trainer_state.json

{
  "best_global_step": 34,
  "best_metric": 0.621063768863678,
  "best_model_checkpoint": "bert_base_km_20_v2_cola/checkpoint-34",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 204,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.376166582107544,
      "learning_rate": 4.9e-05,
      "loss": 0.6167,
      "step": 34
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.621063768863678,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.6946,
      "eval_samples_per_second": 1501.655,
      "eval_steps_per_second": 7.199,
      "step": 34
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.7012274265289307,
      "learning_rate": 4.8e-05,
      "loss": 0.5958,
      "step": 68
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6893576383590698,
      "eval_loss": 0.6219270825386047,
      "eval_matthews_correlation": 0.03026012901076368,
      "eval_runtime": 0.6717,
      "eval_samples_per_second": 1552.861,
      "eval_steps_per_second": 7.444,
      "step": 68
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.7330247163772583,
      "learning_rate": 4.7e-05,
      "loss": 0.557,
      "step": 102
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.665388286113739,
      "eval_loss": 0.6299116015434265,
      "eval_matthews_correlation": 0.026175355246044623,
      "eval_runtime": 0.6905,
      "eval_samples_per_second": 1510.598,
      "eval_steps_per_second": 7.242,
      "step": 102
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.546388864517212,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.5014,
      "step": 136
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6673058271408081,
      "eval_loss": 0.6749251484870911,
      "eval_matthews_correlation": 0.05449082428560162,
      "eval_runtime": 0.6849,
      "eval_samples_per_second": 1522.792,
      "eval_steps_per_second": 7.3,
      "step": 136
    },
    {
      "epoch": 5.0,
      "grad_norm": 3.188899278640747,
      "learning_rate": 4.5e-05,
      "loss": 0.4346,
      "step": 170
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6433365345001221,
      "eval_loss": 0.7010342478752136,
      "eval_matthews_correlation": 0.05917342054031377,
      "eval_runtime": 0.6738,
      "eval_samples_per_second": 1547.853,
      "eval_steps_per_second": 7.42,
      "step": 170
    },
    {
      "epoch": 6.0,
      "grad_norm": 3.2128183841705322,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.3592,
      "step": 204
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6318312287330627,
      "eval_loss": 0.7965663075447083,
      "eval_matthews_correlation": 0.06675062723029203,
      "eval_runtime": 0.6912,
      "eval_samples_per_second": 1508.938,
      "eval_steps_per_second": 7.234,
      "step": 204
    },
    {
      "epoch": 6.0,
      "step": 204,
      "total_flos": 6749587903150080.0,
      "train_loss": 0.5107854113859289,
      "train_runtime": 101.2199,
      "train_samples_per_second": 4223.971,
      "train_steps_per_second": 16.795
    }
  ],
  "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": 6749587903150080.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}