run-22/checkpoint-428/trainer_state.json

{
  "best_metric": 0.7082601054481547,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-22/checkpoint-428",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 428,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 8.196887969970703,
      "learning_rate": 6.743640293411462e-05,
      "loss": 0.5127,
      "step": 214
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7076742823667252,
      "eval_f1": 0.4674493062966916,
      "eval_loss": 0.48254257440567017,
      "eval_mcc": 0.29111772391065377,
      "eval_precision": 0.595108695652174,
      "eval_recall": 0.38488576449912126,
      "eval_runtime": 3.1636,
      "eval_samples_per_second": 539.577,
      "eval_steps_per_second": 17.069,
      "step": 214
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.403902053833008,
      "learning_rate": 0.0,
      "loss": 0.4749,
      "step": 428
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7082601054481547,
      "eval_f1": 0.4969696969696969,
      "eval_loss": 0.47061875462532043,
      "eval_mcc": 0.30463705772080735,
      "eval_precision": 0.5843230403800475,
      "eval_recall": 0.43233743409490333,
      "eval_runtime": 3.1948,
      "eval_samples_per_second": 534.299,
      "eval_steps_per_second": 16.902,
      "step": 428
    }
  ],
  "logging_steps": 500,
  "max_steps": 428,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "total_flos": 1049551328880.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.819893835685884,
    "learning_rate": 0.00013487280586822924,
    "num_train_epochs": 2,
    "temperature": 17
  }
}