run-17/checkpoint-192/trainer_state.json

{
  "best_metric": 0.8013698630136986,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-17/checkpoint-192",
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 192,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.7164396047592163,
      "learning_rate": 0.0007649073963933848,
      "loss": 0.5715,
      "step": 48
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7397260273972602,
      "eval_f1": 0.7861736334405145,
      "eval_loss": 0.5001023411750793,
      "eval_precision": 0.6671214188267395,
      "eval_recall": 0.9569471624266145,
      "eval_runtime": 27.2812,
      "eval_samples_per_second": 37.462,
      "eval_steps_per_second": 0.586,
      "step": 48
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.8946006298065186,
      "learning_rate": 0.0005736805472950385,
      "loss": 0.4948,
      "step": 96
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7455968688845401,
      "eval_f1": 0.7893030794165317,
      "eval_loss": 0.4874393045902252,
      "eval_precision": 0.673582295988935,
      "eval_recall": 0.9530332681017613,
      "eval_runtime": 27.1224,
      "eval_samples_per_second": 37.681,
      "eval_steps_per_second": 0.59,
      "step": 96
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.64782452583313,
      "learning_rate": 0.0003824536981966924,
      "loss": 0.4839,
      "step": 144
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7661448140900196,
      "eval_f1": 0.8023159636062862,
      "eval_loss": 0.4675934612751007,
      "eval_precision": 0.6948424068767909,
      "eval_recall": 0.949119373776908,
      "eval_runtime": 26.9999,
      "eval_samples_per_second": 37.852,
      "eval_steps_per_second": 0.593,
      "step": 144
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.7006142139434814,
      "learning_rate": 0.0001912268490983462,
      "loss": 0.4531,
      "step": 192
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8013698630136986,
      "eval_f1": 0.8031037827352084,
      "eval_loss": 0.4373694658279419,
      "eval_precision": 0.7961538461538461,
      "eval_recall": 0.8101761252446184,
      "eval_runtime": 27.2608,
      "eval_samples_per_second": 37.49,
      "eval_steps_per_second": 0.587,
      "step": 192
    }
  ],
  "logging_steps": 500,
  "max_steps": 240,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 942780789120.0,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.8956137687880612,
    "learning_rate": 0.000956134245491731,
    "num_train_epochs": 5,
    "temperature": 12
  }
}