trainer_state.json

{
  "best_global_step": 6,
  "best_metric": 0.7483770847320557,
  "best_model_checkpoint": "bert_base_km_5_v1_wnli/checkpoint-6",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 21,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.9307972192764282,
      "learning_rate": 4.9e-05,
      "loss": 0.7537,
      "step": 3
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.4084507042253521,
      "eval_loss": 0.7666977643966675,
      "eval_runtime": 0.0573,
      "eval_samples_per_second": 1239.276,
      "eval_steps_per_second": 17.455,
      "step": 3
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.165887713432312,
      "learning_rate": 4.8e-05,
      "loss": 0.7111,
      "step": 6
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.4507042253521127,
      "eval_loss": 0.7483770847320557,
      "eval_runtime": 0.0564,
      "eval_samples_per_second": 1257.822,
      "eval_steps_per_second": 17.716,
      "step": 6
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.135459065437317,
      "learning_rate": 4.7e-05,
      "loss": 0.6976,
      "step": 9
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.23943661971830985,
      "eval_loss": 0.7711266279220581,
      "eval_runtime": 0.0608,
      "eval_samples_per_second": 1167.858,
      "eval_steps_per_second": 16.449,
      "step": 9
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.0129027366638184,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.6883,
      "step": 12
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.22535211267605634,
      "eval_loss": 0.7964349985122681,
      "eval_runtime": 0.0566,
      "eval_samples_per_second": 1253.829,
      "eval_steps_per_second": 17.66,
      "step": 12
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.573878288269043,
      "learning_rate": 4.5e-05,
      "loss": 0.6862,
      "step": 15
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.23943661971830985,
      "eval_loss": 0.8099692463874817,
      "eval_runtime": 0.0554,
      "eval_samples_per_second": 1281.475,
      "eval_steps_per_second": 18.049,
      "step": 15
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.3363782167434692,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.6752,
      "step": 18
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.16901408450704225,
      "eval_loss": 0.8518927097320557,
      "eval_runtime": 0.0571,
      "eval_samples_per_second": 1242.596,
      "eval_steps_per_second": 17.501,
      "step": 18
    },
    {
      "epoch": 7.0,
      "grad_norm": 1.0673614740371704,
      "learning_rate": 4.3e-05,
      "loss": 0.6723,
      "step": 21
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.16901408450704225,
      "eval_loss": 0.8786036968231201,
      "eval_runtime": 0.0633,
      "eval_samples_per_second": 1122.393,
      "eval_steps_per_second": 15.808,
      "step": 21
    },
    {
      "epoch": 7.0,
      "step": 21,
      "total_flos": 584764320537600.0,
      "train_loss": 0.6977946417672294,
      "train_runtime": 21.9814,
      "train_samples_per_second": 1444.404,
      "train_steps_per_second": 6.824
    }
  ],
  "logging_steps": 1,
  "max_steps": 150,
  "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": 584764320537600.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}