{
  "best_metric": 0.6923967003822327,
  "best_model_checkpoint": "tiny_bert_rand_50_v1_wnli/checkpoint-6",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 21,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.956745445728302,
      "learning_rate": 4.9e-05,
      "loss": 0.6996,
      "step": 3
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5352112676056338,
      "eval_loss": 0.6930016279220581,
      "eval_runtime": 0.036,
      "eval_samples_per_second": 1973.842,
      "eval_steps_per_second": 27.801,
      "step": 3
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.7020816206932068,
      "learning_rate": 4.8e-05,
      "loss": 0.6965,
      "step": 6
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5633802816901409,
      "eval_loss": 0.6923967003822327,
      "eval_runtime": 0.0355,
      "eval_samples_per_second": 2000.79,
      "eval_steps_per_second": 28.18,
      "step": 6
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.49778902530670166,
      "learning_rate": 4.7e-05,
      "loss": 0.6961,
      "step": 9
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.36619718309859156,
      "eval_loss": 0.7064810991287231,
      "eval_runtime": 0.0772,
      "eval_samples_per_second": 919.852,
      "eval_steps_per_second": 12.956,
      "step": 9
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.6517056226730347,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.6933,
      "step": 12
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.352112676056338,
      "eval_loss": 0.704830527305603,
      "eval_runtime": 0.037,
      "eval_samples_per_second": 1921.187,
      "eval_steps_per_second": 27.059,
      "step": 12
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.8543090224266052,
      "learning_rate": 4.5e-05,
      "loss": 0.6941,
      "step": 15
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.4507042253521127,
      "eval_loss": 0.703125,
      "eval_runtime": 0.0372,
      "eval_samples_per_second": 1910.796,
      "eval_steps_per_second": 26.913,
      "step": 15
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.6381711959838867,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.6911,
      "step": 18
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.39436619718309857,
      "eval_loss": 0.7073062658309937,
      "eval_runtime": 0.044,
      "eval_samples_per_second": 1614.322,
      "eval_steps_per_second": 22.737,
      "step": 18
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.4603327214717865,
      "learning_rate": 4.3e-05,
      "loss": 0.6905,
      "step": 21
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.4084507042253521,
      "eval_loss": 0.710222065448761,
      "eval_runtime": 0.0359,
      "eval_samples_per_second": 1978.182,
      "eval_steps_per_second": 27.862,
      "step": 21
    },
    {
      "epoch": 7.0,
      "step": 21,
      "total_flos": 116563804707840.0,
      "train_loss": 0.6944480964115688,
      "train_runtime": 8.3963,
      "train_samples_per_second": 3781.441,
      "train_steps_per_second": 17.865
    }
  ],
  "logging_steps": 1,
  "max_steps": 150,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 50,
  "save_steps": 500,
  "total_flos": 116563804707840.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}