trainer_state.json

{
  "best_metric": 0.32672205567359924,
  "best_model_checkpoint": "bert_base_lda_20_v1_book_qnli/checkpoint-820",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 2870,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 3.378965377807617,
      "learning_rate": 4.9e-05,
      "loss": 0.4723,
      "step": 410
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8398315943620721,
      "eval_loss": 0.37053045630455017,
      "eval_runtime": 3.6347,
      "eval_samples_per_second": 1503.002,
      "eval_steps_per_second": 6.053,
      "step": 410
    },
    {
      "epoch": 2.0,
      "grad_norm": 5.707330226898193,
      "learning_rate": 4.8e-05,
      "loss": 0.3383,
      "step": 820
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.862529745561047,
      "eval_loss": 0.32672205567359924,
      "eval_runtime": 3.6323,
      "eval_samples_per_second": 1504.003,
      "eval_steps_per_second": 6.057,
      "step": 820
    },
    {
      "epoch": 3.0,
      "grad_norm": 4.867852210998535,
      "learning_rate": 4.7e-05,
      "loss": 0.2411,
      "step": 1230
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8628958447739338,
      "eval_loss": 0.33844906091690063,
      "eval_runtime": 3.6254,
      "eval_samples_per_second": 1506.873,
      "eval_steps_per_second": 6.068,
      "step": 1230
    },
    {
      "epoch": 4.0,
      "grad_norm": 4.006015300750732,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.1581,
      "step": 1640
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.847702727439136,
      "eval_loss": 0.4427230954170227,
      "eval_runtime": 3.5785,
      "eval_samples_per_second": 1526.615,
      "eval_steps_per_second": 6.148,
      "step": 1640
    },
    {
      "epoch": 5.0,
      "grad_norm": 5.629159927368164,
      "learning_rate": 4.5e-05,
      "loss": 0.1048,
      "step": 2050
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8513637195680029,
      "eval_loss": 0.5107147693634033,
      "eval_runtime": 3.5257,
      "eval_samples_per_second": 1549.498,
      "eval_steps_per_second": 6.24,
      "step": 2050
    },
    {
      "epoch": 6.0,
      "grad_norm": 12.798274993896484,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.0739,
      "step": 2460
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8574043565806333,
      "eval_loss": 0.5324944257736206,
      "eval_runtime": 3.6055,
      "eval_samples_per_second": 1515.2,
      "eval_steps_per_second": 6.102,
      "step": 2460
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.7117020487785339,
      "learning_rate": 4.3e-05,
      "loss": 0.0568,
      "step": 2870
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8451400329489291,
      "eval_loss": 0.6685991287231445,
      "eval_runtime": 3.5648,
      "eval_samples_per_second": 1532.48,
      "eval_steps_per_second": 6.171,
      "step": 2870
    },
    {
      "epoch": 7.0,
      "step": 2870,
      "total_flos": 9.645664445050368e+16,
      "train_loss": 0.20647664950699757,
      "train_runtime": 1243.2195,
      "train_samples_per_second": 4212.571,
      "train_steps_per_second": 16.489
    }
  ],
  "logging_steps": 1,
  "max_steps": 20500,
  "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": 9.645664445050368e+16,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}