trainer_state.json

{
  "best_metric": 0.7000991106033325,
  "best_model_checkpoint": "bert_tiny_lda_5_v1_book_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": 1.3964866399765015,
      "learning_rate": 4.9e-05,
      "loss": 0.7081,
      "step": 3
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.4225352112676056,
      "eval_loss": 0.7032899260520935,
      "eval_runtime": 0.0287,
      "eval_samples_per_second": 2474.701,
      "eval_steps_per_second": 34.855,
      "step": 3
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.9558442234992981,
      "learning_rate": 4.8e-05,
      "loss": 0.7026,
      "step": 6
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5633802816901409,
      "eval_loss": 0.7000991106033325,
      "eval_runtime": 0.0298,
      "eval_samples_per_second": 2383.852,
      "eval_steps_per_second": 33.575,
      "step": 6
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.478176474571228,
      "learning_rate": 4.7e-05,
      "loss": 0.6963,
      "step": 9
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.29577464788732394,
      "eval_loss": 0.716989278793335,
      "eval_runtime": 0.0321,
      "eval_samples_per_second": 2209.789,
      "eval_steps_per_second": 31.124,
      "step": 9
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.4897487759590149,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.6918,
      "step": 12
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.36619718309859156,
      "eval_loss": 0.7249119281768799,
      "eval_runtime": 0.0286,
      "eval_samples_per_second": 2486.541,
      "eval_steps_per_second": 35.022,
      "step": 12
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.6434975862503052,
      "learning_rate": 4.5e-05,
      "loss": 0.6921,
      "step": 15
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.28169014084507044,
      "eval_loss": 0.7210608720779419,
      "eval_runtime": 0.0283,
      "eval_samples_per_second": 2513.17,
      "eval_steps_per_second": 35.397,
      "step": 15
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.5997876524925232,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.6919,
      "step": 18
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.30985915492957744,
      "eval_loss": 0.724141538143158,
      "eval_runtime": 0.0288,
      "eval_samples_per_second": 2465.767,
      "eval_steps_per_second": 34.729,
      "step": 18
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.5266169309616089,
      "learning_rate": 4.3e-05,
      "loss": 0.6883,
      "step": 21
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.3380281690140845,
      "eval_loss": 0.7300835847854614,
      "eval_runtime": 0.0289,
      "eval_samples_per_second": 2458.459,
      "eval_steps_per_second": 34.626,
      "step": 21
    },
    {
      "epoch": 7.0,
      "step": 21,
      "total_flos": 116563804707840.0,
      "train_loss": 0.6958753835587275,
      "train_runtime": 8.7747,
      "train_samples_per_second": 3618.346,
      "train_steps_per_second": 17.095
    }
  ],
  "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": 116563804707840.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}