trainer_state.json

{
  "best_global_step": 10,
  "best_metric": 0.6880781650543213,
  "best_model_checkpoint": "bert_base_rand_5_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": 1.3935413360595703,
      "learning_rate": 4.9e-05,
      "loss": 0.696,
      "step": 10
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5126353790613718,
      "eval_loss": 0.6880781650543213,
      "eval_runtime": 0.1889,
      "eval_samples_per_second": 1466.288,
      "eval_steps_per_second": 10.587,
      "step": 10
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.137573719024658,
      "learning_rate": 4.8e-05,
      "loss": 0.6677,
      "step": 20
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.51985559566787,
      "eval_loss": 0.7654381394386292,
      "eval_runtime": 0.2023,
      "eval_samples_per_second": 1369.56,
      "eval_steps_per_second": 9.889,
      "step": 20
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.546442985534668,
      "learning_rate": 4.7e-05,
      "loss": 0.5906,
      "step": 30
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5306859205776173,
      "eval_loss": 0.8287155032157898,
      "eval_runtime": 0.1921,
      "eval_samples_per_second": 1441.968,
      "eval_steps_per_second": 10.411,
      "step": 30
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.615215539932251,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.4538,
      "step": 40
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.49097472924187724,
      "eval_loss": 1.0560897588729858,
      "eval_runtime": 0.1943,
      "eval_samples_per_second": 1425.781,
      "eval_steps_per_second": 10.294,
      "step": 40
    },
    {
      "epoch": 5.0,
      "grad_norm": 4.1631622314453125,
      "learning_rate": 4.5e-05,
      "loss": 0.3089,
      "step": 50
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5415162454873647,
      "eval_loss": 1.1951771974563599,
      "eval_runtime": 0.1886,
      "eval_samples_per_second": 1468.542,
      "eval_steps_per_second": 10.603,
      "step": 50
    },
    {
      "epoch": 6.0,
      "grad_norm": 5.564742088317871,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.2171,
      "step": 60
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.5126353790613718,
      "eval_loss": 1.4987684488296509,
      "eval_runtime": 0.2236,
      "eval_samples_per_second": 1238.861,
      "eval_steps_per_second": 8.945,
      "step": 60
    },
    {
      "epoch": 6.0,
      "step": 60,
      "total_flos": 1965439583539200.0,
      "train_loss": 0.48901474475860596,
      "train_runtime": 37.6671,
      "train_samples_per_second": 3305.275,
      "train_steps_per_second": 13.274
    }
  ],
  "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": 1965439583539200.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}