trainer_state.json

{
  "best_metric": 0.704170286655426,
  "best_model_checkpoint": "bert_uncased_L-4_H-128_A-2_wnli/checkpoint-6",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 21,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.7738248705863953,
      "learning_rate": 4.9e-05,
      "loss": 0.6968,
      "step": 3
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.4225352112676056,
      "eval_loss": 0.7060407996177673,
      "eval_runtime": 0.0325,
      "eval_samples_per_second": 2182.852,
      "eval_steps_per_second": 30.744,
      "step": 3
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.40101733803749084,
      "learning_rate": 4.8e-05,
      "loss": 0.6898,
      "step": 6
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.352112676056338,
      "eval_loss": 0.704170286655426,
      "eval_runtime": 0.0317,
      "eval_samples_per_second": 2239.3,
      "eval_steps_per_second": 31.539,
      "step": 6
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.43040701746940613,
      "learning_rate": 4.7e-05,
      "loss": 0.6912,
      "step": 9
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.36619718309859156,
      "eval_loss": 0.705545961856842,
      "eval_runtime": 0.0392,
      "eval_samples_per_second": 1811.695,
      "eval_steps_per_second": 25.517,
      "step": 9
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.8497236371040344,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.695,
      "step": 12
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.3380281690140845,
      "eval_loss": 0.7075263857841492,
      "eval_runtime": 0.0342,
      "eval_samples_per_second": 2076.113,
      "eval_steps_per_second": 29.241,
      "step": 12
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.7524233460426331,
      "learning_rate": 4.5e-05,
      "loss": 0.6946,
      "step": 15
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.30985915492957744,
      "eval_loss": 0.709892213344574,
      "eval_runtime": 0.0299,
      "eval_samples_per_second": 2374.008,
      "eval_steps_per_second": 33.437,
      "step": 15
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.6198353171348572,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.6921,
      "step": 18
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.30985915492957744,
      "eval_loss": 0.7122030854225159,
      "eval_runtime": 0.0308,
      "eval_samples_per_second": 2307.312,
      "eval_steps_per_second": 32.497,
      "step": 18
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.444954514503479,
      "learning_rate": 4.3e-05,
      "loss": 0.6874,
      "step": 21
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.323943661971831,
      "eval_loss": 0.7139633893966675,
      "eval_runtime": 0.0309,
      "eval_samples_per_second": 2299.456,
      "eval_steps_per_second": 32.387,
      "step": 21
    },
    {
      "epoch": 7.0,
      "step": 21,
      "total_flos": 5531069537280.0,
      "train_loss": 0.6924250693548293,
      "train_runtime": 3.1238,
      "train_samples_per_second": 10163.884,
      "train_steps_per_second": 48.018
    }
  ],
  "logging_steps": 1,
  "max_steps": 150,
  "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": 5531069537280.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}