trainer_state.json

{
  "best_global_step": 10,
  "best_metric": 0.6921324729919434,
  "best_model_checkpoint": "bert_base_rand_50_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": 0.5368039608001709,
      "learning_rate": 4.9e-05,
      "loss": 0.6947,
      "step": 10
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5234657039711191,
      "eval_loss": 0.6921324729919434,
      "eval_runtime": 0.1956,
      "eval_samples_per_second": 1416.49,
      "eval_steps_per_second": 10.227,
      "step": 10
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.2625293731689453,
      "learning_rate": 4.8e-05,
      "loss": 0.678,
      "step": 20
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5126353790613718,
      "eval_loss": 0.7364938259124756,
      "eval_runtime": 0.1903,
      "eval_samples_per_second": 1455.762,
      "eval_steps_per_second": 10.511,
      "step": 20
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.143974542617798,
      "learning_rate": 4.7e-05,
      "loss": 0.6121,
      "step": 30
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5054151624548736,
      "eval_loss": 0.8218866586685181,
      "eval_runtime": 0.2324,
      "eval_samples_per_second": 1192.039,
      "eval_steps_per_second": 8.607,
      "step": 30
    },
    {
      "epoch": 4.0,
      "grad_norm": 3.1051809787750244,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.4991,
      "step": 40
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.48375451263537905,
      "eval_loss": 1.0314902067184448,
      "eval_runtime": 0.1962,
      "eval_samples_per_second": 1411.577,
      "eval_steps_per_second": 10.192,
      "step": 40
    },
    {
      "epoch": 5.0,
      "grad_norm": 3.6076419353485107,
      "learning_rate": 4.5e-05,
      "loss": 0.35,
      "step": 50
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.4693140794223827,
      "eval_loss": 1.1965441703796387,
      "eval_runtime": 0.1924,
      "eval_samples_per_second": 1439.97,
      "eval_steps_per_second": 10.397,
      "step": 50
    },
    {
      "epoch": 6.0,
      "grad_norm": 8.849884986877441,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.2498,
      "step": 60
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.48736462093862815,
      "eval_loss": 1.7301607131958008,
      "eval_runtime": 0.1936,
      "eval_samples_per_second": 1430.557,
      "eval_steps_per_second": 10.329,
      "step": 60
    },
    {
      "epoch": 6.0,
      "step": 60,
      "total_flos": 1965439583539200.0,
      "train_loss": 0.5139366706212362,
      "train_runtime": 37.1849,
      "train_samples_per_second": 3348.133,
      "train_steps_per_second": 13.446
    }
  ],
  "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
}