trainer_state.json

{
  "best_global_step": 264,
  "best_metric": 0.43955448269844055,
  "best_model_checkpoint": "distilbert_rand_50_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": 7.4908223152160645,
      "learning_rate": 4.9e-05,
      "loss": 0.3925,
      "step": 264
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8245412844036697,
      "eval_loss": 0.43955448269844055,
      "eval_runtime": 0.3705,
      "eval_samples_per_second": 2353.603,
      "eval_steps_per_second": 10.796,
      "step": 264
    },
    {
      "epoch": 2.0,
      "grad_norm": 8.066856384277344,
      "learning_rate": 4.8e-05,
      "loss": 0.219,
      "step": 528
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8268348623853211,
      "eval_loss": 0.47368353605270386,
      "eval_runtime": 0.3774,
      "eval_samples_per_second": 2310.281,
      "eval_steps_per_second": 10.598,
      "step": 528
    },
    {
      "epoch": 3.0,
      "grad_norm": 5.2119340896606445,
      "learning_rate": 4.7e-05,
      "loss": 0.1609,
      "step": 792
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8084862385321101,
      "eval_loss": 0.49463415145874023,
      "eval_runtime": 0.374,
      "eval_samples_per_second": 2331.799,
      "eval_steps_per_second": 10.696,
      "step": 792
    },
    {
      "epoch": 4.0,
      "grad_norm": 5.589087009429932,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.1177,
      "step": 1056
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8176605504587156,
      "eval_loss": 0.6155182719230652,
      "eval_runtime": 0.3725,
      "eval_samples_per_second": 2340.769,
      "eval_steps_per_second": 10.737,
      "step": 1056
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.820615768432617,
      "learning_rate": 4.5e-05,
      "loss": 0.0889,
      "step": 1320
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8107798165137615,
      "eval_loss": 0.6215694546699524,
      "eval_runtime": 0.3663,
      "eval_samples_per_second": 2380.263,
      "eval_steps_per_second": 10.919,
      "step": 1320
    },
    {
      "epoch": 6.0,
      "grad_norm": 5.1633477210998535,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.0698,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8061926605504587,
      "eval_loss": 0.765625,
      "eval_runtime": 0.3636,
      "eval_samples_per_second": 2398.386,
      "eval_steps_per_second": 11.002,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "step": 1584,
      "total_flos": 2.676464049624883e+16,
      "train_loss": 0.1747924005142366,
      "train_runtime": 379.2232,
      "train_samples_per_second": 8879.864,
      "train_steps_per_second": 34.808
    }
  ],
  "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": 2.676464049624883e+16,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}