trainer_state.json

{
  "best_global_step": 10,
  "best_metric": 0.7072145938873291,
  "best_model_checkpoint": "distilbert_rand_10_v2_rte/checkpoint-10",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 60,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 2.0832531452178955,
      "learning_rate": 4.9e-05,
      "loss": 0.7299,
      "step": 10
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.48014440433212996,
      "eval_loss": 0.7072145938873291,
      "eval_runtime": 0.1371,
      "eval_samples_per_second": 2020.125,
      "eval_steps_per_second": 14.586,
      "step": 10
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.0068490505218506,
      "learning_rate": 4.8e-05,
      "loss": 0.6804,
      "step": 20
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5126353790613718,
      "eval_loss": 0.7083075642585754,
      "eval_runtime": 0.1292,
      "eval_samples_per_second": 2143.16,
      "eval_steps_per_second": 15.474,
      "step": 20
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.9002633690834045,
      "learning_rate": 4.7e-05,
      "loss": 0.6441,
      "step": 30
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5487364620938628,
      "eval_loss": 0.7337650656700134,
      "eval_runtime": 0.1414,
      "eval_samples_per_second": 1958.766,
      "eval_steps_per_second": 14.143,
      "step": 30
    },
    {
      "epoch": 4.0,
      "grad_norm": 3.7300801277160645,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.5634,
      "step": 40
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5054151624548736,
      "eval_loss": 0.8555057644844055,
      "eval_runtime": 0.1291,
      "eval_samples_per_second": 2145.979,
      "eval_steps_per_second": 15.494,
      "step": 40
    },
    {
      "epoch": 5.0,
      "grad_norm": 3.1470835208892822,
      "learning_rate": 4.5e-05,
      "loss": 0.434,
      "step": 50
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5379061371841155,
      "eval_loss": 0.9622868895530701,
      "eval_runtime": 0.1657,
      "eval_samples_per_second": 1671.559,
      "eval_steps_per_second": 12.069,
      "step": 50
    },
    {
      "epoch": 6.0,
      "grad_norm": 8.392926216125488,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.3059,
      "step": 60
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.5306859205776173,
      "eval_loss": 1.2856022119522095,
      "eval_runtime": 0.1462,
      "eval_samples_per_second": 1894.326,
      "eval_steps_per_second": 13.677,
      "step": 60
    },
    {
      "epoch": 6.0,
      "step": 60,
      "total_flos": 989531467960320.0,
      "train_loss": 0.559611996014913,
      "train_runtime": 23.2453,
      "train_samples_per_second": 5355.916,
      "train_steps_per_second": 21.51
    }
  ],
  "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": 989531467960320.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}