trainer_state.json

{
  "best_metric": 0.3492174744606018,
  "best_model_checkpoint": "distilbert_lda_20_v1_book_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": 4.688103199005127,
      "learning_rate": 4.9e-05,
      "loss": 0.4872,
      "step": 410
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8314113124656782,
      "eval_loss": 0.3770684599876404,
      "eval_runtime": 2.4665,
      "eval_samples_per_second": 2214.88,
      "eval_steps_per_second": 8.92,
      "step": 410
    },
    {
      "epoch": 2.0,
      "grad_norm": 4.475072860717773,
      "learning_rate": 4.8e-05,
      "loss": 0.3637,
      "step": 820
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8451400329489291,
      "eval_loss": 0.3492174744606018,
      "eval_runtime": 2.3745,
      "eval_samples_per_second": 2300.696,
      "eval_steps_per_second": 9.265,
      "step": 820
    },
    {
      "epoch": 3.0,
      "grad_norm": 4.083785533905029,
      "learning_rate": 4.7e-05,
      "loss": 0.2718,
      "step": 1230
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8491671242906828,
      "eval_loss": 0.3612353801727295,
      "eval_runtime": 2.3979,
      "eval_samples_per_second": 2278.287,
      "eval_steps_per_second": 9.175,
      "step": 1230
    },
    {
      "epoch": 4.0,
      "grad_norm": 8.397468566894531,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.1905,
      "step": 1640
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8213435841112942,
      "eval_loss": 0.5011342763900757,
      "eval_runtime": 2.4627,
      "eval_samples_per_second": 2218.3,
      "eval_steps_per_second": 8.933,
      "step": 1640
    },
    {
      "epoch": 5.0,
      "grad_norm": 5.871656894683838,
      "learning_rate": 4.5e-05,
      "loss": 0.1298,
      "step": 2050
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8336079077429983,
      "eval_loss": 0.509253740310669,
      "eval_runtime": 2.4621,
      "eval_samples_per_second": 2218.835,
      "eval_steps_per_second": 8.935,
      "step": 2050
    },
    {
      "epoch": 6.0,
      "grad_norm": 7.910794734954834,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.0919,
      "step": 2460
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8411129416071755,
      "eval_loss": 0.6079259514808655,
      "eval_runtime": 2.4735,
      "eval_samples_per_second": 2208.609,
      "eval_steps_per_second": 8.894,
      "step": 2460
    },
    {
      "epoch": 7.0,
      "grad_norm": 5.707584857940674,
      "learning_rate": 4.3e-05,
      "loss": 0.0682,
      "step": 2870
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8242723778143877,
      "eval_loss": 0.6701850295066833,
      "eval_runtime": 2.371,
      "eval_samples_per_second": 2304.135,
      "eval_steps_per_second": 9.279,
      "step": 2870
    },
    {
      "epoch": 7.0,
      "step": 2870,
      "total_flos": 4.856261458098893e+16,
      "train_loss": 0.2290068164519732,
      "train_runtime": 731.2636,
      "train_samples_per_second": 7161.781,
      "train_steps_per_second": 28.034
    }
  ],
  "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
}