trainer_state.json

{
  "best_global_step": 264,
  "best_metric": 0.4363076388835907,
  "best_model_checkpoint": "bert_base_km_20_v1_sst2/checkpoint-264",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 1584,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 6.855026721954346,
      "learning_rate": 4.9e-05,
      "loss": 0.4071,
      "step": 264
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8084862385321101,
      "eval_loss": 0.4363076388835907,
      "eval_runtime": 0.5669,
      "eval_samples_per_second": 1538.269,
      "eval_steps_per_second": 7.056,
      "step": 264
    },
    {
      "epoch": 2.0,
      "grad_norm": 5.597779750823975,
      "learning_rate": 4.8e-05,
      "loss": 0.2234,
      "step": 528
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8027522935779816,
      "eval_loss": 0.5351740121841431,
      "eval_runtime": 0.5619,
      "eval_samples_per_second": 1551.884,
      "eval_steps_per_second": 7.119,
      "step": 528
    },
    {
      "epoch": 3.0,
      "grad_norm": 6.781604766845703,
      "learning_rate": 4.7e-05,
      "loss": 0.1578,
      "step": 792
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7981651376146789,
      "eval_loss": 0.7095787525177002,
      "eval_runtime": 0.5629,
      "eval_samples_per_second": 1549.196,
      "eval_steps_per_second": 7.106,
      "step": 792
    },
    {
      "epoch": 4.0,
      "grad_norm": 7.563427448272705,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.1145,
      "step": 1056
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7981651376146789,
      "eval_loss": 0.5919294953346252,
      "eval_runtime": 0.5621,
      "eval_samples_per_second": 1551.216,
      "eval_steps_per_second": 7.116,
      "step": 1056
    },
    {
      "epoch": 5.0,
      "grad_norm": 7.101842880249023,
      "learning_rate": 4.5e-05,
      "loss": 0.0841,
      "step": 1320
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.783256880733945,
      "eval_loss": 0.7547828555107117,
      "eval_runtime": 0.5846,
      "eval_samples_per_second": 1491.658,
      "eval_steps_per_second": 6.842,
      "step": 1320
    },
    {
      "epoch": 6.0,
      "grad_norm": 6.3622589111328125,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.0635,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7798165137614679,
      "eval_loss": 0.901915967464447,
      "eval_runtime": 0.5612,
      "eval_samples_per_second": 1553.681,
      "eval_steps_per_second": 7.127,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "step": 1584,
      "total_flos": 5.316079940232192e+16,
      "train_loss": 0.17508280758905892,
      "train_runtime": 659.8178,
      "train_samples_per_second": 5103.606,
      "train_steps_per_second": 20.006
    }
  ],
  "logging_steps": 1,
  "max_steps": 13200,
  "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": 5.316079940232192e+16,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}