run-6/checkpoint-856/trainer_state.json

{
  "best_metric": 0.6666666666666666,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-6/checkpoint-214",
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 856,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.7624236345291138,
      "learning_rate": 1.5910505716425373e-05,
      "loss": 0.3605,
      "step": 214
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6666666666666666,
      "eval_f1": 0.0,
      "eval_loss": 0.3405539095401764,
      "eval_mcc": 0.0,
      "eval_precision": 0.0,
      "eval_recall": 0.0,
      "eval_runtime": 3.1408,
      "eval_samples_per_second": 543.484,
      "eval_steps_per_second": 17.193,
      "step": 214
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.9066030383110046,
      "learning_rate": 1.0607003810950248e-05,
      "loss": 0.3392,
      "step": 428
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6666666666666666,
      "eval_f1": 0.0,
      "eval_loss": 0.331169456243515,
      "eval_mcc": 0.0,
      "eval_precision": 0.0,
      "eval_recall": 0.0,
      "eval_runtime": 3.1414,
      "eval_samples_per_second": 543.396,
      "eval_steps_per_second": 17.19,
      "step": 428
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.5507936477661133,
      "learning_rate": 5.303501905475124e-06,
      "loss": 0.333,
      "step": 642
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6666666666666666,
      "eval_f1": 0.0,
      "eval_loss": 0.3254913091659546,
      "eval_mcc": 0.0,
      "eval_precision": 0.0,
      "eval_recall": 0.0,
      "eval_runtime": 3.1383,
      "eval_samples_per_second": 543.924,
      "eval_steps_per_second": 17.207,
      "step": 642
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.9257184267044067,
      "learning_rate": 0.0,
      "loss": 0.3297,
      "step": 856
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6666666666666666,
      "eval_f1": 0.0,
      "eval_loss": 0.3236803710460663,
      "eval_mcc": 0.0,
      "eval_precision": 0.0,
      "eval_recall": 0.0,
      "eval_runtime": 3.1403,
      "eval_samples_per_second": 543.576,
      "eval_steps_per_second": 17.196,
      "step": 856
    }
  ],
  "logging_steps": 500,
  "max_steps": 856,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "total_flos": 2099102657760.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.5123346903654048,
    "learning_rate": 2.1214007621900497e-05,
    "num_train_epochs": 4,
    "temperature": 10
  }
}