trainer_state.json

{
  "best_metric": 0.6178625822067261,
  "best_model_checkpoint": "bert_base_lda_100_v1_cola/checkpoint-102",
  "epoch": 8.0,
  "eval_steps": 500,
  "global_step": 272,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.8181986808776855,
      "learning_rate": 4.9e-05,
      "loss": 0.6207,
      "step": 34
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.6229906678199768,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.7078,
      "eval_samples_per_second": 1473.659,
      "eval_steps_per_second": 7.065,
      "step": 34
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.3978294134140015,
      "learning_rate": 4.8e-05,
      "loss": 0.6135,
      "step": 68
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.6196237802505493,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.6873,
      "eval_samples_per_second": 1517.614,
      "eval_steps_per_second": 7.275,
      "step": 68
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.9987629652023315,
      "learning_rate": 4.7e-05,
      "loss": 0.6074,
      "step": 102
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.6178625822067261,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.6803,
      "eval_samples_per_second": 1533.121,
      "eval_steps_per_second": 7.35,
      "step": 102
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.9304279088974,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.5969,
      "step": 136
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.6277948617935181,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.6899,
      "eval_samples_per_second": 1511.79,
      "eval_steps_per_second": 7.247,
      "step": 136
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.3234096765518188,
      "learning_rate": 4.5e-05,
      "loss": 0.5522,
      "step": 170
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6615532040596008,
      "eval_loss": 0.623870849609375,
      "eval_matthews_correlation": 0.055798711473707796,
      "eval_runtime": 0.6815,
      "eval_samples_per_second": 1530.554,
      "eval_steps_per_second": 7.337,
      "step": 170
    },
    {
      "epoch": 6.0,
      "grad_norm": 2.008174419403076,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.4812,
      "step": 204
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6845637559890747,
      "eval_loss": 0.6719733476638794,
      "eval_matthews_correlation": 0.12095523242891448,
      "eval_runtime": 0.6877,
      "eval_samples_per_second": 1516.73,
      "eval_steps_per_second": 7.271,
      "step": 204
    },
    {
      "epoch": 7.0,
      "grad_norm": 2.082489490509033,
      "learning_rate": 4.3e-05,
      "loss": 0.416,
      "step": 238
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.6251198649406433,
      "eval_loss": 0.7045519351959229,
      "eval_matthews_correlation": 0.10103917683294436,
      "eval_runtime": 0.6722,
      "eval_samples_per_second": 1551.541,
      "eval_steps_per_second": 7.438,
      "step": 238
    },
    {
      "epoch": 8.0,
      "grad_norm": 3.2819783687591553,
      "learning_rate": 4.2e-05,
      "loss": 0.3693,
      "step": 272
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.6558005809783936,
      "eval_loss": 0.8737056255340576,
      "eval_matthews_correlation": 0.1049535667908795,
      "eval_runtime": 0.6691,
      "eval_samples_per_second": 1558.794,
      "eval_steps_per_second": 7.473,
      "step": 272
    },
    {
      "epoch": 8.0,
      "step": 272,
      "total_flos": 8999450537533440.0,
      "train_loss": 0.5321676205186283,
      "train_runtime": 137.2005,
      "train_samples_per_second": 3116.242,
      "train_steps_per_second": 12.391
    }
  ],
  "logging_steps": 1,
  "max_steps": 1700,
  "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": 8999450537533440.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}