trainer_state.json

{
  "best_metric": 0.3705822825431824,
  "best_model_checkpoint": "distilbert_lda_5_v1_qnli/checkpoint-820",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 2870,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 4.038874626159668,
      "learning_rate": 4.9e-05,
      "loss": 0.497,
      "step": 410
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8303130148270181,
      "eval_loss": 0.38713347911834717,
      "eval_runtime": 2.272,
      "eval_samples_per_second": 2404.487,
      "eval_steps_per_second": 9.683,
      "step": 410
    },
    {
      "epoch": 2.0,
      "grad_norm": 4.284979820251465,
      "learning_rate": 4.8e-05,
      "loss": 0.3635,
      "step": 820
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8434925864909391,
      "eval_loss": 0.3705822825431824,
      "eval_runtime": 2.3214,
      "eval_samples_per_second": 2353.364,
      "eval_steps_per_second": 9.477,
      "step": 820
    },
    {
      "epoch": 3.0,
      "grad_norm": 4.197302341461182,
      "learning_rate": 4.7e-05,
      "loss": 0.2745,
      "step": 1230
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8376349990847519,
      "eval_loss": 0.37501493096351624,
      "eval_runtime": 2.303,
      "eval_samples_per_second": 2372.144,
      "eval_steps_per_second": 9.553,
      "step": 1230
    },
    {
      "epoch": 4.0,
      "grad_norm": 6.512701511383057,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.1968,
      "step": 1640
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8308621636463481,
      "eval_loss": 0.45666348934173584,
      "eval_runtime": 2.3076,
      "eval_samples_per_second": 2367.403,
      "eval_steps_per_second": 9.534,
      "step": 1640
    },
    {
      "epoch": 5.0,
      "grad_norm": 4.087521553039551,
      "learning_rate": 4.5e-05,
      "loss": 0.1394,
      "step": 2050
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8321435108914516,
      "eval_loss": 0.5239177346229553,
      "eval_runtime": 2.333,
      "eval_samples_per_second": 2341.607,
      "eval_steps_per_second": 9.43,
      "step": 2050
    },
    {
      "epoch": 6.0,
      "grad_norm": 8.565244674682617,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.1001,
      "step": 2460
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8341570565623284,
      "eval_loss": 0.5522398352622986,
      "eval_runtime": 2.2746,
      "eval_samples_per_second": 2401.711,
      "eval_steps_per_second": 9.672,
      "step": 2460
    },
    {
      "epoch": 7.0,
      "grad_norm": 2.2605297565460205,
      "learning_rate": 4.3e-05,
      "loss": 0.0764,
      "step": 2870
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8121911037891268,
      "eval_loss": 0.6609876155853271,
      "eval_runtime": 2.3532,
      "eval_samples_per_second": 2321.488,
      "eval_steps_per_second": 9.349,
      "step": 2870
    },
    {
      "epoch": 7.0,
      "step": 2870,
      "total_flos": 4.856261458098893e+16,
      "train_loss": 0.23536453379986594,
      "train_runtime": 707.5282,
      "train_samples_per_second": 7402.037,
      "train_steps_per_second": 28.974
    }
  ],
  "logging_steps": 1,
  "max_steps": 20500,
  "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": 4.856261458098893e+16,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}