run-30/checkpoint-428/trainer_state.json

{
  "best_metric": 0.7299355594610427,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-30/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": 6.404624938964844,
      "learning_rate": 0.0006827366491550371,
      "loss": 0.5674,
      "step": 214
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.70298769771529,
      "eval_f1": 0.47461139896373056,
      "eval_loss": 0.5547006130218506,
      "eval_mcc": 0.28558083057134703,
      "eval_precision": 0.5782828282828283,
      "eval_recall": 0.4024604569420035,
      "eval_runtime": 3.121,
      "eval_samples_per_second": 546.936,
      "eval_steps_per_second": 17.302,
      "step": 214
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.9503083229064941,
      "learning_rate": 0.0005973945680106575,
      "loss": 0.518,
      "step": 428
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7299355594610427,
      "eval_f1": 0.5252317198764161,
      "eval_loss": 0.5166522860527039,
      "eval_mcc": 0.35438337904683964,
      "eval_precision": 0.6343283582089553,
      "eval_recall": 0.44815465729349735,
      "eval_runtime": 3.144,
      "eval_samples_per_second": 542.932,
      "eval_steps_per_second": 17.175,
      "step": 428
    }
  ],
  "logging_steps": 500,
  "max_steps": 1926,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "save_steps": 500,
  "total_flos": 1049551328880.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.9213962687632572,
    "learning_rate": 0.0007680787302994168,
    "num_train_epochs": 9,
    "temperature": 25
  }
}