run-1/checkpoint-384/trainer_state.json

{
  "best_metric": 0.538160469667319,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-1/checkpoint-384",
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 384,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.40490686893463135,
      "learning_rate": 1.4953262133881219e-05,
      "loss": 0.1703,
      "step": 96
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5,
      "eval_f1": 0.0,
      "eval_loss": 0.11485283076763153,
      "eval_precision": 0.0,
      "eval_recall": 0.0,
      "eval_runtime": 25.7126,
      "eval_samples_per_second": 39.747,
      "eval_steps_per_second": 1.245,
      "step": 96
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.2503454089164734,
      "learning_rate": 1.2817081829041045e-05,
      "loss": 0.1095,
      "step": 192
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5,
      "eval_f1": 0.0038986354775828462,
      "eval_loss": 0.0978715717792511,
      "eval_precision": 0.5,
      "eval_recall": 0.0019569471624266144,
      "eval_runtime": 25.921,
      "eval_samples_per_second": 39.427,
      "eval_steps_per_second": 1.235,
      "step": 192
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.29292032122612,
      "learning_rate": 1.0680901524200872e-05,
      "loss": 0.0978,
      "step": 288
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5352250489236791,
      "eval_f1": 0.18244406196213422,
      "eval_loss": 0.0902138277888298,
      "eval_precision": 0.7571428571428571,
      "eval_recall": 0.10371819960861056,
      "eval_runtime": 25.3689,
      "eval_samples_per_second": 40.286,
      "eval_steps_per_second": 1.261,
      "step": 288
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.24781617522239685,
      "learning_rate": 8.544721219360697e-06,
      "loss": 0.0933,
      "step": 384
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.538160469667319,
      "eval_f1": 0.1945392491467577,
      "eval_loss": 0.08735258132219315,
      "eval_precision": 0.76,
      "eval_recall": 0.11154598825831702,
      "eval_runtime": 25.5717,
      "eval_samples_per_second": 39.966,
      "eval_steps_per_second": 1.251,
      "step": 384
    }
  ],
  "logging_steps": 500,
  "max_steps": 768,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "save_steps": 500,
  "total_flos": 942780789120.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.09656764013507046,
    "learning_rate": 1.7089442438721394e-05,
    "num_train_epochs": 8,
    "temperature": 20
  }
}