{
  "best_global_step": 5688,
  "best_metric": 0.4227236807346344,
  "best_model_checkpoint": "tiny_bert_km_20_v2_qqp/checkpoint-5688",
  "epoch": 9.0,
  "eval_steps": 500,
  "global_step": 12798,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.547630786895752,
      "learning_rate": 4.9e-05,
      "loss": 0.497,
      "step": 1422
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7812515458817709,
      "eval_combined_score": 0.7269934461202816,
      "eval_f1": 0.6727353463587922,
      "eval_loss": 0.4555601477622986,
      "eval_runtime": 11.612,
      "eval_samples_per_second": 3481.747,
      "eval_steps_per_second": 13.607,
      "step": 1422
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.2651309967041016,
      "learning_rate": 4.8e-05,
      "loss": 0.4091,
      "step": 2844
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8003215434083601,
      "eval_combined_score": 0.7544965950234396,
      "eval_f1": 0.708671646638519,
      "eval_loss": 0.42515942454338074,
      "eval_runtime": 11.4412,
      "eval_samples_per_second": 3533.723,
      "eval_steps_per_second": 13.81,
      "step": 2844
    },
    {
      "epoch": 3.0,
      "grad_norm": 3.216604232788086,
      "learning_rate": 4.7e-05,
      "loss": 0.3508,
      "step": 4266
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8117981696759832,
      "eval_combined_score": 0.7715736249643906,
      "eval_f1": 0.7313490802527981,
      "eval_loss": 0.4284079074859619,
      "eval_runtime": 11.4348,
      "eval_samples_per_second": 3535.707,
      "eval_steps_per_second": 13.818,
      "step": 4266
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.6022121906280518,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.3022,
      "step": 5688
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8167697254513975,
      "eval_combined_score": 0.7768227401143468,
      "eval_f1": 0.7368757547772963,
      "eval_loss": 0.4227236807346344,
      "eval_runtime": 11.4203,
      "eval_samples_per_second": 3540.181,
      "eval_steps_per_second": 13.835,
      "step": 5688
    },
    {
      "epoch": 5.0,
      "grad_norm": 3.1953468322753906,
      "learning_rate": 4.5e-05,
      "loss": 0.2622,
      "step": 7110
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8211723967350977,
      "eval_combined_score": 0.7873671650319432,
      "eval_f1": 0.7535619333287886,
      "eval_loss": 0.43976229429244995,
      "eval_runtime": 11.4345,
      "eval_samples_per_second": 3535.798,
      "eval_steps_per_second": 13.818,
      "step": 7110
    },
    {
      "epoch": 6.0,
      "grad_norm": 2.6762185096740723,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.2268,
      "step": 8532
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8229285184269107,
      "eval_combined_score": 0.7912378796077724,
      "eval_f1": 0.759547240788634,
      "eval_loss": 0.4587453305721283,
      "eval_runtime": 11.4824,
      "eval_samples_per_second": 3521.049,
      "eval_steps_per_second": 13.76,
      "step": 8532
    },
    {
      "epoch": 7.0,
      "grad_norm": 4.207704067230225,
      "learning_rate": 4.3e-05,
      "loss": 0.2005,
      "step": 9954
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8228048478852338,
      "eval_combined_score": 0.7915229058703277,
      "eval_f1": 0.7602409638554217,
      "eval_loss": 0.4909394085407257,
      "eval_runtime": 11.6006,
      "eval_samples_per_second": 3485.161,
      "eval_steps_per_second": 13.62,
      "step": 9954
    },
    {
      "epoch": 8.0,
      "grad_norm": 3.09031081199646,
      "learning_rate": 4.2e-05,
      "loss": 0.1759,
      "step": 11376
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8241899579520159,
      "eval_combined_score": 0.7929931023004047,
      "eval_f1": 0.7617962466487935,
      "eval_loss": 0.5125080943107605,
      "eval_runtime": 11.475,
      "eval_samples_per_second": 3523.301,
      "eval_steps_per_second": 13.769,
      "step": 11376
    },
    {
      "epoch": 9.0,
      "grad_norm": 4.291172981262207,
      "learning_rate": 4.1e-05,
      "loss": 0.1564,
      "step": 12798
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.8236210734603018,
      "eval_combined_score": 0.7958329158083151,
      "eval_f1": 0.7680447581563282,
      "eval_loss": 0.5786588788032532,
      "eval_runtime": 11.4812,
      "eval_samples_per_second": 3521.414,
      "eval_steps_per_second": 13.762,
      "step": 12798
    },
    {
      "epoch": 9.0,
      "step": 12798,
      "total_flos": 8.587209601564877e+16,
      "train_loss": 0.2867661023218197,
      "train_runtime": 1805.0479,
      "train_samples_per_second": 10078.569,
      "train_steps_per_second": 39.39
    }
  ],
  "logging_steps": 1,
  "max_steps": 71100,
  "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": 8.587209601564877e+16,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}