trainer_state.json

{
  "best_global_step": 20,
  "best_metric": 0.7117660045623779,
  "best_model_checkpoint": "distilbert_km_20_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": 1.4335824251174927,
      "learning_rate": 4.9e-05,
      "loss": 0.7025,
      "step": 10
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.51985559566787,
      "eval_loss": 0.7155982851982117,
      "eval_runtime": 0.1559,
      "eval_samples_per_second": 1777.006,
      "eval_steps_per_second": 12.83,
      "step": 10
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.13246488571167,
      "learning_rate": 4.8e-05,
      "loss": 0.667,
      "step": 20
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.516245487364621,
      "eval_loss": 0.7117660045623779,
      "eval_runtime": 0.1311,
      "eval_samples_per_second": 2112.757,
      "eval_steps_per_second": 15.255,
      "step": 20
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.9504374861717224,
      "learning_rate": 4.7e-05,
      "loss": 0.6331,
      "step": 30
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5342960288808665,
      "eval_loss": 0.7130563259124756,
      "eval_runtime": 0.1583,
      "eval_samples_per_second": 1750.352,
      "eval_steps_per_second": 12.638,
      "step": 30
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.0548288822174072,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.581,
      "step": 40
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.51985559566787,
      "eval_loss": 0.7594307065010071,
      "eval_runtime": 0.1331,
      "eval_samples_per_second": 2080.538,
      "eval_steps_per_second": 15.022,
      "step": 40
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.0273826122283936,
      "learning_rate": 4.5e-05,
      "loss": 0.5128,
      "step": 50
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.516245487364621,
      "eval_loss": 0.8006331324577332,
      "eval_runtime": 0.1283,
      "eval_samples_per_second": 2159.14,
      "eval_steps_per_second": 15.589,
      "step": 50
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.810105323791504,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.4258,
      "step": 60
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.4657039711191336,
      "eval_loss": 0.8966765403747559,
      "eval_runtime": 0.1443,
      "eval_samples_per_second": 1920.218,
      "eval_steps_per_second": 13.864,
      "step": 60
    },
    {
      "epoch": 7.0,
      "grad_norm": 1.4864951372146606,
      "learning_rate": 4.3e-05,
      "loss": 0.3149,
      "step": 70
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.4693140794223827,
      "eval_loss": 1.0948337316513062,
      "eval_runtime": 0.1342,
      "eval_samples_per_second": 2064.383,
      "eval_steps_per_second": 14.905,
      "step": 70
    },
    {
      "epoch": 7.0,
      "step": 70,
      "total_flos": 1154453379287040.0,
      "train_loss": 0.548187266077314,
      "train_runtime": 26.4605,
      "train_samples_per_second": 4705.12,
      "train_steps_per_second": 18.896
    }
  ],
  "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
}