trainer_state.json

{
  "best_global_step": 10,
  "best_metric": 0.6984360814094543,
  "best_model_checkpoint": "bert_base_km_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": 0.9742558002471924,
      "learning_rate": 4.9e-05,
      "loss": 0.7036,
      "step": 10
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5090252707581228,
      "eval_loss": 0.6984360814094543,
      "eval_runtime": 0.1976,
      "eval_samples_per_second": 1401.721,
      "eval_steps_per_second": 10.121,
      "step": 10
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.1415458917617798,
      "learning_rate": 4.8e-05,
      "loss": 0.66,
      "step": 20
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5451263537906137,
      "eval_loss": 0.7282441854476929,
      "eval_runtime": 0.2029,
      "eval_samples_per_second": 1365.505,
      "eval_steps_per_second": 9.859,
      "step": 20
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.28670334815979,
      "learning_rate": 4.7e-05,
      "loss": 0.6065,
      "step": 30
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.516245487364621,
      "eval_loss": 0.7693426609039307,
      "eval_runtime": 0.1975,
      "eval_samples_per_second": 1402.229,
      "eval_steps_per_second": 10.124,
      "step": 30
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.6812015771865845,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.4844,
      "step": 40
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.4693140794223827,
      "eval_loss": 0.9045701026916504,
      "eval_runtime": 0.1918,
      "eval_samples_per_second": 1444.226,
      "eval_steps_per_second": 10.428,
      "step": 40
    },
    {
      "epoch": 5.0,
      "grad_norm": 3.42217755317688,
      "learning_rate": 4.5e-05,
      "loss": 0.3615,
      "step": 50
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5379061371841155,
      "eval_loss": 1.0748211145401,
      "eval_runtime": 0.1984,
      "eval_samples_per_second": 1396.461,
      "eval_steps_per_second": 10.083,
      "step": 50
    },
    {
      "epoch": 6.0,
      "grad_norm": 3.6273467540740967,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.26,
      "step": 60
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.4693140794223827,
      "eval_loss": 1.225681185722351,
      "eval_runtime": 0.1903,
      "eval_samples_per_second": 1455.261,
      "eval_steps_per_second": 10.507,
      "step": 60
    },
    {
      "epoch": 6.0,
      "step": 60,
      "total_flos": 1965439583539200.0,
      "train_loss": 0.5126757184664409,
      "train_runtime": 37.3675,
      "train_samples_per_second": 3331.771,
      "train_steps_per_second": 13.381
    }
  ],
  "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
}