{
  "best_global_step": 7670,
  "best_metric": 0.8323174118995667,
  "best_model_checkpoint": "tiny_bert_km_20_v2_mnli/checkpoint-7670",
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 15340,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.1402794122695923,
      "learning_rate": 4.9e-05,
      "loss": 0.9993,
      "step": 1534
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5527254202750892,
      "eval_loss": 0.9375848770141602,
      "eval_runtime": 2.866,
      "eval_samples_per_second": 3424.627,
      "eval_steps_per_second": 13.608,
      "step": 1534
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.3853436708450317,
      "learning_rate": 4.8e-05,
      "loss": 0.9067,
      "step": 3068
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5892002037697401,
      "eval_loss": 0.8823896646499634,
      "eval_runtime": 2.7936,
      "eval_samples_per_second": 3513.424,
      "eval_steps_per_second": 13.961,
      "step": 3068
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.4370675086975098,
      "learning_rate": 4.7e-05,
      "loss": 0.8527,
      "step": 4602
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6076413652572593,
      "eval_loss": 0.8542468547821045,
      "eval_runtime": 2.8078,
      "eval_samples_per_second": 3495.574,
      "eval_steps_per_second": 13.89,
      "step": 4602
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.9635460376739502,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.8046,
      "step": 6136
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.621395822720326,
      "eval_loss": 0.8475756645202637,
      "eval_runtime": 2.7922,
      "eval_samples_per_second": 3515.198,
      "eval_steps_per_second": 13.968,
      "step": 6136
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.8043975830078125,
      "learning_rate": 4.5e-05,
      "loss": 0.7574,
      "step": 7670
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6300560366785533,
      "eval_loss": 0.8323174118995667,
      "eval_runtime": 2.7823,
      "eval_samples_per_second": 3527.703,
      "eval_steps_per_second": 14.017,
      "step": 7670
    },
    {
      "epoch": 6.0,
      "grad_norm": 2.153754711151123,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.7096,
      "step": 9204
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6352521650534896,
      "eval_loss": 0.8395307064056396,
      "eval_runtime": 2.8595,
      "eval_samples_per_second": 3432.38,
      "eval_steps_per_second": 13.639,
      "step": 9204
    },
    {
      "epoch": 7.0,
      "grad_norm": 2.124469041824341,
      "learning_rate": 4.3e-05,
      "loss": 0.666,
      "step": 10738
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.6446255731023943,
      "eval_loss": 0.8471359014511108,
      "eval_runtime": 2.8009,
      "eval_samples_per_second": 3504.256,
      "eval_steps_per_second": 13.924,
      "step": 10738
    },
    {
      "epoch": 8.0,
      "grad_norm": 2.1384940147399902,
      "learning_rate": 4.2e-05,
      "loss": 0.6211,
      "step": 12272
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.6423841059602649,
      "eval_loss": 0.876538097858429,
      "eval_runtime": 2.7815,
      "eval_samples_per_second": 3528.638,
      "eval_steps_per_second": 14.021,
      "step": 12272
    },
    {
      "epoch": 9.0,
      "grad_norm": 2.8161027431488037,
      "learning_rate": 4.1e-05,
      "loss": 0.5794,
      "step": 13806
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.6421803362200713,
      "eval_loss": 0.9186814427375793,
      "eval_runtime": 2.8072,
      "eval_samples_per_second": 3496.388,
      "eval_steps_per_second": 13.893,
      "step": 13806
    },
    {
      "epoch": 10.0,
      "grad_norm": 3.022850751876831,
      "learning_rate": 4e-05,
      "loss": 0.539,
      "step": 15340
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.6424859908303617,
      "eval_loss": 0.9312591552734375,
      "eval_runtime": 2.8099,
      "eval_samples_per_second": 3493.058,
      "eval_steps_per_second": 13.88,
      "step": 15340
    },
    {
      "epoch": 10.0,
      "step": 15340,
      "total_flos": 1.029836100834816e+17,
      "train_loss": 0.7435631265851633,
      "train_runtime": 2079.7547,
      "train_samples_per_second": 9441.065,
      "train_steps_per_second": 36.879
    }
  ],
  "logging_steps": 1,
  "max_steps": 76700,
  "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": 1.029836100834816e+17,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}