trainer_state.json

{
  "best_global_step": 264,
  "best_metric": 0.45707353949546814,
  "best_model_checkpoint": "bert_base_km_100_v2_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": 8.074118614196777,
      "learning_rate": 4.9e-05,
      "loss": 0.4294,
      "step": 264
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8084862385321101,
      "eval_loss": 0.45707353949546814,
      "eval_runtime": 0.5857,
      "eval_samples_per_second": 1488.734,
      "eval_steps_per_second": 6.829,
      "step": 264
    },
    {
      "epoch": 2.0,
      "grad_norm": 9.72170352935791,
      "learning_rate": 4.8e-05,
      "loss": 0.2304,
      "step": 528
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7924311926605505,
      "eval_loss": 0.5690664649009705,
      "eval_runtime": 0.5784,
      "eval_samples_per_second": 1507.691,
      "eval_steps_per_second": 6.916,
      "step": 528
    },
    {
      "epoch": 3.0,
      "grad_norm": 7.4511284828186035,
      "learning_rate": 4.7e-05,
      "loss": 0.1661,
      "step": 792
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7821100917431193,
      "eval_loss": 0.6635616421699524,
      "eval_runtime": 0.588,
      "eval_samples_per_second": 1482.892,
      "eval_steps_per_second": 6.802,
      "step": 792
    },
    {
      "epoch": 4.0,
      "grad_norm": 4.683515548706055,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.1227,
      "step": 1056
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7901376146788991,
      "eval_loss": 0.651644229888916,
      "eval_runtime": 0.5658,
      "eval_samples_per_second": 1541.306,
      "eval_steps_per_second": 7.07,
      "step": 1056
    },
    {
      "epoch": 5.0,
      "grad_norm": 5.742991924285889,
      "learning_rate": 4.5e-05,
      "loss": 0.092,
      "step": 1320
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8004587155963303,
      "eval_loss": 0.6851325631141663,
      "eval_runtime": 0.5631,
      "eval_samples_per_second": 1548.519,
      "eval_steps_per_second": 7.103,
      "step": 1320
    },
    {
      "epoch": 6.0,
      "grad_norm": 2.8473849296569824,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.0687,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7935779816513762,
      "eval_loss": 0.8138066530227661,
      "eval_runtime": 0.5617,
      "eval_samples_per_second": 1552.347,
      "eval_steps_per_second": 7.121,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "step": 1584,
      "total_flos": 5.316079940232192e+16,
      "train_loss": 0.18486634288171325,
      "train_runtime": 665.9432,
      "train_samples_per_second": 5056.662,
      "train_steps_per_second": 19.822
    }
  ],
  "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
}