trainer_state.json

{
  "best_metric": 0.6964953541755676,
  "best_model_checkpoint": "bert_tiny_lda_50_v1_book_wnli/checkpoint-3",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 18,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.8155308365821838,
      "learning_rate": 4.9e-05,
      "loss": 0.7018,
      "step": 3
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5633802816901409,
      "eval_loss": 0.6964953541755676,
      "eval_runtime": 0.029,
      "eval_samples_per_second": 2450.307,
      "eval_steps_per_second": 34.511,
      "step": 3
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.7931452989578247,
      "learning_rate": 4.8e-05,
      "loss": 0.6999,
      "step": 6
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.39436619718309857,
      "eval_loss": 0.7075263857841492,
      "eval_runtime": 0.029,
      "eval_samples_per_second": 2447.93,
      "eval_steps_per_second": 34.478,
      "step": 6
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.8391414284706116,
      "learning_rate": 4.7e-05,
      "loss": 0.6971,
      "step": 9
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.4225352112676056,
      "eval_loss": 0.7252969145774841,
      "eval_runtime": 0.0293,
      "eval_samples_per_second": 2419.391,
      "eval_steps_per_second": 34.076,
      "step": 9
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.7865285873413086,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.6921,
      "step": 12
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.4225352112676056,
      "eval_loss": 0.7110476493835449,
      "eval_runtime": 0.0297,
      "eval_samples_per_second": 2387.656,
      "eval_steps_per_second": 33.629,
      "step": 12
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.0633721351623535,
      "learning_rate": 4.5e-05,
      "loss": 0.6941,
      "step": 15
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.39436619718309857,
      "eval_loss": 0.716549277305603,
      "eval_runtime": 0.0288,
      "eval_samples_per_second": 2463.482,
      "eval_steps_per_second": 34.697,
      "step": 15
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.7382026314735413,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.689,
      "step": 18
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.3380281690140845,
      "eval_loss": 0.7329993844032288,
      "eval_runtime": 0.029,
      "eval_samples_per_second": 2448.071,
      "eval_steps_per_second": 34.48,
      "step": 18
    },
    {
      "epoch": 6.0,
      "step": 18,
      "total_flos": 99911832606720.0,
      "train_loss": 0.6956634124120077,
      "train_runtime": 7.7222,
      "train_samples_per_second": 4111.506,
      "train_steps_per_second": 19.424
    }
  ],
  "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": 99911832606720.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}