{
  "best_global_step": 6136,
  "best_metric": 0.7564880847930908,
  "best_model_checkpoint": "tiny_bert_km_5_v1_mnli/checkpoint-6136",
  "epoch": 9.0,
  "eval_steps": 500,
  "global_step": 13806,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.197121262550354,
      "learning_rate": 4.9e-05,
      "loss": 0.9825,
      "step": 1534
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5957208354559348,
      "eval_loss": 0.8823983669281006,
      "eval_runtime": 2.813,
      "eval_samples_per_second": 3489.144,
      "eval_steps_per_second": 13.864,
      "step": 1534
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.8529736995697021,
      "learning_rate": 4.8e-05,
      "loss": 0.8526,
      "step": 3068
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6437086092715232,
      "eval_loss": 0.8074957728385925,
      "eval_runtime": 2.7722,
      "eval_samples_per_second": 3540.56,
      "eval_steps_per_second": 14.068,
      "step": 3068
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.7989853620529175,
      "learning_rate": 4.7e-05,
      "loss": 0.775,
      "step": 4602
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6681609780947529,
      "eval_loss": 0.7722101807594299,
      "eval_runtime": 2.7687,
      "eval_samples_per_second": 3545.029,
      "eval_steps_per_second": 14.086,
      "step": 4602
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.2518186569213867,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.7197,
      "step": 6136
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6725420275089149,
      "eval_loss": 0.7564880847930908,
      "eval_runtime": 2.7713,
      "eval_samples_per_second": 3541.681,
      "eval_steps_per_second": 14.073,
      "step": 6136
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.1857810020446777,
      "learning_rate": 4.5e-05,
      "loss": 0.6704,
      "step": 7670
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6799796230259807,
      "eval_loss": 0.7602323293685913,
      "eval_runtime": 2.7824,
      "eval_samples_per_second": 3527.502,
      "eval_steps_per_second": 14.017,
      "step": 7670
    },
    {
      "epoch": 6.0,
      "grad_norm": 2.1004514694213867,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.6234,
      "step": 9204
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6842587875700459,
      "eval_loss": 0.7736557126045227,
      "eval_runtime": 2.7697,
      "eval_samples_per_second": 3543.731,
      "eval_steps_per_second": 14.081,
      "step": 9204
    },
    {
      "epoch": 7.0,
      "grad_norm": 2.3393008708953857,
      "learning_rate": 4.3e-05,
      "loss": 0.5804,
      "step": 10738
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.6852776362710138,
      "eval_loss": 0.7831186652183533,
      "eval_runtime": 2.7714,
      "eval_samples_per_second": 3541.587,
      "eval_steps_per_second": 14.073,
      "step": 10738
    },
    {
      "epoch": 8.0,
      "grad_norm": 2.4777815341949463,
      "learning_rate": 4.2e-05,
      "loss": 0.538,
      "step": 12272
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.6861946001018848,
      "eval_loss": 0.811160147190094,
      "eval_runtime": 2.7764,
      "eval_samples_per_second": 3535.118,
      "eval_steps_per_second": 14.047,
      "step": 12272
    },
    {
      "epoch": 9.0,
      "grad_norm": 3.404972553253174,
      "learning_rate": 4.1e-05,
      "loss": 0.4956,
      "step": 13806
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.6777381558838512,
      "eval_loss": 0.9172862768173218,
      "eval_runtime": 2.7785,
      "eval_samples_per_second": 3532.473,
      "eval_steps_per_second": 14.036,
      "step": 13806
    },
    {
      "epoch": 9.0,
      "step": 13806,
      "total_flos": 9.268524907513344e+16,
      "train_loss": 0.6930639208114339,
      "train_runtime": 1871.0512,
      "train_samples_per_second": 10494.154,
      "train_steps_per_second": 40.993
    }
  ],
  "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": 9.268524907513344e+16,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}