trainer_state.json

{
  "best_global_step": 10,
  "best_metric": 0.6832129955291748,
  "best_model_checkpoint": "tiny_bert_rand_20_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": 0.6792683005332947,
      "learning_rate": 4.9e-05,
      "loss": 0.6969,
      "step": 10
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5415162454873647,
      "eval_loss": 0.6832129955291748,
      "eval_runtime": 0.1039,
      "eval_samples_per_second": 2665.451,
      "eval_steps_per_second": 19.245,
      "step": 10
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.8212109208106995,
      "learning_rate": 4.8e-05,
      "loss": 0.6608,
      "step": 20
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5487364620938628,
      "eval_loss": 0.6854340434074402,
      "eval_runtime": 0.1012,
      "eval_samples_per_second": 2737.693,
      "eval_steps_per_second": 19.767,
      "step": 20
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.8642770648002625,
      "learning_rate": 4.7e-05,
      "loss": 0.6234,
      "step": 30
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5342960288808665,
      "eval_loss": 0.6942301988601685,
      "eval_runtime": 0.1233,
      "eval_samples_per_second": 2246.148,
      "eval_steps_per_second": 16.218,
      "step": 30
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.2472058534622192,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.5633,
      "step": 40
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5270758122743683,
      "eval_loss": 0.731425940990448,
      "eval_runtime": 0.1093,
      "eval_samples_per_second": 2533.217,
      "eval_steps_per_second": 18.29,
      "step": 40
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.4485448598861694,
      "learning_rate": 4.5e-05,
      "loss": 0.4815,
      "step": 50
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5523465703971119,
      "eval_loss": 0.8199784159660339,
      "eval_runtime": 0.1057,
      "eval_samples_per_second": 2619.803,
      "eval_steps_per_second": 18.916,
      "step": 50
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.830456018447876,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.3824,
      "step": 60
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.5234657039711191,
      "eval_loss": 0.9300938248634338,
      "eval_runtime": 0.0983,
      "eval_samples_per_second": 2817.071,
      "eval_steps_per_second": 20.34,
      "step": 60
    },
    {
      "epoch": 6.0,
      "step": 60,
      "total_flos": 391780256993280.0,
      "train_loss": 0.5680781046549479,
      "train_runtime": 13.5748,
      "train_samples_per_second": 9171.378,
      "train_steps_per_second": 36.833
    }
  ],
  "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
}