run-23/checkpoint-96/trainer_state.json

{
  "best_metric": 0.7896281800391389,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-23/checkpoint-96",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 96,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.5342754125595093,
      "learning_rate": 0.00022562537376582823,
      "loss": 0.5823,
      "step": 48
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7632093933463796,
      "eval_f1": 0.7962962962962963,
      "eval_loss": 0.48644715547561646,
      "eval_precision": 0.6986706056129985,
      "eval_recall": 0.9256360078277887,
      "eval_runtime": 26.8879,
      "eval_samples_per_second": 38.01,
      "eval_steps_per_second": 0.595,
      "step": 48
    },
    {
      "epoch": 2.0,
      "grad_norm": 5.700287342071533,
      "learning_rate": 0.0001504169158438855,
      "loss": 0.4818,
      "step": 96
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7896281800391389,
      "eval_f1": 0.7954329210275928,
      "eval_loss": 0.4546288549900055,
      "eval_precision": 0.774074074074074,
      "eval_recall": 0.8180039138943248,
      "eval_runtime": 26.9896,
      "eval_samples_per_second": 37.866,
      "eval_steps_per_second": 0.593,
      "step": 96
    }
  ],
  "logging_steps": 500,
  "max_steps": 192,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "total_flos": 471390394560.0,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.9259836100395409,
    "learning_rate": 0.000300833831687771,
    "num_train_epochs": 4,
    "temperature": 27
  }
}