{
  "best_global_step": 2844,
  "best_metric": 0.3963772654533386,
  "best_model_checkpoint": "bert_base_km_10_v2_qqp/checkpoint-2844",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 9954,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.4397708177566528,
      "learning_rate": 4.9e-05,
      "loss": 0.4814,
      "step": 1422
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7895622062824635,
      "eval_combined_score": 0.7323898983771456,
      "eval_f1": 0.6752175904718277,
      "eval_loss": 0.43921375274658203,
      "eval_runtime": 25.8733,
      "eval_samples_per_second": 1562.618,
      "eval_steps_per_second": 6.107,
      "step": 1422
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.1709325313568115,
      "learning_rate": 4.8e-05,
      "loss": 0.3764,
      "step": 2844
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8140737076428395,
      "eval_combined_score": 0.7833653721327619,
      "eval_f1": 0.7526570366226844,
      "eval_loss": 0.3963772654533386,
      "eval_runtime": 25.3306,
      "eval_samples_per_second": 1596.09,
      "eval_steps_per_second": 6.238,
      "step": 2844
    },
    {
      "epoch": 3.0,
      "grad_norm": 3.0599889755249023,
      "learning_rate": 4.7e-05,
      "loss": 0.2953,
      "step": 4266
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8251051199604255,
      "eval_combined_score": 0.7957035663589019,
      "eval_f1": 0.7663020127573784,
      "eval_loss": 0.40289756655693054,
      "eval_runtime": 25.5862,
      "eval_samples_per_second": 1580.146,
      "eval_steps_per_second": 6.175,
      "step": 4266
    },
    {
      "epoch": 4.0,
      "grad_norm": 3.135880947113037,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.2253,
      "step": 5688
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8326737571110562,
      "eval_combined_score": 0.7987869219960146,
      "eval_f1": 0.7649000868809731,
      "eval_loss": 0.4317234754562378,
      "eval_runtime": 25.1067,
      "eval_samples_per_second": 1610.326,
      "eval_steps_per_second": 6.293,
      "step": 5688
    },
    {
      "epoch": 5.0,
      "grad_norm": 3.71781587600708,
      "learning_rate": 4.5e-05,
      "loss": 0.17,
      "step": 7110
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8364580756863715,
      "eval_combined_score": 0.8038662316995577,
      "eval_f1": 0.7712743877127439,
      "eval_loss": 0.5111433863639832,
      "eval_runtime": 25.5197,
      "eval_samples_per_second": 1584.268,
      "eval_steps_per_second": 6.191,
      "step": 7110
    },
    {
      "epoch": 6.0,
      "grad_norm": 3.7161002159118652,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.1295,
      "step": 8532
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8370764283947564,
      "eval_combined_score": 0.8060975327593938,
      "eval_f1": 0.7751186371240313,
      "eval_loss": 0.5110228061676025,
      "eval_runtime": 25.2963,
      "eval_samples_per_second": 1598.256,
      "eval_steps_per_second": 6.246,
      "step": 8532
    },
    {
      "epoch": 7.0,
      "grad_norm": 3.5093371868133545,
      "learning_rate": 4.3e-05,
      "loss": 0.1021,
      "step": 9954
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8337620578778135,
      "eval_combined_score": 0.8055066213559684,
      "eval_f1": 0.7772511848341233,
      "eval_loss": 0.5918472409248352,
      "eval_runtime": 25.242,
      "eval_samples_per_second": 1601.696,
      "eval_steps_per_second": 6.259,
      "step": 9954
    },
    {
      "epoch": 7.0,
      "step": 9954,
      "total_flos": 3.3506166766980096e+17,
      "train_loss": 0.25428547398354695,
      "train_runtime": 4254.8392,
      "train_samples_per_second": 4275.673,
      "train_steps_per_second": 16.71
    }
  ],
  "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": 3.3506166766980096e+17,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}