run-22/checkpoint-192/trainer_state.json

{
  "best_metric": 0.7984344422700587,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-22/checkpoint-192",
  "epoch": 2.0,
  "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": 2.464465618133545,
      "learning_rate": 0.0001794635836369996,
      "loss": 0.565,
      "step": 96
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7788649706457925,
      "eval_f1": 0.8078231292517007,
      "eval_loss": 0.4691426157951355,
      "eval_precision": 0.7142857142857143,
      "eval_recall": 0.9295499021526419,
      "eval_runtime": 25.0337,
      "eval_samples_per_second": 40.825,
      "eval_steps_per_second": 1.278,
      "step": 96
    },
    {
      "epoch": 2.0,
      "grad_norm": 4.053184509277344,
      "learning_rate": 0.00015703063568237466,
      "loss": 0.4676,
      "step": 192
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7984344422700587,
      "eval_f1": 0.8170515097690941,
      "eval_loss": 0.4367915391921997,
      "eval_precision": 0.7479674796747967,
      "eval_recall": 0.9001956947162426,
      "eval_runtime": 25.2941,
      "eval_samples_per_second": 40.405,
      "eval_steps_per_second": 1.265,
      "step": 192
    }
  ],
  "logging_steps": 500,
  "max_steps": 864,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "save_steps": 500,
  "total_flos": 471390394560.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.8956145079239534,
    "learning_rate": 0.00020189653159162458,
    "num_train_epochs": 9,
    "temperature": 8
  }
}