trainer_state.json

{
  "best_global_step": 20,
  "best_metric": 0.7014539241790771,
  "best_model_checkpoint": "distilbert_rand_5_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": 1.9665743112564087,
      "learning_rate": 4.9e-05,
      "loss": 0.729,
      "step": 10
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.4657039711191336,
      "eval_loss": 0.7057479619979858,
      "eval_runtime": 0.1544,
      "eval_samples_per_second": 1793.611,
      "eval_steps_per_second": 12.95,
      "step": 10
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.5788726210594177,
      "learning_rate": 4.8e-05,
      "loss": 0.6878,
      "step": 20
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5018050541516246,
      "eval_loss": 0.7014539241790771,
      "eval_runtime": 0.1433,
      "eval_samples_per_second": 1932.724,
      "eval_steps_per_second": 13.955,
      "step": 20
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.9091008305549622,
      "learning_rate": 4.7e-05,
      "loss": 0.6494,
      "step": 30
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5379061371841155,
      "eval_loss": 0.7377594709396362,
      "eval_runtime": 0.1392,
      "eval_samples_per_second": 1990.055,
      "eval_steps_per_second": 14.369,
      "step": 30
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.5707993507385254,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.567,
      "step": 40
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5234657039711191,
      "eval_loss": 0.8509384989738464,
      "eval_runtime": 0.1403,
      "eval_samples_per_second": 1974.411,
      "eval_steps_per_second": 14.256,
      "step": 40
    },
    {
      "epoch": 5.0,
      "grad_norm": 4.262302875518799,
      "learning_rate": 4.5e-05,
      "loss": 0.4454,
      "step": 50
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.4981949458483754,
      "eval_loss": 1.0468066930770874,
      "eval_runtime": 0.1352,
      "eval_samples_per_second": 2049.387,
      "eval_steps_per_second": 14.797,
      "step": 50
    },
    {
      "epoch": 6.0,
      "grad_norm": 3.614743232727051,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.304,
      "step": 60
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.5018050541516246,
      "eval_loss": 1.3210716247558594,
      "eval_runtime": 0.1464,
      "eval_samples_per_second": 1891.633,
      "eval_steps_per_second": 13.658,
      "step": 60
    },
    {
      "epoch": 7.0,
      "grad_norm": 3.6419012546539307,
      "learning_rate": 4.3e-05,
      "loss": 0.1812,
      "step": 70
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.48736462093862815,
      "eval_loss": 1.6215654611587524,
      "eval_runtime": 0.1263,
      "eval_samples_per_second": 2193.065,
      "eval_steps_per_second": 15.834,
      "step": 70
    },
    {
      "epoch": 7.0,
      "step": 70,
      "total_flos": 1154453379287040.0,
      "train_loss": 0.5091195225715637,
      "train_runtime": 26.6073,
      "train_samples_per_second": 4679.16,
      "train_steps_per_second": 18.792
    }
  ],
  "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
}