trainer_state.json

{
  "best_global_step": 20,
  "best_metric": 0.6859487891197205,
  "best_model_checkpoint": "tiny_bert_rand_20_v2_rte/checkpoint-20",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 70,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.596297025680542,
      "learning_rate": 4.9e-05,
      "loss": 0.6973,
      "step": 10
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5451263537906137,
      "eval_loss": 0.6897563338279724,
      "eval_runtime": 0.0983,
      "eval_samples_per_second": 2818.254,
      "eval_steps_per_second": 20.348,
      "step": 10
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.6815950870513916,
      "learning_rate": 4.8e-05,
      "loss": 0.6885,
      "step": 20
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5703971119133574,
      "eval_loss": 0.6859487891197205,
      "eval_runtime": 0.1168,
      "eval_samples_per_second": 2371.569,
      "eval_steps_per_second": 17.123,
      "step": 20
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.5316508412361145,
      "learning_rate": 4.7e-05,
      "loss": 0.6679,
      "step": 30
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5703971119133574,
      "eval_loss": 0.7017534971237183,
      "eval_runtime": 0.0972,
      "eval_samples_per_second": 2848.546,
      "eval_steps_per_second": 20.567,
      "step": 30
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.6141400337219238,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.6325,
      "step": 40
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5487364620938628,
      "eval_loss": 0.7494218349456787,
      "eval_runtime": 0.0989,
      "eval_samples_per_second": 2802.11,
      "eval_steps_per_second": 20.232,
      "step": 40
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.5151808261871338,
      "learning_rate": 4.5e-05,
      "loss": 0.58,
      "step": 50
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5379061371841155,
      "eval_loss": 0.8294788002967834,
      "eval_runtime": 0.1094,
      "eval_samples_per_second": 2531.153,
      "eval_steps_per_second": 18.275,
      "step": 50
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.656488060951233,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.5061,
      "step": 60
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.5270758122743683,
      "eval_loss": 0.8910938501358032,
      "eval_runtime": 0.1145,
      "eval_samples_per_second": 2419.934,
      "eval_steps_per_second": 17.472,
      "step": 60
    },
    {
      "epoch": 7.0,
      "grad_norm": 2.2175965309143066,
      "learning_rate": 4.3e-05,
      "loss": 0.3931,
      "step": 70
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.4981949458483754,
      "eval_loss": 1.1315115690231323,
      "eval_runtime": 0.1083,
      "eval_samples_per_second": 2556.99,
      "eval_steps_per_second": 18.462,
      "step": 70
    },
    {
      "epoch": 7.0,
      "step": 70,
      "total_flos": 457076966492160.0,
      "train_loss": 0.595053608076913,
      "train_runtime": 15.2908,
      "train_samples_per_second": 8142.147,
      "train_steps_per_second": 32.699
    }
  ],
  "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": 457076966492160.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}