
      "eval_samples_per_second": 2522.629,
      "eval_steps_per_second": 18.214,
      "step": 40
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.7022738456726074,
      "learning_rate": 4.5e-05,
      "loss": 0.5651,
      "step": 50
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5379061371841155,
      "eval_loss": 0.8309101462364197,
      "eval_runtime": 0.1032,
      "eval_samples_per_second": 2684.463,
      "eval_steps_per_second": 19.382,
      "step": 50
    },
    {
      "epoch": 6.0,
      "grad_norm": 3.098724603652954,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.4986,
      "step": 60
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.5126353790613718,
      "eval_loss": 0.9334157109260559,
      "eval_runtime": 0.095,
      "eval_samples_per_second": 2914.596,
      "eval_steps_per_second": 21.044,
      "step": 60
    },
    {
      "epoch": 6.0,
      "step": 60,
      "total_flos": 391780256993280.0,
      "train_loss": 0.6219117085138957,
      "train_runtime": 13.3103,
      "train_samples_per_second": 9353.64,
      "train_steps_per_second": 37.565
    }
  ],
  "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": 391780256993280.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}