trainer_state.json

{
  "best_global_step": 20,
  "best_metric": 0.699292778968811,
  "best_model_checkpoint": "distilbert_km_50_v2_rte/checkpoint-20",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 70,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.335757851600647,
      "learning_rate": 4.9e-05,
      "loss": 0.7065,
      "step": 10
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.4729241877256318,
      "eval_loss": 0.7252510190010071,
      "eval_runtime": 0.1272,
      "eval_samples_per_second": 2178.161,
      "eval_steps_per_second": 15.727,
      "step": 10
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.8837642073631287,
      "learning_rate": 4.8e-05,
      "loss": 0.6735,
      "step": 20
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5379061371841155,
      "eval_loss": 0.699292778968811,
      "eval_runtime": 0.1245,
      "eval_samples_per_second": 2224.507,
      "eval_steps_per_second": 16.061,
      "step": 20
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.1381245851516724,
      "learning_rate": 4.7e-05,
      "loss": 0.6465,
      "step": 30
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.4981949458483754,
      "eval_loss": 0.7214222550392151,
      "eval_runtime": 0.1319,
      "eval_samples_per_second": 2100.238,
      "eval_steps_per_second": 15.164,
      "step": 30
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.9131489992141724,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.5982,
      "step": 40
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5342960288808665,
      "eval_loss": 0.7387977242469788,
      "eval_runtime": 0.1344,
      "eval_samples_per_second": 2060.717,
      "eval_steps_per_second": 14.879,
      "step": 40
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.0394362211227417,
      "learning_rate": 4.5e-05,
      "loss": 0.5339,
      "step": 50
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5306859205776173,
      "eval_loss": 0.7753897309303284,
      "eval_runtime": 0.1238,
      "eval_samples_per_second": 2238.203,
      "eval_steps_per_second": 16.16,
      "step": 50
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.1707080602645874,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.4565,
      "step": 60
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.5090252707581228,
      "eval_loss": 0.8549531698226929,
      "eval_runtime": 0.1488,
      "eval_samples_per_second": 1861.468,
      "eval_steps_per_second": 13.44,
      "step": 60
    },
    {
      "epoch": 7.0,
      "grad_norm": 1.2546998262405396,
      "learning_rate": 4.3e-05,
      "loss": 0.368,
      "step": 70
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.5126353790613718,
      "eval_loss": 0.9558070302009583,
      "eval_runtime": 0.1307,
      "eval_samples_per_second": 2119.309,
      "eval_steps_per_second": 15.302,
      "step": 70
    },
    {
      "epoch": 7.0,
      "step": 70,
      "total_flos": 1154453379287040.0,
      "train_loss": 0.5690259763172695,
      "train_runtime": 26.0206,
      "train_samples_per_second": 4784.668,
      "train_steps_per_second": 19.216
    }
  ],
  "logging_steps": 1,
  "max_steps": 500,
  "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": 1154453379287040.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}