{
  "best_global_step": 410,
  "best_metric": 0.6509972214698792,
  "best_model_checkpoint": "tiny_bert_km_100_v2_qnli/checkpoint-410",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 2460,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.396662712097168,
      "learning_rate": 4.9e-05,
      "loss": 0.6693,
      "step": 410
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6112026359143328,
      "eval_loss": 0.6509972214698792,
      "eval_runtime": 1.6178,
      "eval_samples_per_second": 3376.822,
      "eval_steps_per_second": 13.599,
      "step": 410
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.5475722551345825,
      "learning_rate": 4.8e-05,
      "loss": 0.6414,
      "step": 820
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.615595826468973,
      "eval_loss": 0.6562298536300659,
      "eval_runtime": 1.6093,
      "eval_samples_per_second": 3394.562,
      "eval_steps_per_second": 13.67,
      "step": 820
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.388838052749634,
      "learning_rate": 4.7e-05,
      "loss": 0.6031,
      "step": 1230
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6205381658429434,
      "eval_loss": 0.6521785259246826,
      "eval_runtime": 1.5873,
      "eval_samples_per_second": 3441.744,
      "eval_steps_per_second": 13.86,
      "step": 1230
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.951145648956299,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.5387,
      "step": 1640
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6108365367014461,
      "eval_loss": 0.6984445452690125,
      "eval_runtime": 1.5725,
      "eval_samples_per_second": 3474.069,
      "eval_steps_per_second": 13.99,
      "step": 1640
    },
    {
      "epoch": 5.0,
      "grad_norm": 3.9199023246765137,
      "learning_rate": 4.5e-05,
      "loss": 0.4639,
      "step": 2050
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6137653304045396,
      "eval_loss": 0.7590162754058838,
      "eval_runtime": 1.5543,
      "eval_samples_per_second": 3514.803,
      "eval_steps_per_second": 14.154,
      "step": 2050
    },
    {
      "epoch": 6.0,
      "grad_norm": 4.7078399658203125,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.3926,
      "step": 2460
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.603880651656599,
      "eval_loss": 0.9158124923706055,
      "eval_runtime": 1.5503,
      "eval_samples_per_second": 3523.835,
      "eval_steps_per_second": 14.191,
      "step": 2460
    },
    {
      "epoch": 6.0,
      "step": 2460,
      "total_flos": 1.6480417453111296e+16,
      "train_loss": 0.5515205445328379,
      "train_runtime": 340.597,
      "train_samples_per_second": 15376.384,
      "train_steps_per_second": 60.188
    }
  ],
  "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": 1.6480417453111296e+16,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}